Statistics and State-Society Relations in the Early People's Republic of China, 1949-1959

Home , Chen Xian (politician), History of Uzbekistan, Hong Liangji, Liaoxi, Ma Yinchu, Music of Heilongjiang

Making it Count: Statistics and State-Society Relations in the Early People’s Republic of China, 1949-1959

Arunabh Ghosh

Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the Graduate School of Arts and Sciences

COLUMBIA UNIVERSITY

2014

ABSTRACT

Making it Count: Statistics and State-Society Relations in the Early People’s Republic of China, 1949-1959

Arunabh Ghosh

This dissertation offers new perspectives on China’s transition to socialism by investigating a fundamental question—how did the state build capacity to know the nation through numbers? With the establishment of the People’s Republic in 1949, jubilant Chinese revolutionaries were confronted by the dual challenge of a nearly nonexistent statistical infrastructure and the pressing need to escape the universalist claims of capitalist statistics. At stake for revolutionary statisticians and economists was a fundamental difficulty: how to accurately ascertain social scientific fact. Resolving this difficulty involved not just epistemological and theoretical debates on the unity or disunity of statistical science but also practical considerations surrounding state-capacity building. The resultant shift toward a socialist definition of statistics, achieved by explicitly following the Soviet Union’s example, was instrumental in shaping new bureaus, designing statistical work, and training personnel. New classificatory schemes and methods of data collection also raised issues of authority and policy, ultimately not just remolding state-society relations but also informing new conceptions of everyday life and work. By the mid-1950s, however, growing disaffection with the efficacy of Soviet methods led the Chinese, in a surprising turn of events, to seek out Indian statisticians in an unprecedented instance of Chinese participation in South-South scientific exchange.

At the heart of these exchanges was the desire to learn more about large-scale random sampling, an emergent statistical technology, which, while technically complex, held great practical salience for large countries like China and India.

“Making it Count” engages with and contributes to scholarship on the history of modern China and on the global and Cold War histories of science and social science.

While the historiography on statistics and quantification has focused primarily on the early-modern and nineteenth century world, the dissertation brings this history into the twentieth century, when states, multi-national institutions, and private actors, regardless of their ideological hue, mobilized statistics on behalf of positivist social science and statecraft. By examining the collection and deployment of data, a process critical to the ambitions of the revolutionary PRC state but one that has largely been overlooked in the historical literature, the dissertation also provides an alternative account for a decade often portrayed as lurching from one mass campaign to another. Finally, the examination of the Sino-Indian statistical links reveals that pioneering innovation took place in many contexts after 1945 and challenges Cold War paradigms that are predisposed to assume the United States or the Soviet Union as the primary nodes from which scientific and other forms of modern knowledge emanated.

CONTENTS

Charts, Graphs, and Illustrations ii Acronyms iii Acknowledgments iv

I. Introduction 1

PART I: A Crisis in Counting

II. Revolutionary Statistics in the early 1950s 22

III. Lies, Damned Lies, and (Bourgeois) Statistics—Ascertaining 60 Social Fact

IV. No “Mean” Solution: The Reformulation of Statistical Science 99

PART II: Statistics in Action

V. The Nature of Statistical Work 144

VI. To “Ardently Love Statistical Work”: State (In-) capacity, 190 Professionalization, and their Discontents

VII. Statistical State, Quantified Society 238

PART III: Alternatives

VIII. The Turn to India: Sino-Indian Statistical Exchanges, 1951-1959 268

IX. Epilogue: A ‘Great Leap’ in Statistics 337

REFERENCE MATTER Bibliography 347

CHARTS, GRAPHS, AND ILLUSTRATIONS

TABLES

3.1 Statistics and Socialist Statistics 78

3.2 Soviet Statistical Experts at the SSB 89

3.3 Collected Translations in Statistics 95

4.1 Frequency of Articles Criticizing “Bourgeois” 114 Statistics, 1949-1958

4.2 Introductory Statistics Textbooks 126

4.3 Comparison of Statistics Textbooks 128

4.4 Industrial and Agricultural Production 139

4.5 Growth in Modern Industry 140

5.1 Establishment of National, Regional, Provincial, 154 And City Level Statistical Bureaus

5.2 Composition of Central and Survey Team 169

5.3 Excess Forms at Two Nanjing Factories 171

6.1 Renmin University Statistics Graduates/Enrollees 202

6.2 Establishment of Dedicated Statistical Schools 204 or Departments

6.3 BFES—Early Characteristics 206

CHARTS AND ILLUSTRATIONS

5.1 The State Statistical System 160

5.2 Annual Report Cover Page 161

5.3 The Dual Track Reporting System 177

ACRONYMS

BFES Beijing Finance and Economics School BMA Beijing Municipal Archives DSJ Dashiji (Chronicle of Events in Statistics in China) (Zhonghua renmin gongheguo tongji dashiji, 1949-2009 《中华人民共和国统计大事记》) NAI National Archives of India NMML Nehru Memorial Museum and Library NSB Northeast Statistics Bureau PCMMMA P.C. Mahalanobis Memorial Museum and Archives, Kolkata RMRB Renmin Ribao (People’s Daily) RWTG Re’ai women the tongji gongzuo (Ardently Love our Statistcal Work 《热爱我们的统计工作》 SSB State Statistics Bureau TJGZ Tongji gongzuo (Statistical Work) TJGZTX Tongji gongzuo tongxun (Statistical Work Bulletin) WJB Waijiao bu (Foreign Ministry Archives), Beijing

iii

ACKNOWLEDGEMENTS

I have incurred more debts writing this dissertation than I can hope to repay in one lifetime. Madeleine Zelin took me under her wing when I was a confused MA student.

Her unstinting support, encouragement, and intellectual guidance over the years have significantly shaped my own thinking and approach to Chinese history. Eugenia Lean, as near as possible to a second advisor, has been a constant guide and influence in the classroom and outside ever since I was a student in her modern China seminar several years ago. Jacob Eyferth’s generosity and infectious enthusiasm are only outpaced by the depth of his knowledge of the early People’s Republic. I am grateful for the interest he has taken in my work and hope to keep learning from our ongoing conversations. The

India part of this story would not have been possible without the interest, insights, and advice of Partha Chatterjee, who also generously made available to me his considerable network of friends and colleagues in Kolkata. Matt Connelly has been a source of inspiration right from when I took his seminar on population control and encountered a history of the twentieth century that was defined as much by transnational connections as other factors. I am grateful for the interest, feedback, and encouragement I have received over the years from Lydia Liu, Matt Jones, Betsy Blackmar, Dorothy Ko, Gray Tuttle, and Ken Prewitt. Lydia Liu was also instrumental in helping me find an intellectual and institutional home at Tsinghua University in Beijing during my year of fieldwork in

China. Bill McAllister, the Director of the Mellon Interdisciplinary Graduate Fellows

Program at INCITE, set a tremendous example of fellowship and intellectual commitment, which I shall endeavor to emulate. A special word of thanks to Bob Hymes,

iv who put the EALAC department firmly in my corner as I took my first tentative steps in

East Asian studies.

I first became interested in Chinese history in Paul Smith’s seminars at Haverford

College. Paul’s friendship and encouragement over the years, especially as I summoned up the courage to embark on a career in Chinese history without yet having begun any language study, have been more significant than he probably realizes. I am grateful that I attended Haverford at a time when the Junior Research Seminar was a central component of the History Major. The early opportunity to produce new scholarship using historical artifacts and documents remains one of the highlights of my undergraduate years and played no small role in nudging me towards making a career out of historical research.

Friends at Columbia and elsewhere have made the years in graduate school intellectually absorbing and enormously enjoyable. Meha Priyadarshini, Sayaka Chatani,

Liza and Collin Lawrence, Daniel Asen, Ryan Martin, Peiting Li, and Timothy Yang have been fellow journeymen from the very beginning. Anatoly Detwyler and Richard So have been fast friends and intellectual co-conspirators for some years now. I thank Manan

Ahmed, Jennifer Altehenger, Bao Xianqing, Joshua Batts, Glenda Chao, Buyun Chen,

Cyrus Chen, Chen Kaijun, Divya Cherian, Christopher Craig, Joanna Dee Das, Rohit De,

Andre Deckrow, Clay Eaton, Manfred Elfstrom, Seema Golestaneh, Gal Gvili, Reto

Hofmann, Colin Jones, Abhishek Kaicker, Alyssa Kim, Yumi Kim, Ulug Kuzuoglu,

Jenny Lah, Brian Lander, Julia del Palacio Langer, Andy Liu, Luo Weiwei, Olivia Nicol,

Greg Patterson, Chelsea Schieder, Beatrice Schraa, Myra Sun, Anand Taneja, Mehmet

Tezgul, Simon Taylor, Rob Tuck, Tal Unreich, Sören Urbansky, Stacey Van Vleet,

Sixiang Wang, Jake Werner, Stephen Wertheim, Zhang Jian, Zhong Yurou, and many

v others in the Columbia community for the warmth of their friendship and support. Liza

Lawrence, Toby Harper, and Kristen Roebuck generously read portions of the final draft and offered trenchant comments. Gaurav Pant, Abhimanyu Yadav, Meha Desai, Matt

Spigelman, Rob Barry, Eveline Chao, Garrett McVaugh, and Falguni Sen made it easy to escape into other worlds whenever I needed to get away from Morningside Heights.

Numerous colleagues from within the China field and outside have given of their time and energy to comment on parts of the dissertation or discuss points of mutual interest, whether at conferences, workshops or over a drink or two. I would like to thank

Zuoyue Wang, Fa-ti Fan, Tong Lam, Jahanvi Phalkey, Sigrid Schmalzer, Leon Rocha,

David Luesink, Malcolm Thompson, Victor Seow, Jeremy Brown, Philip Thai, Yeh

Wen-hsin, Aminda Smith, Emma Rothschild, Tansen Sen, Margherita Zanasi, Elisabeth

Köll, Benno Weiner, Alex Cook, and Thomas Mullaney. This dissertation would be much the poorer without their insights, probing questions, and invaluable suggestions.

Over the course of several summers and an year-long stint of research in Beijing,

I benefited from the support of a host of scholars and institutions. Professor Yuan Wei at

Renmin University took a personal interest in my research. Without his help, I would have been unable to meet and interview Professors Ni Jiaxun and Lin Fude, or Gong

Jianyao and Wu Hui. I am grateful to these four elderly gentlemen for taking the time to sit down with me and recall the events of more than half a century ago. The dissertation is substantially informed and shaped by what I learned from them. I am thankful to Prof.

Wu Cangping for sharing some of his memories of the 1950s with me. Prof. Liu

Beicheng, with whom I enjoyed numerous long and wide-ranging chats in cafes around

Beijing Normal University, was a constant source of encouragement, always ready with a

vi fresh insight every time I shared my latest archival findings. I thank Professors Yang

Kuisong, Xu Jianqing, and Zhang Jiuchen, who at various times provided me with their help or offered me a venue to share my ongoing work.

I am grateful for the professionalism and friendliness of the entire staff of the

Beijing Municipal Archives; the Foreign Ministry Archives; the Archives of Tsinghua

University; the Resource Room of the Statistics Department at Renmin University; the library of the Institute of Economics at the Chinese Academy of Social Sciences; the

National Library of China; and the Library of Tsinghua University. Margaret Tillman,

Amanda Shuman, and Zach Scarlett ensured a sense of shared enterprise during our daily grind at the BMA. Adrian Sandiford and Hannah Oussedik opened their apartment to me over two summers. Special thanks to James Cheah, Amy Qin, Jamie Chu, Badri

Chillarige, Julia Chuang, James DeSantis, Jee-Eun Choi, Angie Baecker, Li Li, Gao

Lanfang, Anders Hansen, Austin Ramzy, Julian Fisher, and Zhang Jian for making

Beijing such a fun place to live in.

In India, I would like to thank the staff at the Nehru Memorial Museum and

Library; the Indian Council of World Affairs; the Institute of Economic Growth; the

National Archives of India; and the National Library. Dr. Nibedita Ganguly and Ms

Krishna Bhattacharya of the Indian Statistical Institute made my visits to the P.C.

Mahalanobis Memorial Museum and Archives in Kolkata a real pleasure. Their generosity and enthusiasm for archiving Mahalanobis’ activities made possible the substantiation of a fascinating episode in post-colonial Sino-Indian history. Professors

Manoranjan Mohanty, Madhavi Thampi, and Jabin Jacob offered me a venue to share my work and introduced me to the China studies community in New Delhi. Thanks to

vii

Narayan Sen, Debkumar Bose, and Meera Kosambi, who spoke to me about China and

India in the 1950s. Jasbir Singh Malik, Preetika Suri, Samarendra Bhuiyan, Shoili

Kanungo, Abhinav Madan, Warisha Farasat, Kaustabh Chakraborty, and Dhruv Swamini have been stalwart friends; perhaps not always clear on what exactly I was up to, but never letting that come in the way of our friendship.

The writing of this dissertation has been supported by the financial munificence of a number of sources. For their generous support, I am grateful to the Andrew W. Mellon

Foundation, the American Council of Learned Societies, the Social Science Research

Council, the Graduate School of Arts and Sciences (Columbia), the Interdisciplinary

Center for Innovate Theory and Empirics (Columbia), the Weatherhead East Asian

Institute (Columbia), the Department of History (Columbia), the Department of East

Asian Languages and Cultures (Columbia), and the Inter-University Program for Chinese

Language Studies (Berkeley).

The final words are reserved for family, to whom I owe everything. My parents,

Indira and Partha Sarathy, whose support has been unconditional in spite of their misgivings about another generation entering the academy, to my sister, Suparna, who has kept it more real than her brother and parents combined, and to the extended Ghosh and Kaul families—my deepest gratitude.

viii

“…Thou shalt not sit With statisticians nor commit A social science.”

Under Which Lyre A Reactionary Tract for the Times (Phi Beta Kappa Poem, Harvard, 1946) W. H. Auden

CHAPTER I

Introduction

A Crisis in Counting

On 1 October 1949, as Mao Zedong stood on the ramparts of the Gate of Heavenly Peace in Beijing and declared the establishment of the People’s Republic of China (PRC), there was no doubt a tremendous sense of optimism and anticipation. The Chinese Communist

Party (CCP) had won the Civil War against staggering odds and a new era of socialist peace and prosperity beckoned. But challenges were legion too. And for a state that promised to fundamentally re-engineer Chinese society and economy via socialist planning, one of the biggest challenges was that of data. After all, planning requires data, and good planning presumably requires good data. In 1949, such data was hard to come by. Over the preceding four decades, starting with the collapse of the Qing Empire in

1911, China had experienced warlordism, a Japanese invasion, the Second World War, and a debilitating Civil War. What the Communists inherited was a fractured and withered state. The Nationalist Government’s central statistical agency on the eve of

1949, for instance, commanded a mere 5,000 personnel and, despite numerous attempts, had never been able to conduct a nationwide census.1 For many CCP statisticians and economists, the long-term prospects of transforming the PRC into a true socialist utopia hinged, to a large degree, on being able to resolve this crisis of counting.

Today, any mention of China, statistics, and the 1950s is bound to evince in the reader specters of a very different crisis. One of the twentieth century’s worst tragedies, the famine of 1959-61 and the Great Leap Forward (GLF) that in many ways presaged it form a teleological end-point in early PRC history, often constraining our ability to study the 1950s on their own terms. Indeed, for those living in 1959, the year symbolized a

1 Zhu Junyi (Jennings P. Chu) 朱君毅. “The Independently Controlled Statistical System of the Chinese Government.” In Proceedings of the International Statistical Conferences, Vol. II, 92-97. Washngton, D.C.: U.N. World Statistical Confernce, 1947, p. 96.

very different milestone: the People’s Republic was turning ten. To celebrate the occasion, China’s State Statistics Bureau (SSB) produced the volume “Ten Great Years,” and declared in its preface that “an epic of world-shaking importance, forever worthy of being recalled,” had undeniably begun to unfold. Page after successive page full of numbers, tables, and charts followed—a veritable barrage of statistical data corralled to provide indisputable proof that the Chinese had experienced “ten years of leaping progress in economy and culture.”2 And so, we are presented with two starkly different images—the crisis of 1949 and the celebration of 1959. How do we reconcile them?

What kind of statistical infrastructure did the Chinese build during the 1950s? How did they count and know what China was? What impact did this counting have on state- building and state-society relations? This dissertation addresses such questions.

That statistics are a core concern for all modernizing states goes without saying.

In general, one can think of two idealized models that a national statistical system might gravitate toward: centralized or non-centralized. In a non-centralized system, a variety of agencies—central and local government organs, trade bodies, private institutions, research organizations, NGOs, etc.—all periodically collect and publish quantitative data on social and economic activity. The overall quality and comprehensiveness of data rests on the number and diversity of agencies collecting data. When their density is high, the data they produce can represent a national whole. An example of this system is the

United States. On the other extreme is a centralized system in which a nationwide agency is responsible for standardization (of methods, concepts, and schedules), supervision and coordination (of public and private enterprises), and which has centralized control over

2 State Statistics Bureau. Weida de shinian 伟大的十年 (Ten Great Years). Beijing: People’s Press, 1959. p. 1 (Chinese language edition)

the utilization and release of all national data. Centralized statistics are especially important for socialist states that rely on centrally planned economic growth. The former

USSR is an obvious example. The case of PRC after 1949 (and at least for its first decade) is no different.3

In its simplest form, the crisis of counting in the PRC was understood as a problem of building such a centralized statistical system. In December 1950, Zhang

Youyu 张友渔, Vice Mayor of Beijing, the first metropolitan area where the CCP formed a government, offered the following analysis:4

Were there statistics in the past? No matter in liberated areas or in areas under the old regime, we cannot say there were no statistics; just that they were full of inadequacies. It is not that they did not value statistics; for example, in the liberated areas, county and district committee reports did carry reports and tables, but these materials in all likelihood were incomplete, inaccurate, and unsystematic, and therefore couldn’t serve as the basis [for anything]. As for areas under the old regime, their numbers are even more unreliable since they are a product of formalism [形式主义].5

This was indeed a familiar criticism, in line with the basic imperatives of state-building, wherein expansion of state capacity is a central task for any government seeking to establish order after decades of strife and civil war. A decade later, the economist Li

Choh-Ming’s dismissal of the statistical infrastructure inherited by the CCP would largely echo Zhang’s assessment:

3 Li, Choh-ming. The Statistical System of Communist China. Berkeley: Univ. of California Press, 1962.

4 Zhang Youyu 张友渔 (1898-1992) as first Vice Mayor was in charge of the city’s day-to-day operations. Appointed in April 1949 or 1950 (verify), he was one of two vice mayors till 1957 (the other was Wu Han, head of the Beijing branch of the Democratic League). Peng Zhen was only appointed Mayor in February 1951, and Zhang claimed that for much of the 1950s it was he who ran the city, not Peng Zhen. For more see: Mary G. Mazur. “The United Front Redefined for the Party-State: A Case Study of Transition and Legitimation.” in New Perspectives in State Socialism in China, edited by Timothy Cheek and Tony Saich, 66-68. New York: M.E. Sharpe, 1997.

5 “BMA 002-020-000969: 2.

“Since there was hardly any statistical system to speak of before 1949, did Peking manage to set one up that was actually workable? When did this happen and how did it develop? Where were official statistics produced and finalized? Were they used for planning purposes at different government levels? ...What were the size and quality of the statistical work force?”6

Within months of Zhang’s analysis, however, a second, much more fundamental criticism of pre-1949 statistics was articulated by Li Fuchun 李富春, the deputy head of the Central Economic and Financial Commission. This second critique did not waste time lamenting the lack of statistical data or institutions. After all, statistical infrastructure and activities could be established where none or little existed. Instead, this critique called for the wholesale repudiation of existing statistical thought and practice:

In the past, China was a semi-colonial, semi-feudal country; strictly speaking, it did not possess any statistics [worth speaking of]. Statistics in Old China was learned from the Anglo-American bourgeoisie. This kind of statistics cannot serve as our weapon, it is unsuitable for [the tasks of] managing and supervising the country…we need to build [a new] statistics for a new China…7

According to such a critique, the problem with Anglo-American bourgeois statistics was that it served capitalists, whose sole purpose was profit via the exploitation of labor. This argument would be developed and deployed through the rest of the decade by a whole range of interlocutors. To cite one representative instance, a mid-1950s essay made the case in the following way:

Bourgeois statistics exists in order to strengthen the exploitation of workers, in order to serve the interests of capitalists; it uses unscientific formalist mathematical doctrine to conceal the economic dangers of capitalism, whitewash class-conflict, and deceive people. The viewpoints

6 Choh-Ming Li. “Statistics and Planning at the Hsien Level in Communist China.” The China Quarterly, no. 9 (Jan. – Mar. 1962): 112-123. See p. 112 for quote.

7 Li Fuchun 李富春 (July 1951). Speech at the 1st National Statistics Meeting. Li was the deputy head of the Central Economic and Financial Commission, and prior to that had served in the Northeast.

and methods of such statistical theory cannot meet the needs of national construction work and will directly endanger its progress.8

A year after Li’s dismissal of pre-1949 statistics, Vice Premier Zhu De 朱德 noted that the establishment of a new comprehensive statistical system had already become an important task and anyone who lacked sufficient awareness of its significance was in error.9

How this call to arms—to set up a new statistics for a new China—was answered is the principal subject of this dissertation. Central to the activities of early PRC statisticians was a key question: what is the best way to ascertain social fact when you need to count masses of things? The solution that the Chinese chose presumed the impossibility of a universal statistical science and accepted instead a binary that contrasted socialist with non-socialist (capitalist; bourgeois) statistics. Essential to this distinction was the place of probabilistic theory in the study of the natural and the social world. This dissertation uncovers the reasoning behind that choice and traces its implications.

Statistics and tongji 统计—An Etymological Excursus

The Chinese word for statistics is a compound of two characters: tong 统, which means

“all” or “together,” and ji 计, which means to “count” or “calculate.”10 Together, tongji

8 Xu Qian 徐前 and Liu Xin 刘新 . “关于资产阶级统计理论的批判.” Tongji gongzuo tongxun, (1955.08): 28.

9 Xu Qian 徐前 and Liu Xin 刘新 . “关于资产阶级统计理论的批判.” Tongji gongzuo tongxun, (1955.08): 28.

10 Commercial Press. Xinhua zidian—hanyying shuangjie 新华字典—汉英双解 (Xinhua Dictionary with English Translation). Beijing: Commercial Press, 2000, p. 287, 652.

统计 is thus defined as the “collection, sorting, calculation, and analysis of numerical data associated with a given phenomenon,” or more simply as “summary calculation” 总

括地计算.11 This is, of course, a relatively recent definition. The Hanyu dacidian

(Unabridged Dictionary of Sinitic) lists instances of its use in this context as early as the

Ming dynasty (CE 1368-1644). Explicit links between tongji and statecraft, however, do not appear to have been made at that time. Instead, it is the word kuaiji 会计 (accounting) that was the standard word for talking about numbers and statecraft through much of

Chinese history. As several scholars have contended, in ancient and late imperial China the relationship between numbers and statecraft was captured in the statement, “[if] accounting is proper [correct], [then] it is already enough [and governance will be effective].”12 The phrase can be found in The Mencius, purportedly uttered by Confucius when he was a minor official in charge of warehousing 委吏. At least one Chinese statistician has made the case that the phrase’s invocation through succeeding dynasties ought to be considered evidence of the importance accorded to proper statistical work in

Chinese history.13

The rise of statistics as a modern discipline in China, however, is frequently traced to the work of Roger Hart at the Chinese Maritime Customs Bureau during the late nineteenth century.14 Hart’s understanding of statistics was informed by its evolution

11 “指对某一现象有关的数据的搜集、整理、计算和分析等.” See Hanyu dacidian Press. Hanyu dacidian—di jiu juan 汉语大词典—第九卷. Beiing: Hanyu dacidian Press, p. 846

12 “孔子嘗為委吏矣，曰『會計當而已矣』.” See: Mencius, Wan Zhang—II 孟子—万章下. Accessed at the Chinese Text Project: http://ctext.org/mengzi/wan-zhang-ii

13 See: Yu Yue 于越. “会计当而已矣.” Tongji yanjiu 统计研究, (1986.01): 74.

14 Andrea Br ard. “Robert Hart and China’s Statistical Revolution.” Modern Asian Studies, 40, no. 3 (2006): 605–629. In Chapter II, I provide an overview of pre-1949 statistics in China.

within a European context. In the English language, the word statistics is usually traced to

French (statistique) and German (Statistik) antecedents dating from the mid-eighteenth century. In both the French and German cases, the word originally referred to “the study of the state, of statecraft, or of the conditions, circumstances, and politics of a state, the study of numerical data concerning society.”15 Prior to the adoption of statistics, the term that incorporated a similar meaning in English was “political arithmetic,” which is credited to the seventeenth century English economist William Petty.16 An understanding that also incorporated “techniques of mathematical interpretation applied to phenomena for which an exhaustive study of all data is impractical,” i.e. probability theory, took nearly another century to make an appearance.17 In China, this modern understanding of statistics was denoted by tongji, which reappeared as a “return graphic loan:” a “Kanji” term derived from Classical Chinese and used in Japanese to translate modern European words that was subsequently reimported into modern Chinese in the nineteenth century.18

Both these definitions of tongji/statistics—the one that links statistics to statecraft and the other that is primarily mathematical in nature—and the different ways in which they were perceived in the early People’s Republic are central to the arguments of this dissertation.

15 See, for instance, entries for statistic and statistics in the Oxford English Dictionary. Oxford: Oxford University Press. (Online edition)

16 For more on Petty and political arithmetic, see: Ted McCormick 2009. William Petty: And the Ambitions of Political Arithmetic. Oxford: Oxford University Press

17 Oxford English Dictionary (Online Edition). The latter mathematical definition was first recorded in English in 1843.

18 Lydia H. Liu. Translingual Practice—Literature, National Culture, and Translated Modernity, 1900- 1937. Stanford: Stanford University Press, 1995, 302 (fn), 338.

An “Impossible Dream”

In 1974, Leo Orleans, then the China Research Specialist at the Library of Congress, wrote a short article pithily entitled “Chinese Statistics: The Impossible Dream.”19

Whether he was inspired by 1972’s Man of La Mancha and Peter O’Toole’s haunted rendition of the famous song cannot be known.20 But his description of Chinese statistics as “notorious” and the Chinese as a “nation that does not value accuracy” suggest he may very well have. As he noted in the article:

The contention here is that this is not the case, but that in fact, the Chinese themselves lack much of the statistical data we so desperately seek and that their traditional nonchalance toward accuracy in statistics persists among many of the people presently responsible for record-keeping.21

For long, the study of modern China consistently presented a paradox to the China specialist, to whom data was rarely made available.22 As the economist Thomas Rawski noted in a recent review of scholarship on China’s economy: “The common method underlying these studies, particularly those focused on the People’s Republic, was the use

19 Leo A. Orleans. “Chinese Statistics: The Impossible Dream” The American Statistician, May, 28, no. 2 (May 1974): 47-52.

20 The movie was based on the eponymous 1965 Broadway musical, which in turn was adapted from the non-musical 1959 teleplay I, Don Quixote. The original inspiration for all of these works was, of course, Miguel de Cervantes’ seventeenth century The Ingenious Gentleman Don Quixote of La Mancha (El ingenioso hidalgo don Quijote de la Mancha). The musical and movie are both famous for the song “Impossible Dream,” in which Don Quixote expounds on his quest and knightly duties.

21 Orleans 1974: 47. And consider the opening lines of the song: To dream ... the impossible dream ... To fight ... the unbeatable foe ... To bear ... with unbearable sorrow ... To run ... where the brave dare not go ... To right ... the unrightable wrong ... To love ... pure and chaste from afar ... To try ... when your arms are too weary ... To reach ... the unreachable star ...

22 Throughout the 1950s and the ensuing decades, statistics in the PRC remained a state secret.

of fragmentary data to reconstruct the broader economic landscape. This approach arose from the paucity of systematic data prior to 1949 and from the limited publication of economic statistics under the People’s Republic until the appearance of the China

Statistics Yearbook [《中国统计年鉴》] in 1981.”23 As a result, each occasion when the

Chinese released data (e.g. census numbers in 1954, Ten Great Years in 1959, production figures in 1970) generated tremendous excitement.24

And so, one gets the impossible dream of the China scholar tilting at the proverbial windmill of Chinese data. To abuse another metaphor, for many decades, statistics from the People’s Republic of China were treated like gold dust: near impossible to obtain, but with the additional (and often justifiable) fear that they were the product of alchemy—any moment one may discover what one held in one’s hands was dust alright, but hardly gold. Orleans explained the frustrations of the China watchers by noting: “Never having known the advantages (and, yes, the frustrations) of a good statistical base, they [the Chinese] don’t miss it nearly as much as do many of us who look at China from the outside.”25

The mainstream response in the West to the problem of Chinese data has been characterized by a tendency to not take Chinese statistics from the Mao years too

23 Thomas G. Rawski. “Studies of China’s Economy.” In A Scholarly Review of Chinese Studies in North America, 175. Asia Past & Present: New Research from AAS, 2013.

24 For a sense of the surprise and skepticism generated by the 1953 Census numbers, see John S. Aird. “China’s Population, Census and Vital Statistics.” Journal of the American Statistical Association 56 (1961): 443; John S. Aird. “Estimating China’s Population.” The Annals of the American Academy of Political and Social Science 369 (1967): 67-72; John S. Aird. “Sources of Demographic Data on Mainland China.” American Sociological Review 24, no. 5 (1959): 623-630; Ansley J. Coale. Rapid Population Change in China 1952-1982. Washington, D.C.: National Academy Press, 1984; G.B. Cressey. “The 1953 Census of China.” The Far Eastern Quarterly 14, no. 3 (1955): 387-388. On the 1970 production figures, see: Orleans 1974: 51

25 Orleans 1974: 47

seriously. The two extant book-length studies on PRC statistics exhibit this tendency, but in contrasting fashions. The first study was produced by the economist Li Choh-Ming in the early 1960s and the second by the anthropologist Liu Xin much more recently. Li was primarily invested in gauging the accuracy of statistical data produced in the early PRC and in explaining the statistical failure associated with the famine of 1959-1961.26 Liu’s more recent anthropological study has focused on the transition from Maoist subjective mass science to objective science after 1978.27 If the latter paints the entire Mao period from 1949 to 1978 with the broad brush of subjective mass science, then the former tends to neglect the scientific basis for the design of statistical work during the 1950s, blaming its collapse largely on the Great Leap Forward of 1958. In this dissertation, I complicate both these analyses. I demonstrate that the picture within the PRC during its first decade was one of data richness and a great degree of professional engagement and deliberation.

While public access to data may have been limited, the 1950s witnessed a veritable explosion in data collection, management, and analysis.

Statistics and Big Data Positivism

To know something through numbers remains one of the most powerful ways of knowing in the modern world. Powerful not because it is necessarily or always nearer the truth

(were we to grant the singularity of such a thing), but powerful because it offers a tool of persuasion and a basis for rational, methodical, calibrated, and repeatable action that

26 Li 1962a.

27 Xin Liu. The Mirage of China: Anti-Humanism, Narcissism, and Corporeality of the Contemporary World. New York: Berghahn Books, 2009.

remains unmatched.28 Such power has become all-the-more desirable as we have come to realize that common-sense understandings of the world are often erroneous. 29 It is for these reasons that quantification has gained such traction over the past few centuries. Our current all-pervasive zeal for Big Data is symptomatic of this general impetus to quantify.

But such enthusiasm for quantification is not especially new. By my count, since the nineteenth century we have experienced at least three major waves of quantitative positivism. The first of these was during the late nineteenth and early twentieth century, when the use of numbers to produce actionable knowledge in society received a major boost through the activities of Francis Galton and Karl Pearson.30 The ethos—confidence in quantitative analysis—that drove Galton, Pearson, and their contemporaries can arguably be traced to Auguste Comte and his mid-nineteenth century work on positivism.31 This first major wave of quantitative positivism gave rise to disciplines such as statistics, demography, and sociology.

The second major wave of quantitative positivism was in the 1950s, the period of time that this dissertation focuses on. In chapters III and VIII I provide a more systematic treatment of the promise of post-war statistics. For now, I would like to stress that the belief that any problem could be remedied and diagnosed as long as enough data was collected pervaded not just the world of science, social science, and governance, but was rather a part of the very zeitgeist itself. This faith, for that is what it was, is perhaps most

28 Ted Porter. Trust in Numbers: The Pursuit of Objectivity in Science and Public Life. Princton: Princton University Press, 1996.

29 On this, see for instance: Duncan Watts. Everything Is Obvious*: *Once You Know the Answer. Crown Business, 2011.

30 Daniel J. Kevles, In the Name of Eugenics: Genetics and the Uses of Human Heredity. Cambridge, MA: Harvard University Press, 1985.

31 Auguste Comte. A general view of positivism [Discours sur l'Esprit positif 1844] London, 1856

vividly captured in Isaac Asimov’s 1940s and 1950s Foundation Series of science-fiction novels.32 In the series, the “psychohistorian” Hari Seldon, calculates that the Galactic

Empire is in terminal decline and the galaxy will slowly unravel into a long period of chaos. Seldon performs this analysis by using the science of psychohistory, a field that he developed from “a set of vague axioms” to “a profound statistical science.” A neologism coined by Asimov, psychohistory combined history, psychology, sociology, and mathematical statistics to make general predictions about the future of large masses of people: “The individual human being is unpredictable, but the reactions of human mobs,

Seldon found, could be treated statistically. The larger the mob, the greater the accuracy that could be achieved.”33 Enthusiasm for quantitative positivism clearly was not just limited to the domain of stuffy debates about policy and economics.

And yet, even if the 1950s were a moment of pervasive big data positivism, such positivism manifested itself in more than one way. As this dissertation will demonstrate, the dreams of Big Data in China were theoretically and substantially different from those in the West or in places like India. Chinese statisticians in the early PRC categorically rejected the role of probabilistic thinking in the social world and favored counting things exhaustively.

32 Isaac Asimov Asimov, Isaac. The Foundation Trilogy. London: The Folio Society, 2012. The nine stories that comprise the Foundation trilogy were originally published serially in the science fiction magazine Astounding over an eight year period in the 1940s. They were recompiled into three volumes in the early 1950s.

33 Asimov was clearly applying the law of large numbers to prospective large scale human action. Elsewhere in the series, he wrote: “The laws of Psychohistory are statistical in nature and are rendered invalid if the actions of individual men are not random in nature. If a sizable group of human beings learned of key details of the Plan, their actions would be governed by that knowledge and would no longer be random in the meaning of the axioms of Psychohistory. In other words, they would no longer be perfectly predictable.”

In a somewhat ironic twist, our current wave of Big Data positivism, the third in my reckoning, has witnessed the return of the dream of total information, though unsurprisingly it is also accompanied by legitimate fears about the growth of the all- powerful and all-seeing state and corporation.

Engaging Fields

As the first historical study of the development of statistical science in modern China, this dissertation fills a critical gap in our general understanding of contemporary Chinese history. I draw inspiration from work on the history of science and social science in

China and elsewhere, the growing scholarship on the history of post-1949 China, as well as recent trends in global history. Scholarship on the history of science in modern China has persuasively demonstrated modern China’s active participation in globally evolving social and political technologies. These studies have, however, tended to focus either on the Republican years (CE 1911-1949) or on the post-reform era (CE 1978- ), leaving the first few decades of the PRC relatively under-explored.34 Building on these studies, my research situates itself in the growing subfield of scholarship on 1950s China. Among the earliest arguments against understanding 1949 as a moment of rupture was the work of

William Kirby.35 In more recent years, no longer hostage to Cold War geo-politics and

34 See, for instance: Yung-chen Chiang. Social Engineering and the Social Sciences in China, 1919-1949. Cambridge: Cambridge University Press, 2001; Susan Greenhalgh. Just One Child—Science and Policy in Deng's China. Berkeley: Univ. of California Press, 2008; Ruth Rogaski. Hygienic Modernity: meanings of health and disease in treaty-port China. Berkeley: University of California Press, 2004; Paul B. Trescott. Jingji Xue: The History of the Introduction of Western Economic Ideas into China, 1850-1950. Hong Kong: Chinese University Press, 2007.; and Tong Lam. A Passion for Facts: Social Surveys and the Construction of the Chinese Nation-State, 1900–1949. Berkeley: University of California Press, 2011.

35 William C. Kirby. “Continuity and Change in Modern China: Chinese Economic Planning on the Mainland and on Taiwan, 1943-1958.” Australian Journal of Chinese Affairs 24 (July 1990): 121-141.

disciplinary or temporal boundaries, and spurred by the increasing openness of local archives, historians have offered new perspectives on the early PRC.36 While some have further facilitated a reassessment of 1949 as a rupture,37 others have investigated aspects of the transition to Communist rule, exploring subjects such as marriage, gender relations, skill and rural industries, urban transformations, film, urban outcasts, and much else.38

History of science, in particular socialist mass science, has also begun to receive much well-deserved attention.39

This dissertation is also informed by the rich history of science literature on statistics and quantification in the pre-twentieth century world.40 Equally significant are

36 Julia Strauss ed. “The History of the People's Republic of China, 1949-1976.” The China Quarterly Special Issue. Cambridge: Cambridge University Press, 2007. See, in particular, Strauss’ introduction to the volume.

37 Xiabo Lü and Elizabeth Perry, eds. Danwei: the Changing Chinese Workplace in Historical and Comparative Perspective. New York: M.E. Sharpe, 1997; and Strauss 2007.

38 See: Neil J. Diamant. Revolutionizing the Family—Politics, Love, and Divorce in Urban and Rural China, 1949–1968. Berkeley: University of California Press, 2000; Jacob Eyferth. Eating Rice from Bamboo Roots—The Social History of a Community of Handicraft Papermakers in Rural Sichuan, 1920– 2000. Cambridge: Harvard University Press, 2008; James Z. Gao. The Communist Takeover of Hangzhou: The Transformation of City and Cadre, 1949–1954. Honolulu: University of Hawai‘i Press, 2004; Gail Hershatter. The Gender of Memory—Rural Women and China’s Collective Past. Berkeley: University of California Press, 2011; Paul G. Pickowics and Jeremy Brown, eds. Dilemmas of Victory--The Early Years of the People's Republic of China. Cambridge: Harvard University Press, 2007; Aminda Smith. Thought Reform and China's Dangerous Classes Reeducation, Resistance, and the People. Rowman and Littlefield, 2013; and Strauss 2007.

39 See: Focus: Science in Modern China. Isis 8, no. 5. Chicago: University of Chicago Press, 2007; Thomas Mullaney, Thomas. Coming to Terms with the Nation: Ethnic Classification in Modern China. Berkeley: University of California Press, 2011; and Sigrid Schmalzer. The People’s Peking Man: Popular Science and Human Identity in Twentieth-Century China. Chicago: University of Chicago Press, 2008.

40 See, for instance: Lorraine Daston. Classical Probability in the Enlightenment. Princeton: Princeton University Press, 1988; Alain Desrosières. The Politics of Large Numbers—History of Statistical Reasoning. Translated by Camille Naish. Cambridge: Harvard University Press, 1998; Ian Hacking. “Biopower and the Avalanche of Printed Numbers.” Humanities in Society 5 (1982): 279–95; Ian Hacking. The Taming of Chance. Cambridge: Cambridge University Press, 1990; Ian Hacking . “Making up People.” (Essay based on “Tenth Annual British Academy Lecture” of April 2006; the full text of this talk will appear in the Proceedingsof the B.A.) London Review of Books, (17 August 2006): 7-11; Lorenz Krüger, Lorraine J. Daston, and Michael Heidelberger, eds. The Probabilistic Revolution - Vol. 1: Ideas in History. Cambridge: MIT Press, 1987; Silvana Patriarca.Numbers and Nationhood: Writings Statistics in Nineteenth-Century Italy. Cambridge: Cambridge University Press, 1996; Mary Poovey. A History of the

the works of sociologists, political scientists, anthropologists, and social theorists who have investigated various aspects of knowledge creation, scientific research, state-power, and the role of numbers in such processes.41 Finally, my focus on the mobilization of new technologies in the service of new governance agendas in a global setting goes hand-in- hand with the wider turn away from Cold War paradigms. The Sino-Indian statistical exchanges of the 1950s remind us that experimentation and innovation took place in many contexts after 1945. Such transnational flows of ideas, information, practices, and people draw attention to reciprocal linkages across places during a century in which cooperation between governments, organizations, and individual researchers was the norm, and not the exception.42

Organization

The dissertation is divided into three substantive parts and seven chapters. The three chapters in Part I, “A Crisis in Counting,” explore what was new about statistics in the

Modern Fact: Problems of Knowledge in the Sciences of Wealth and Society. Chicago: University of Chicago Press, 1998; Theodore M. Porter. The Rise of Statistical Thinking, 1820-1900. Princeton, Princeton University Press, 1986; and Theodore M. Porter 1994. “Chance Subdued by Science.” Poetics Today 15:3 (Fall 1994): 467-478.

41 See, for instance: William Alonso and Paul Starr, eds. The Politics of Numbers. New York: Russell Sage Foundation, 1987; Bernard Cohn. An Anthropologist among Historians. Oxford: Oxford University Press, 1990; Michel Foucault. “Governmentality.” In The Foucault Effect: Studies in Governmentality, edited by G Burchell, C Gordon and P Miller. London: Harvester Wheatsheaf, 1981; Greenhalgh 2008; Bruno Latour. Science in Action, How to Follow Scientists and Engineers through Society. Cambridge: Harvard University Press, 1987; Donald A. MacKenzie. An Engine, Not a Camera: How Financial Models Shape Markets. Cambridge: MIT Press, 2006; Timothy Mitchell. Rule of Experts: Egypt, Techno-Politics, Modernity. Berkeley: Univ. of California Press, 2002; James C. Scott. Seeing Like a State: How Certain Schemes to Improve the Human Condition Have Failed. New Haven: Yale University Press, 1998; and James C. Scott. The Art of Not Being Governed—An Anarchist History of Upland Southeast Asia. New Haven: Yale University Press, 2009.

42 See, for instance: Matthew Connelly. Fatal Misconception: The Struggle to Control World Population. Cambridge: Belknap (Harvard University Press), 2008; and Akira Iriye. “Internationalizing International History.” In Rethinking American History in the Global Age, edited by Thomas Bender, 48-60. Ewing, NJ: University of California Press, 2002.

People’s Republic after 1949. Chapter II, “Revolutionary Statistics in the Early 1950s,” explores early statistical work in the PRC’s Northeast, arguing that this work and the practical experience so gained was the foundation upon which the rest of the country’s statistical apparatus was based. The establishment of socialist statistical work thus preceded its theoretical and ideological justification, which then becomes the subject of

Chapter III: “Lies, Damned Lies, and (Bourgeois) Statistics: Ascertaining Social Fact.”

The Chapter also provides an assessment of Soviet technical aid and of the Soviet statistical experts who were instrumental to the organization of statistical activity in the

PRC. Chapter IV, “No “Mean” Solution: The Reformulation of Statistical Science,” explores how this new understanding of statistics became dominant in the 1950s and how it affected the valuation of key concepts and methods.

Taken together, the three chapters in Part I return us to an important issue in early

PRC history: that of understanding the extent to which the shift to Communist rule after

1949 was a rupture, and tracing the continuities that persisted nevertheless. Statistics exhibits elements of both rupture and continuity, depending on our time-frame as well as geographic focus. A closer look at the statistical activities in the Northeast produces a strong case for substantial rupture—new statistical methods and practices were introduced under the direction of Soviet experts and quickly became extensive. The picture for the rest of the China till 1952 is much less clear. Thus, in these early years the rupture appears regionally determined.

In contrasting fashion, statistical education continued unchanged during these first two to three years. Many academic statisticians maintained their positions after 1949 and were involved in the training of statistical workers. It was only in 1951, and more

properly after 1952, that a concerted effort was mounted to redefine statistics, both by changing the curriculum, and through targeted criticisms of well-established statisticians and their textbooks. Changes in practice thus preceded changes in education and academic discourse. Such findings help us rethink 1949 not as a singular moment of rupture but instead call for the separation of the rhetoric of rupture (promoted by the CCP) from the actual changes on the ground, which were temporally and regionally variegated.

The three chapters in Part II, “Statistical State, Quantified Society,” focus on statistics in practice, thereby tracing some of the implications of the theoretical, scientific, and administrative decisions investigated in Part I. Chapter V, “The Nature of Statistical

Work,” draws upon materials from the Beijing Municipal Archives, including various statistical reports generated across all levels of the statistical system, internal bulletins of work, and materials from conferences of statistical workers (at local, provincial, and national levels), to explore the messiness (contra the idealization unpacked in Chapter II) of actual statistical work. It captures not just the centralizing impetus of the expansion, but also the varieties of challenges that were encountered in putting into practice the methods that were at the heart of socialist statistics: (1) the periodical report system 定期

报表制度, and (2) various forms of typical sampling 典型调查. Even though acknowledgement of the problem of excess issuing of reports 滥发报表 and the chaos they generated had been present since before 1949, the problem became increasingly ominous through the 1950s, fueled in part by a table-as-product rationale, which is explored in Chapter VI: “To “Ardently Love Statistical Work”: State (In) capacity,

Professionalization, and their Discontents.” This chapter analyzes the variety of stratagems—training, supplementary training, self-study, motivation, and reward—that

were employed to professionalize and maintain a cadre of statistical workers that numbered as high as 200,000 by 1956. The final chapter in Part II, “Statistical State,

Quantified Society,” investigates the uses and reception of statistics through two case studies: economic planning and the institution of the work-points system in the rural countryside.

Part III consists of one chapter, “The Turn to India: Sino-Indian Statistical

Exchanges, 1951-1959.” By the 1950s, India had become one of the centers of global innovation in statistical technology. And starting as early as 1951, and with increasing urgency after 1956, Chinese and Indian statisticians traded visits as they sought to learn from each other’s experiences. At the heart of these exchanges was the desire to learn more about mathematical statistics, in particular the then cutting-edge statistical method of random sampling. Involving such key figures as the deputy director of the SSB, Wang

Sihua 王思华 and the Indian statistician P.C. Mahalanobis, the unearthing of these exchanges forces a reappraisal of Cold-War paradigms and sets out in stark relief the extent to which Chinese statisticians and leaders clearly understood the strengths and shortcomings of their own statistical system.

The Epilogue, “A Great Leap in Statistics,” traces very briefly the impact and legacy of an “on-the-spot meeting” 现场会议 in the city of Baoding 保定 in the summer of 1958 that was organized to discuss a Great Leap Forward in statistics. Over the ensuing months, the tussle between socialist statistics and its alternatives was largely overwhelmed by Maoist mass science. In statistics this meant the rebadging and valorization of typical sampling, which was now explained as Mao’s synthesis of

Marxist–Leninist theory with the practice of revolution in China. Mao’s 1927 Report on

the Peasant Movement in Hunan became the foundational text for this ‘revolutionary’ method. It is this notion of mass science, with its anti-expertise and anti- professionalization, that has come to inform our understanding of much of the Mao era.

But as this dissertation demonstrates, the picture till 1958 was quite different, far more contested, and open to very different paths.

PART I

A CRISIS IN COUNTING

CHAPTER II

Revolutionary Statistics in the Early 1950s

In July 1950, Wang Sihua 王思华, recently designated head of the Northeast Statistical

Bureau (NSB), proudly declared to his audience that not only were they participating in the first meeting of statistical workers in the Northeast’s history, but that their meeting was also the largest in the nation’s history.43 This was clear evidence, according to him, that New China was already on the path towards achieving a planned economy; for no planned economy could be possible without the presence of scientific statistical work. In his address, Wang observed that since the promulgation of the “Decision Regarding

Strengthening Statistical Work” on 10 April 1950, statistical work had entered a “new phase.”44 As a result, “statistical work in the northeast had begun to move from a state of decentralization to integration, from chaos to the correct path, from the non-scientific to the scientific.” Wang explained that help offered by Soviet experts was critical to this change. The Soviet Union possessed over thirty years of experience and it was their expertise and aid that allowed Wang to declare: “We have set up a statistical system, we possess method, [and] we possess confidence.”45

A few years later, in December 1952, while addressing the third annual meeting of the Northeast Statistical Bureau, Wang was even more emphatic. Over the previous three years, he would claim, the bureau had successfully established “a new type of

43 Wang Sihua. “Tongji gongzuo the renwu 统计工作的任务 (The Tasks of Statistical Work).” In Tongji gongzuo—di yi xuanji 统计工作—第一选集 (Statistical Work—The First Anthology), 7-14. Shenyang: Northeast Statistical Press, February 1952. A slightly edited version of this address was included in a collection of Wang’s works published in 1986. See: Wang Sihua. Wang Sihua tongji lunwenji 王思华统计论文集 (Collected Statistical Works of Wang Sihua). Beijing: China Statistics Press, 1986, 1-14.

44 Wang Sihua 1952: 7. For a text of the decision, see “Guanyu jiaqiang tongji gongzuo de jueding 关于加强统计工作的决定 (Decision Regarding Strengthening Statistical Work).” In In Tongji gongzuo—di yi xuanji 统计工作—第一选集 (Statistical Work—The First Anthology), 127. Shenyang: Northeast Statistical Press, 1951.02.

45 Wang Sihua 1952: 8.

standardized statistical work” 新型正规化的统计工作.46 For him, such standardized statistical work was the hallmark of socialist countries and was fundamentally different from the statistics of capitalist countries. It was also different from the older small- scale, handicraft-like practices of conducting the odd statistical survey common in the pre-49 years.47

At the time Wang delivered the latter speech, a national bureau of statistics (The

State Statistics Bureau; 国家统计局) had recently been set up in Beijing and, as will be described in Part II of the dissertation, the rapid expansion of a national network of statistical offices and personnel at provincial, county, and village levels was underway.

As the first statistical agency in the PRC to fully incorporate Soviet/socialist approaches to statistics, the Northeast Statistical Bureau became the model for this rapid expansion.48

46 Wang Sihua. “San nian lai dongbei tongjiju gongzuo zongjie 三年来东北统计局工作总结 (Summary of Statistical Work Performed by the Northeast Statistical Bureau over the Past Three Years).” Tongji gongzuo—di wu xuanji 统计工作—第五选集 (Statistical Work—Fifth Anthology), Northeast Finance and Economics Press, January 1953, 1-10. An edited version was published in Wang’s collected works; see: Wang Sihua 1986: 31-49.

47 Wang Sihua 1953: 3.

48 Oral history interview with Wu Cangping 乌沧萍 (1922- ), Beijing, 11 March 2011. Wu, a demographer and statistician was trained in the West before being invited back by the Communist government in the early 1950s. Offered a position at People’s University, he was active in demographic research and was a member of the four-man team that visited India in late 1956. Since the 1980s he has been among the pioneers in the field of gerontology in China.

For the Northeast’s distinctiveness and prototypical qualities in the early PRC more generally, see The Cambridge History of China, 14, part 1 (The People’s Republic), 82-83: “The outstanding case of regional particularity was the Northeast…it reflected the fact that the Northeast was the most advanced region and served as a bellwether for the rest of the country for a number of reasons. First, having benefited from industrialization under Japanese rule, the Northeast had the most developed economic base. It provided 34% of China’s industrial output in 1949 and 52% in 1952. Second, by virtue of being the first region totally liberated, the Northeast could move more quickly toward comprehensive policies and was able to begin regional planning by 1950. And finally, proximity to the Soviet Union and Soviet holdings in the regional railroads and the port of Luta (Port Arthur-Dairen) combined to provide easy access to Soviet aid and influence. Thus the northeast instituted Soviet methods of economic management, albeit with difficulty due to shortages of skilled personnel, and these methods were generally endorsed by the central leadership in Peking for extension to China as a whole.”

Under Wang’s leadership, the Bureau had not only carried out statistical work since early

1950, but also undertook the first major attempt at translating and publishing Soviet treatises on statistical theory and methods.49

The NSB’s activities had a seminal impact not just on the ways in which statistical work was organized and carried out, but also on the very definition of statistics, itself. By valorizing the Soviet Union’s ‘advanced experience,’ debates within China starting as early as 1950 came to reformulate statistics as a social science in contradistinction to a natural or universal science. This science’s function lay in helping build a socialist society. As a result, the new statistics stood bifurcated—rhetorically and substantively—from what was labeled the tainted, bourgeois, and socially unproductive pursuit of mathematical statistics. Statistical work, first most notably in the northeast under Wang Sihua, and then progressively elsewhere, followed suit. By 1952, these changes began to be implemented within statistical education as part of a general reorganization of the higher education system.

And yet, in studies of PRC statistics, this early period garners almost no attention.

Li Choh-Ming, one of the first scholars to focus his attention on the question of statistics in the PRC, was dismissive of these initial years. As he wrote in 1961, “from 1949 to

1952 no effort was made to develop a statistical system national in scope.”50 This emphasis on a national system led him to begin his longer book-length study on PRC statistics not with the founding of the PRC in 1949, nor with statistical activities in the

49 For more see: Liu Chang 刘畅 and Zhang Yun 张云. Ershi shiji zhongguo de tongjixue 二十世纪中国的统计学 (Statistical Education in 20th Century China). Beijing: Dangjian duwu chubanshe, 2000, 118.

50 Choh-Ming Li. “Communist China’s Statistical System: 1949-1957.” The American Economic Review, 51, no. 2 (Papers and Proceedings of the Seventy-Third Annual Meeting of the American Economic Association) (May 1961): 499-517 (quote from 500).

Northeast, but with the founding of the State Statistics Bureau in Beijing on August 8

1952. To be fair, Li did recognize the Northeast as the only region capable of producing reliable statistical data prior to 1952, and he also acknowledged the early presence of

Soviet statisticians in the Ministries of Railways (1950) and Health (1951-52).51 But, aside from this somewhat perfunctory acknowledgement, his approach only took seriously statistical activities national in scale, which were only made possible after the formation of the SSB in 1952.52

As this chapter will demonstrate, the early period of 1949-1952 is crucial to understanding statistics and statistical work through the decade that followed. Unpacking the constituent elements of Wang’s ‘new type of standardized statistical work,’ will permit an articulation of the ways in which it was regarded as more standardized, correct, and scientific than what preceded it in the years before 1949. At the same time, the chapter also seeks to complicate the neat binary that Chinese (and Soviet) statisticians liked to deploy—of socialist and bourgeois statistics, of pre-liberation and post-liberation statistics. I will show that in the case of the PRC this binary understanding is incomplete.

The 1950s were shot through with the influence, rarely pedagogical or technical in nature but rather all-pervasive and methodological, of what could be termed the Maoist model of ascertaining fact. This Maoist model, with its emphasis on direct experience over technical knowledge or ideological fervor as a source of expertise, made the earlier binary and the clean solutions it offered much more conflicted. As a result, statisticians and statistical workers through the 1950s found themselves not just transitioning from a

51 Choh-Ming Li 1962a: 7-8.

52 Writing a little over 10 years later, Leo A. Orleans would echo this assessment: “The responsibility of standardizing and centralizing all statistical work in the county was finally vested in the State Statistical Bureau in August of 1952.” Orleans 1974: 50.

‘non-socialist’ to a ‘socialist’ system of data collection and analysis, but also operating in a milieu that particularly valued personal individual practical experience.

Statistics in China before 1949

In September 1947, Washington, D.C. played host to the twenty-fifth meeting of the

International Statistical Institute’s biannual conference.53 China was well represented at the nearly two-week long event. The Director of Statistics of the National Government,

Jennings P. Chu (朱君毅 Zhu Junyi, 1892-1963), led a six-member delegation representing a variety of fields and specializations within and outside the government. Of the six members, two were already stationed in Washington, D.C.: D.K. Lieu (Liu Dajun

劉大釣, 1890-1962) was the Commercial Counselor at the Chinese Embassy as well as the Chinese Representative on the United Nations Statistical Commission, and Alfred

Bangnee Liu (Liu Pengnian 刘彭年, 1901-?) was the Director of the Chinese News

Service. These three men were joined by the well-known sociologist, demographer, and

Tsinghua University professor, Chen Da (Ta Ch’en 陈达, 1892-1975), who had traveled from Beijing; by the Shanghai-based Director of the Accounting Department of the

Central Bank of China, Kuo-Pao King (Jin Guobao 金国宝, 1893-1963); and by the Head of the Department of Mathematics of the National Central University in Nanjing, Pei-

Ching Tang (唐培经 Tang Peijing, 1903-1988).54 Born in the late Qing, some of the

53 The names are listed in: Introduction—International Statistical Conferences, September 6-18, 1947, Washington, D.C.— Proceedings of the International Statistical Conference, Volume I. Calcutta: Alka Press, 1951, 42. (Hereafter PISC) Also see: Liu Chang and Zhang Yun 2000: 70.

54 Among them, three delivered papers: Chu delivered a paper titled “The Independently Controlled Statistical System of the Chinese Government”; King delivered a paper titled “A Census of Nine Hsien in

members had benefited from Boxer Indemnity fellowships and received education in the

United States, while others had been trained nearer home in Japan. Some had conducted and published research using statistical data and methods, while others had written textbooks and been involved in the administration of statistical work across China. They were each exposed to contemporary social and scientific theories, and, to a large extent, their approaches reflected the dynamic and varied characteristics of the world of social science research and statistical activity during the first half of the 20th century in China. It was this world and its intellectual and administrative traditions that Wang Sihua, Li

Fuchun, and Zhu De forcefully rejected in the early 1950s.

The dynamism and variety in the world of Chinese social science research in the first half of the twentieth century was itself the outcome, at least partially, of an intense search for the causes and remedies for China’s relative political, economic, and military weakness on the world stage. As a once-powerful empire had transitioned into a weak republic, elites (intellectuals, administrators, and the like) had to contend with what they saw as the long shadow of enduring imperial practices and traditions. Enfolded in this long shadow were the twin practices of statecraft and data collection.

The Qing (1644-1911) state had routinely gathered information on prices, rainfall, salt, grain holdings, famine, and population.55 In 1793, five years before Malthus

Szechwan, China,” and Chen Da delivered a paper titled “Factors of Urban Growth in China.” For additional details, see: PISC, Volume I: 75-76; 78-79; and 94 respectively, and Volume II: 92-97; 167-185.

55 See: Li 2007. Mention can also be made of the decimal based Lijia 里甲 and Baojia 保甲 systems (with their origins in Song during the life of Wang Anshi 王安石 (C.E. 1021-1086), though Qing methods probably cannot be traced further back than the early Ming (C.E. 1368-1644) population censuses, fish registers, and so on. In general, barring central projects like the Grand Canal and famine relief, public services and local security were handled at the local or regional level. For the kinds of tensions this generated in the fiscal structure during the High Qing, and the attendant reforms that were attempted, see: Madeleine Zelin. The Magistrate’s Tael: Rationalizing Fiscal Reform in Eighteenth Century Ch'ing China. Berkeley: University of California Press, 1984.

published his famous tract, the Qing official Hong Liangji 洪亮吉 (1746-1809) had already raised the specter of unchecked population growth, thereby challenging the deeply rooted theory that population increase reflected good governance.56 Towards the end of the Qing, increasing incapacity forced the state to give up on attempts to quantify and collect information. During this same period, heightened interest in Western knowledge forms set off an enduring debate among Chinese elites on the ‘usefulness’ of

Western knowledge in contrast to the ‘essence’ embodied in Chinese learning.57 As a result, institutions such as the Jiangnan Arsenal devoted considerable resources to the translation of Western tracts, including treatises on statecraft that included some statistical analysis.58 The advent of ‘modern’ statistics in China can, arguably, be traced to the 1859 publication of the first Maritime Customs Report.59 By the turn of the century,

56 “China’s Population Problem,” in: Sources of Chinese Tradition, Vol. 2, edited by. William Theodore de Bary and Richard Lufrano, 174-76. New York: Columbia University Press, 2000. In similar fashion, William Rowe has demonstrated that the nineteenth century statecraft thinker and agrarian reformer, Bao Shichen (1775-1855), made national estimates of crop yield and population and dabbled in statistical ideas about state output, future economic and fiscal policies, and conceptions of state income. William T. Rowe. “ Bao Shichen (1775-1855), an Early Nineteenth-Century Agrarian Reformer.” Unpublished paper delivered at the Modern China Seminar, Columbia University, New York, December 15 2012. Available here: http://krieger.jhu.edu/east-asian/documents/Rowe_Seminar_Paper.pdf

57 Formulated in Chinese as ti-yong (体-用; Western learning for its usefulness, Chinese learning for its essence), the seemingly instrumental use of Western knowledge to strengthen an increasingly weak China polarized Chinese intellectuals. For the translator and editor Yan Fu (严复, 1854-1921), without a broader engagement with Western civilization and its political and scientific forms, such an attempt was bound to fail. Others, such as the scholar Feng Guifen (冯桂芬; 1809-1874), held that all China need in order to compete on level terms were the material products of Western civilization.

58 Benjamin Elman. On Their Own Terms: Science in China, 1550-1900. Cambridge: Harvard University Press, 2005, 331-334. Among the examples that Elman cites was the Chinese translation of Henry Fawcett’s A Manual of Political Economy: Fuoguo ce 夫国策. For a digital copy of the book, see: http://www.sino.uni-heidelberg.de/encyc/Fuguo_ce/index.htm

59 D.K. Lieu 刘大钧 (Liu Dajun). “Statistical Work in China.” Presented at the Nineteenth Session of the International Statistical Institute, Tokyo, Japan, 1930, 3. On statistics at the Maritime Customs Bureau and the role of Robert Hart, see: Andrea Br ard. “Robert Hart and China’s Statistical Revolution.” Modern Asian Studies, 40, no. 3 (2006): 605–629. On the development of statistics and the translation of mathematical and statistical works in the nineteenth century, see: Andrea Br ard and Wang Youjun 王幼军. “Tongjixue zai jindai zhongguo de yanbian—cong shehui tongji dao shuli tongji 统计学在近代中国的演

the Qing civil service exam system had begun to incorporate questions drawn from the modern social sciences and mathematics.60 Subsequent reforms announced in 1905 signaled changes that included plans for representative constitutional government and a modern census.61

The fall of the Qing in 1911, while aborting those reforms, helped accelerate the spread of nascent Western social sciences and witnessed attempts at new forms of governance and state-building. A number of Chinese received training as demographers, sociologists, anthropologists, and ethnographers, usually in the West or Japan, and began to work for the Chinese state and collaborate on data collection with Western institutions and scholars. No longer understood purely through an Impact-Response framework, recent research into their activities has stressed the level of hybridity—drawing upon their Western education, but also deeply cognizant of local realities and traditions— that these social scientists achieved in their research.62 Through the ensuing decades of the

1920s and 1930s there emerged what Tong Lam has identified as a ‘Social Survey

Movement,’ where the desire to document, count, tabulate and survey the nation

变—从社会统计到数理统计 (The Evolution of Statistics in Modern China—From Social Statistics to Mathematical Statistics).” Shanghai jiaotong daxue xuebao (zhexue shehui kexue ban)上海交通大学学报 (哲学社会科学版), 18, no. 3 (2010): 61-67, 74.

60 Elman 2005: 333.

61 For more on the 1908 census, see Tong Lam 2012, Chapter III.

62 There has been much recent work on the rise of several modern (social) science disciplines, which highlight the frequently transnational intellectual milieu within which they emerged, the critical importance of European and American influence through training, institutions such as the Rockefeller foundation, and collaboration between international and Chinese scholars, and so on. See for instance: Asen 2012 (Forensics); Buck 1980 (American Science); Chiang 2001 (Sociology), Hu Danian 2005 (Physics), Mullaney 2011 (ethno-linguistic classification), Schmalzer 2008 (Evolutionary Science); Schneider 2003 (Biology and Genetics), and Trescott 2007 (Economics).

permeated not just the elite, trained practitioner of social science but also the lay man.63

For Tong Lam, these surveys operated on two levels. “The idea of social science and social survey research presumed an external social reality with internal mechanisms waiting to be discovered and dissected. On the other hand, Chinese social thinkers themselves also vehemently argued that China did not have a real organic and functioning society and that their goal was to create one.”64 By 1949, social scientists such as the demographer and sociologist who headed the Institute of Census Research in the 1940s, Chen Da 陈达 (1892-1975), the population geographer Hu Huanyong 胡焕庸

(1901-1998), the sociologist Chen Hansheng 陈翰笙 (1897-2004), and the economist Ma

Yinchu 马寅初 (1882-1982), all participated in a well-established, sophisticated, and methodologically and ideologically diverse tradition of social science research and statistical analysis.

In the field of statistics and statistical activity specifically, three broad trends are discernible. The first is the expansion and diversification of educational materials relating to statistics and a concomitant engagement with various approaches to statistical theory and practice. In the years immediately following the establishment of the Republic in

1912, translated texts reflected an awareness and appreciation of the co-existing strands of mathematical statistics (with its focus on probability theory), Anglo-American statistics, as well as the more descriptive statistical practices prevalent in Germany.65

From the 1920s to the 1940s, numerous original works on statistical theory and

63 Lam notes that between 1927 and 1935 nine thousand surveys were carried out in China. Lam 2012: 4

64 Lam 2012: 9-10.

65 In a recent article, Br ard and Wang see the long-term trajectory from the late nineteenth into the first half of the twentieth century to be one of a transition from social statistics to mathematical statistics 从社会统计到数理统计. See: Br ard and Wang 2010.

application were also published.66 Other volumes offered comparisons of the differing statistical traditions and their historical trajectories. For instance, in the 1935 publication

An Outline of Statistics 《统计学纲要》, Liu Hongwan 刘鸿万 observed that:

The view of the School of Social Statistics is that the objects of statistical research are social phenomena, and the method of analysis is large-scale observation [大量观察法]; the view of the School of Mathematical Statistics is that it is a mathematical method that helps arrive at statistical solutions [统计解析].67

In contrast, Wu Zaoxi 吴澡溪, in his Economic Statistics 《经济统计学》, contended that statistics was not an independent social science. Instead, it was “a research method that aided [辅助] the pursuit of other forms of knowledge [学问].”68 Indeed, through much of the first two decades after 1911, statistics was not taught as an independent subject. In 1927, with the establishment of the Central Politics School 中央政治学校, a statistics division was set up within the Economics department. It was another ten years before a dedicated department of statistics was established at National Chongqing

University 国立重庆大学.69 Even so, the general practice in most places remained for

66 Liu and Zhang list at least six texts on statistical theory and another 10 in applied statistics. See Liu and Zhang 2000: 66.

67 Liu Hongwan 刘鸿万. Tongjixue gangyao 统计学纲要 (Outline of Statistics), 1935. Cited in Liu and Zhang 2000: 67.

68 Wu Zaoxi 吴澡溪. Jingji tongjixue 经济统计学 (Economic Statistics), 1943; cited in Liu and Zhang 2000: 68.

69 Liu and Zhang 2000: 89-93. Several schools offering short term courses in statistical methods were around since the mid-1920s. It is also worth noting that the establishment of statistics departments was a late development globally.

statistical course offerings to be made through other departments, such as Economics,

Law, and Business.70

Concurrent with the gradual expansion and diversification of statistical education was a gradual expansion in official statistical work. Within the first few years of the

Republic, offices and bureaus charged with carrying out statistical work were set up at various levels, partly on the foundations of pre-existing local institutional mechanisms. In

1916 a State Council/Cabinet Office was re-established and contained a Bureau of

Statistics within it. The general picture up to 1927 is one of various government bureaus, some dedicated to statistical work, but most involved in a variety of spheres, that collected and disseminated data germane to their briefs.71

As described by D.K. Lieu, the years following 1927 witnessed renewed attempts at centralization and expansion of statistical activity. But Lieu considered much of this

‘mushroom growth’ unsystematic, and in the final analysis, unhealthy, since it meant duplication of work and a significant waste of limited resources. He was among the people who convened a conference in February of 1930 that sought to address these issues. The conference recommended that an independent Central Statistical Bureau be established. Concurrently, it also recommended the establishment of a Central Statistical

Commission charged with coordinating the statistical activities of the different Ministries.

In the time it took to set these institutions up, the conference suggested forming a Central

Statistical Association which could immediately set about coordinating statistical activities already underway. It also suggested carrying out population and agricultural

70 Statistics was introduced as a subject of study within several disciplines as early as 1913, and by 1920s several universities offered courses in it.

71 For a detailed description, see: Lieu 1930: 7-12. Also see Liu and Zhang 2000: 79-85.

censuses in 1931. A final recommendation concerned the creation of a Chinese Statistical

Society, which would provide a private forum for academic discussions of statistics.72

Pursuant to these guidelines, the Directorate General of Budgets, Accounts and Statistics was established on 1April 1931.

D.K. Lieu’s discussion is indicative of two broad features of statistical activity during the latter half of the Republican period. The first is that state statistical activities, though they ranged across such subjects as agriculture, population, industry and labor, communications, finance, and so on, were fragmentary and essentially of a rather limited scope, in terms of the types of data collected, their geographical reach, and their representativeness. As the demographer Leo Orleans noted in a 1974 essay, “the statistical system [of Republican China] never developed beyond the rudimentary stage and almost never reached down to the hsien [county] levels.”73 The point is driven home all the more forcefully by the number of non-state, in many cases non-Chinese, actors whose data collection Lieu notes was an important source for the period. For instance, he noted approvingly the statistical surveys conducted by the likes of Sydney Gamble, John

Lossing Buck, Chen Da, and others. In so doing, Lieu points us to the second broad feature of the period—the lively and fairly diverse private engagement with statistics and statistical work across China through the 1930s and 1940s.74 Indeed, outside of government activity, there had also developed a lively community of scholars who participated in research on social issues, generating copious amounts of statistical data as a result. Such collaborative social research, frequently heavily statistical in nature,

72 Lieu 1930: 12.

73 Orleans 1974: 48.

74 Lieu 1930: 14-27.

involved Western social scientists, Chinese social scientists trained in the West, and locally trained social scientists.75 International organizations such as the Council of

Pacific Relations, and the Rockefeller Foundation played a significant role as well, frequently serving as principal funding or organizing agencies.76

By 1947, Jennings P. Chu 朱君毅 reported to the attendees at the International

Statistical Institute’s biannual conference that the Republic of China possessed a

Statistical Directorate that had proven to “be of immense value in government planning, execution and investigation.”77 He identified five main characteristics of the system, to wit: independence (from political interference), liaisonship (i.e. a combination of centralized and decentralized administration), uniformity (across bureaus and types of data), specialization (of personnel), and finally interdependence (between planning, execution, and investigation). The duties of the Directorate included:

(1) Direction and supervision of statistical personnel and statistical functions (2) Design and issuance of statistical forms and enforcement of uniform procedures (3) Assignment of statistical work across various departments (4) Compiling National Census statistics (5) Compiling a National Statistical Report

75 See, for instance: Daniel Kulp. Country Life in South China: The Sociology of Familism. Volume 1. New York City: Bureau of Publications, Teachers College, Columbia University, 1925; Fei Xiaotong. Peasant Life in China: A Field Study of Country Life in Yangtze Valley. New York : E. P. Dutton, 1939; Fei Xiaotong. Earthbound China: A Case Study of Rural Economy in Yunnan. Chicago: University of Chicago Press, 1945; Sydney Gamble. Ting Hsien: A North China Rural Community. New York: International Secretariat Institute of Pacific Relations, 1954; Sydney Gamble. North China Villages: Social, Political, and Economic Activities Before 1933. Berkeley: University of California Press, 1963; Lin Yueh-hwa. The Golden Wing: A Sociological Study of Chinese Familism. Greenwood Press, 1947; and John Lossing Buck. Land Utilization in China, Atlas: A Study of 16,786 Farms in 168 Localities, and 38,256 Farm Families, in Twenty-Two Provinces in China, 1929-1933. Chicago: University of Chicago Press, 1937.

76 For more on the activities of the Rockefeller Foundation, see: Chiang 2000: Chapter 9.

77 Jennings P. Chu 朱君毅 (Zhu Junyi). “The Independently Controlled Statistical System of the Chinese Government,” PISC, Vol. II, 1947, 93. In 1963 Zhu drafted a book on statistical work during the Republican years. But it would not be published until 1988. See: Zhu Junyi 朱君毅 Minguo shiqi de zhengzhi tongji gongzuo 民国时期的政治统计工作 (Government Statistical Work during the Republican Period). Xiandai tongji zhishi congshu 现代统计知识丛书, 1988.

(6) Investigation of administrative efficiency and compiling statistics of administrative results

The Directorate’s activities were summarized in a variety of publications, such as The

National Statistical Report and Abstract, The Statistical Analysis of National Problems

Series, The Statistical Monthly, and the first edition of the The China Statistical Yearbook, released in 1947.78

All this is suggestive of a large statistical establishment in the postwar years, but the actual numbers of bureaus and personnel that Zhu enumerated paint a different picture. According to Zhu, at the end of June 1947, there were eight Bureaus of Statistics and 660 Offices of Statistics in the principal and subordinate organs of the central government. An additional thirty-two bureaus and 1,196 offices were located in the provinces, municipalities, and counties. Total statistical units in the country numbered

1,905 and the total number of personnel amounted to 5,066, of which 120 were at the

Directorate of Statistics itself.79 In spite of such a seemingly small number of personnel for such a large country, Zhu was optimistic about the future of statistical work in China.

The Japanese had been defeated, and the Communists also seemed to be on the run. A stable future beckoned and Zhu looked forward to the establishment of a complete statistical network, more trained personnel, and the “attainment of comprehensive, accurate, and timely results.”80

78 Jennings P. Chu 1947: 92-95.

79 Jennings P. Chu 1947: 96.

80 Jennings P. Chu 1947: 96. He concluded: “The world is united now more than ever, and international cooperation is aiming at world-wide stabilization and prosperity. All this leads to an extensive demand for accurate and timely statistical information for internal as well as external use, and towards this goal the Chinese Government is determined to keep abreast of the advanced countries of the world.”

Zhu’s optimism, we now know, was premature. By 1948, the tide of the civil war that had raged since 1946 had begun to turn in the CCP’s favor. As such, Leo Orleans perhaps summed-up it up best when he noted that “despite some well-meaning efforts by the Nationalist government, when the Communists took over the mainland in 1949 China continued to be ‘a land where the statistician may perish for want of a few figures, where records are more romantic than mathematical.’”81

Wang Sihua 王思华 and the Northeast Statistics Bureau, 1949-1954

Well before the People’s Republic was officially brought into existence, the CCP had begun to organize statistical work and economic activity in the regions it controlled. After the end of the Second World War, the Party moved into the northeastern provinces vacated by the Japanese. Wang Sihua 王思华 was assigned to these areas to aid in their governance. Working under leaders such as Li Fuchun 李富春 and Gao Gang 高岗,

Wang’s first assignment was to take over the administration of West Liaoning and North

Liaoning. Early in 1947, he was also appointed the Deputy Secretary-General of

Heilongjiang province and the Director of both the Northeast Statistical Office and the

Standing Finance Committee.82

81 Orleans 1974: 48. Orleans himself quoted Bernard Martin. Strange Vigour. Port Washington, NY: Kinnikat Press, 1970, 3.

82 Wang’s biographical details are drawn from the following sources: (1) Wang Sihua 1986: 224-226; (2) “Jinian Wang Sihua tongzhi shishi wuzhou nian 纪念王思华同志逝世五周年 (Commemorating the Fifth Anniversary of the Passing of Comrade Wang Sihua).” Tongji gongzuo 统计工作 (China Statistics), no. 7 (1983): 5-6, 14; (3) Li Deng 李澄. “Ji nian Wang Sihua tongzhi 纪念王思华同志 (Commemorating Comrade Wang Sihua).” Renwen zazhi 人文杂志 (The Journal of Humanities(Bimonthly)), no. 6 (1984): 64-65, 68; and (4) Trescott 2007: 241.

Wang Sihua was well suited to these tasks. He was born in 1904 in coastal

Laoting County 乐亭县 in the northern province of Hebei 河北 and as a student at

Beijing University in the 1920s, he enjoyed frequent contact with fellow Laoting County native, the firebrand intellectual and founding father of the Chinese Communist Party, Li

Dazhao 李大钊. Under Li’s influence, Wang soon became involved in revolutionary work. In 1926, he traveled to Europe and spent four years studying in France and

England. It was during those four years that he began translating Capital into Chinese.

Wang returned to China in the winter of 1930 and was appointed a professor of political economy at Beiping University and at Sino-French University 中法大学.83 At the same time, he also became part of an underground network of left-leaning professors. With the support of the CCP, and the help of Hou Wailu 侯外庐, he continued to translate Capital.

In June 1936, the results of these labors were published as Capital-Volume 1 by the

International Society 国际学社; an accompanying commentary for the convenience of readers was also published.84

A turn towards more active involvement in revolutionary work took place in 1934, after Wang had been arrested and released on bail following charges of participating in intelligence work. Starting that year, Wang not only collaborated with the Anti-

83 Beijing had been renamed Beiping (Northern Peace) in late June 1928 in the wake of the Guomindang’s decision to relocate the national capital to Nanjing.

84 His biography notes that Wang published under his pseudonym Wang Shenming 王慎铭. The companion volume was titled Zibenlun jieshuo 资本论解说. Hou Wailu 侯外庐 (1903 - 1987) was a prominent historian and philosopher, and one of the major experts on Marxism in the PRC. After 1949 he taught at Northwest University 西北大学. I was unable to track information about the 1936 publication. The earliest extant (possibly abridged) editions appear to be from 1949 and 1950: Wang Sihua . Zibenlun yanjiu congshu 资本论研究丛书. Harbin: New China Bookstore 新中国书局, 1949; and Wang Sihua. Zibenlun jieshuo 资本论解说. Beijing: Sanlian Publishers 三联出版社, 1950.

Imperialist League 反帝大同盟 and the Anti-Japanese Society 抗日救国会, he was also selected as a committee member of the Beiping Second War Zone Mobilization

Committee 北平第二战区动员委员会. After the fall of Beiping in early August 1937,

Wang made his way to Yan’an 延安, arriving there in September. He entered the Party rolls the following year and was assigned to teach Political Economy at the Central Party

School 中央党校. During the rectification campaign that began in 1942, Wang, who by then was directing research on China’s economy within the Central Research Institute 中

央研究院, conducted an agricultural survey in northern Shaanxi. The primary research method for this survey was a form of typical sampling 典型调查. The following year, in

1943, Wang became the deputy director of the Northwest Financial Planning Council, which was under the leadership of Chen Yun 陈云 (1905-1995).85

By the time that the Second World War drew to a close in Asia, Wang had garnered significant experience in economic administration and was an ideal candidate to be sent into the Northeast to contribute towards its administration. In October 1948, he was appointed to the newly established Statistical Survey Unit of the Northeast Finance and Economics Committee 东北财经委员会调查统计处. Among his early tasks was to assess the economic conditions in the Northeast. Wang organized surveys of industrial and commercial activity in Harbin, led investigations into the recent economic

85 Chen Yun 陈云 (1905-1995), regarded in China today as one of the ‘eight immortals’ 八大元老, especially for his role in promoting economic reforms during the post-Mao era, was among the senior leaders responsible for economic and financial affairs during the 1940s and 1950s. For more on Chen’s role post-1978, see: Nicholas R. Lardy and Kenneth Lieberthal, eds. Chen Yün's Strategy for China's Development: A Non-Maoist Alternative. Armonk, NY: M.E. Sharpe, 1983

development of the region, and compiled existing statistical data on the pre-liberation economy.86

It was organizing statistical activity for the newly liberated Northeast (and China), however, that was to become his principal mission. In April of 1950, the Northeast

Statistics Bureau (NSB) was officially set up in Shenyang 沈阳 and Wang was appointed its director.87 Under Wang, the NSB laid the foundation of a comprehensive statistical apparatus, began the systematic translation of Soviet statistics treatises, and disseminated information about the Bureau’s work and the experience it gained through its monthly journal Statistical Work 《统计工作》.88 Over the course of its four year existence, the

NSB would register many firsts in PRC history: the first region 大区 to set up a statistics bureau, the first to publish a specialized statistics journal, the first to publish an Annual

Statistical Report 统计年报, and the first to pass a regional ordinance 地区条例 regarding statistical work.89

86 These activities were coordinated through an institute set up to research Japanese rule and the activities of the Manchukuo Puppet Regime from 1932-1945 (伪满经济研究所). The institute’s work was aided by a staff of thirteen Japanese, who helped edit historical materials and also helped calculate the fixed prices for 1943. They returned to Japan after these activities. See: “Huiwang: lishi shang de dongbei tongjiju 回望: 历史上的东北统计局 (Looking Back: The History of the Northeast Statistics Bureau).” Tongji gongzuo 中国统计 (Statistical Work), 2010.8: 22-23.

87 The offices were originally located near Zhongshan Square; later moving to Heping Distict, Northwest Horse Street. Wang’s Deputy Director was Lin Lifu 林里夫 and the Bureau’s Secretary, Yang Jianbai 杨坚白.

88 Not to be confused with the journal of the State Statistics Bureau, with which it shared a name, and for which it was most likely the inspiration. The SSB’s Statistical Work was published starting in January 1953, and most likely assimilated its predecessor.

89 In August 1954, along with the dissolution of the 6 regions 大区, the NSB was also dismantled and most of its personnel were transferred to provincial stats bureaus; some 50 odd were also sent to Beijing. See: “Huiwang: lishi shang…,” 2010: 22.

Under Wang’s leadership, the Northeast Statistics Bureau witnessed a more than five-fold expansion in its rank and file. In his July 1950 address, Wang had reported that altogether there were 5,860 cadre involved in statistical work (this estimate also included part-time workers). Of these, the preponderance was occupied in tracking industrial statistics (4778), with the remainder divided unevenly across trade, railways, and agriculture.90 Two years later, in his December 1952 address, Wang noted that the NSB now had 21,000 full-time statistical cadre who were aided by about 9,000 part-time workers.91 He was also pleased that during the same time various statistical forms 统计报

告表格 had been unified and rationalized, their overall number having been brought under control: over the nearly three year period, the bureau had managed to reduce the number of forms from 3,318 to a much more manageable 396. This was the first step in controlling duplication and the excess issuing of forms 滥发报表, he explained. 92

The monthly journal Statistical Work 《统计工作》, which began publication in

May of 1950, became the primary venue through which the bureau discussed a wide variety of statistical issues. It continued publication till the end of 1952, and a total of thirty-two issues were produced before the journal was absorbed into its national level namesake, which began publication in January 1953.93 Periodically, important articles

90 Wang Sihua 1952: 8

91 Wang Sihua 1953: 1. The head office in Shenyang at this time had about 125 statistical workers.

92 Wang Sihua 1952: 8 and Wang Sihua 1953: 1. As early as July 1950, nearly 600 forms had already been earmarked for deletion. Other strategies involved merging two or more kinds of forms, correcting forms, and, on rare occasions, issuing new ones.

93 Six monthly issues constituted a volume 卷. In 1952, the system of volumes was discontinued and only an issue number was used. Till Vol. 2 No. 2 it was an internal journal; after that it could be purchased by anyone from bookstores. From Vol. 2 No. 5 onward, Xinhua Press 新华书店 began publishing it. See: “Huiwang: lishi shang…,” 2010: 23.

were also compiled and published as anthologies 选辑.94 The subjects covered in these anthologies can broadly be classified into four categories. The first category was instructional and included essays on how to fill-in statistical tables and forms in industry, agriculture, and other areas; how to assess and address problems with calculation, tables, original records, and accounting; the nature of training; how to conduct analysis 分析; and finally, how to provide explanatory textual material to accompany numerical data 文

字说明. A second category consisted of problems that the bureau was encountering and wanted to address. Examples include excess issuing of forms 滥发报表; problems in leadership; and dealing with conservatism in work. Thirdly, they included the publication of important Statistical Laws and Regulations. A final fourth category dealt with the various different kinds of statistical activities carried out by the bureau, and included summaries of work and responsibilities; descriptions of surveys of industry (state and private), crop production, as well as some early essays on typical sampling 典型调查.

The picture that emerges from such categorization is of a clear emphasis on establishing procedure, unifying systems, and training people to do statistical work. There is surprisingly little discussion of actual statistical results.

94 The library of the Institute of Economics of the Chinese Academy of Social Sciences possesses the first, third, fourth, and fifth volumes of the anthologies. Similarly, the Tsinghua University Library possesses the fourth and sixth volumes. In all, I was able to access four, though at least six were produced: 《统计工作—第一选辑》, Northeast Statistical Press 东北统计局出版社, February 1951; 《统计工作—第二选辑》 ? 1951; 《统计工作—第三选辑》, Northeast Finance and Economics Press 东北财经出版社, January 1952; 《统计工作—第四选辑》, Northeast Finance and Economics Press 东北财经出版社, October 1952; 《统计工作—第五选辑》, Northeast Finance and Economics Press 东北财经出版社, January 1953；《统计工作—第六选辑》? 1954. Thus far, I have been unable to track individual issues of the NSB’s Tongji gongzuo 统计工作. They are not part of the digitized China Academic Journals (CAJ) database and were not available in various libraries across Beijing. It is possible that individual issues can still to be found in libraries or archives in Harbin, Shenyang, or Changchun.

The Primacy of Exhaustive Enumeration

In terms of method, the system built under Wang consisted of two principal means of statistical data collection. The first, and by far the more significant, was exhaustive enumeration via the Complete Enumeration Periodical Report System 全面定期统计报

表制度.95 Its bedrock was the institutionalization and rationalization of original records

原始记录. The idea was that data would be collected primarily through a vast network of comprehensive and periodic statistical reports that would span all sectors of the economy.

As Wang noted, each and every table had clear instructions for how to fill-in data and on the timing and sequence of filing forms.96 Three important laws were passed in 1950 to facilitate the process:97

1. The April 10 Law, Regarding the Strengthening of Statistical Work 《关于加

强统计工作的指示》, a notification 通知 of which was reissued on July 8th.

This law established the NSB and called for the setting-up of a centralized

system of reports and required that statutory forms 法定表格 be used at and

by all levels. All other pre-existing forms were to be discarded 废除.

2. The Law of August 22, Regarding the Centralization of the Issuing of

Statistical Numbers 《关于统一公布统计数字的决定》. This law required

95 “Complete Enumeration Periodical Report System” is how the Chinese themselves had translated 全面定期统计报表制度. See: WJB 105-00530-05: 11-12

96 “我们的每一种表格都明确的规定了填报方法、提报时间和提报程序.”

97 For the texts of these laws, see: Tongji gongzuo—diyi xuanji 统计工作—第一选辑, February 1951: 127- 132.

that all reported data be checked by the corresponding statistical office before

being sent to a higher level within the statistical organization.

3. The Law of October 13, Temporary Regulations for Statistical Reports in the

Northeast Region 《东北区统计报告暂行规程》. This was the first

Regional Ordinance 地区条例 for statistics in post-1949 China and clarified

the relationship between different levels of the statistical apparatus. Such a

system also enforced the use of a common set of forms and tables.

In other parts of the country, where statistics bureaus had yet to be set up, similar directives were issued a few months later by the city or provincial governments through the authority of the Statistics Office of the Central Planning Commission’s Finance and

Economics Committee 中央人民政府政务院财政经济委员会统计处. For instance, in

Beijing, a directive to establish and strengthen statistical work was issued on 15

December 1950. It designated the city’s Research Office to act as the coordinating and distributing agency, to which all statistical tables must be sent.98

The following year, in March 1951, the NSB issued The Decision Regarding

Trials of a Two-Track Statistical Reporting System 《关于试行统计报告双轨制的决

定》, which set up a two-track system for reporting and collecting data. In the first track, grassroots units reported data to the enterprise (business, company, or organization) with which they were affiliated, and that enterprise then prepared a consolidated statement for forwarding to the NSB. But these grassroots reporting units were also required to report data using the statistical reporting system. Thus, in theory, data about the same unit

98 BMA 136-001-00014: 1-6; and BMA 002-002-00223: 1-6. Along with the directive, two sets of instructions were also provided, explaining reporting methods 统计报告制度试行办法 and detailing methods to review forms 表格审查办法.

arrived at the statistical offices in Shenyang via two parallel routes: one through the enterprise, the other through the statistical system. The periodicity of these reports could be as short as a ten-day cycle, or as long as an year, and included several possibilities in between: monthly, seasonal, semi-annual, and so on. These reports would be sent up through each successive level till they reached the headquarters of the NSB in Shenyang

(and later Beijing). Some of these seasonal and annual reports were also published, such as the Northeast Report on Implementation of the National Economic Plan 东北区国民

经济计划执行情况的报告, which was published starting in 1951.99 The first Annual

Report 年报 was published on 1 April 1952 and provided data for all of 1951. By1954, this system had expanded to include data from industry, capital construction, agriculture, trade, transport and communication, private enterprise, as well as numerous other fields.

When data was required with an irregular level of expediency, one-time censuses

普查 were also employed. A national census of industry was completed by the end of

1950, and for the first time gave planners in Beijing an idea of the basic production situation in each factory across China. On the factory-side it was hoped the census would help understand production, discover latent potential 后备力量, correct management shortcomings, and implement economic calculation. The census also raised awareness at the factory level about the importance of statistical work, partly through the many additional statistical cadre who had been trained during the census and were now familiar with production processes and related technical issues. The NSB was at the forefront of this survey, carrying it out from April through December of that year. A census of private

99 In the third season it was changed to 《人民经济计划执行情况的报告》and subsequently, in 1952, it was changed to a six-monthly report. “Huiwang: lishi shang…,” 2010: 22-23.

industry was begun shortly thereafter, on 19 July 1950, via the Decision Regarding

Conducting a Census of Private Industry in the Northeast 关于进行全东北私营工业普

查的决定. This census covered six provinces and five cities, enumerated 39,539 private industrial units and took eight months to complete.100

Unlike in industry, statistical work in agriculture did not easily permit exhaustive enumeration, wither via a periodical report system or through one-time censuses.

Accordingly, the second method of data collection relied on sample surveying. The NSB designed surveys to produce statistical estimates of agricultural production. These surveys were invariably based on some form of typical sampling 典型调查, random sampling 随机抽样调查 having been ruled out on account of its mathematical roots. In designing them, the Chinese drew upon not just Soviet expertise but also their own experiences from the 1930s. They discovered that Soviet experts, in general, were not overly enthusiastic about the use of such surveys. 101 In the materials produced by the

NSB, the Soviet statistical experts’ discounting of surveying is perhaps best expressed by the absence of the word for representative-ness 代表性.

The CCP’s experience conducting rural surveys during the 1930s was thus a critical factor in determining the survey design of the early 1950s surveys. In his memoirs,

Xue Muqiao 薛暮桥, the economist and first director of the national State Statistics

Bureau, observed that the purpose of the 1930s surveys had been to estimate production and organize the levying of taxes. Every province and city had established Production

Survey Committees or Production Survey Offices, which were led by the head of the

100 These were categorized as those using machines, or those with 3 or more workers.

101 For more on the place of random sampling and Soviet experts, see Chapters III and VIII.

local Planning Committee. These committees and offices used the method of meetings 会

议的方式 to disseminate what they were doing at the county level. Operating on the maxim “divide by type, select [a] typical, really cut [grain] to really measure,” 8,000 surveyors investigated 1,175 typical 典型 villages.102 First, sown area was surveyed; then areas were divided based on the nature of the harvest (good or bad); typical cases were selected from within these divisions and surveyed after which overall estimates were generated based on the relative weights of good and bad harvest areas.103 Such typical sampling formed the basis of several other surveys carried out during this period. These included surveys of rural surplus, livestock, and farms.104

Other Activities

Besides the collection and collation of data, other statistical activities included the calculation of price indices, the periodic rectification of original records, and the cleaning of statistical tables. In 1950, Soviet methods were used to compile a Price Index. A common program for calculation was subsequently assigned to all provinces and cities so that they could calculate local price indices. For instance, a survey of 2,000 products was carried out in Shenyang in August of 1950 and the following month the NSB released the

Shenyang Wholesale Price Index 沈阳市私商批发物价指数. In October of that year, the

102 “划分类型，选择典型，实割实测.” See: Yang Jianbai 杨坚白 1953. “苏联专家帮助我们建立了科学的统计工作,” Renmin ribao, February 05, pp. 3. As Xue would note, this was basically a rather simple form of survey sampling. See: Xue 2006: 174.

103 Xue Muqiao 薛暮桥. Xue Muqiao—huiyilu 薛暮桥—回忆录 (Xue Muqiao—Memoirs). Tianjin: Tianjin People’s Press, 2006, 174.

104 Various other surveys were conducted: In addition some other surveys, including: investigations of rural surplus, production surveys, farm censuses, rural ancillary activities and livestock surveys, surveys of private industry with more than ten workers, etc. See: “Huiwang: lishi shang…,” 2010: 22-23.

NSB released a State-Owned Whole Sale Price Index 国营批发物价指数. From

December onwards, indices were published monthly in the Northeast Daily 《东北日

报》. The following year, the NSB released a price index for the Northeast’s ten big cities. After that, provincial and city-level indices began to be calculated in the Northeast on a ten-day, monthly, or some other predetermined time-cycle.105

Rectification of original records 整顿原始记录, another practice that would continue through the rest of the decade, was initially begun in the Northeast in 1951 in statistical data for industry, capital construction, and trade. Over the following two years, most industrial units went through two rounds of rectification and their data was widely considered to be reliable. Local state-owned industrial units also went through at least one round of rectification. In capital construction, rectification was modeled on that done in the industrial cities of Angang 鞍钢 and Fushun 抚顺. These cities had begun publishing a newspaper, Section Leader Daily 《组长日报》, which shared basic information on the changing work conditions on a daily basis: work completion, work/labor time, worker’s development, use of materials, and so on. By collecting all such data into one report, these daily papers became the basis for unified, regular, and accurate statistical data about these industries. This experience was subsequently shared with the rest of the country and these methods used everywhere.106

The clean-up of statistical reports was also carried out twice. The first instance was in 1950, per the orders of the Northeast People’s Government. The second instance was in 1953, following a directive from the recently established State Statistics Bureau.

105 See: “Huiwang: lishi shang…,” 2010: 22-23.

106 See: “Huiwang: lishi shang…,” 2010: 23.

The clean-up involved getting rid of illegal reports as well as unnecessary reports and in particular helped reduced the burden on grassroots units. It was also effective in raising the standards of cadre, especially in the agricultural sector.

“A New Type of Standardized Statistical Work”

Taken together, these activities begin to provide a catalog of the basic contours of statistical work in the Northeast. Wang’s “new type of standardized statistical work” (新

型正规化的统计工作 xinxing zhenggui de tongji gongzuo) had three basic features. The first of these was ‘extensiveness’ 广泛化. For Wang this meant that socialist statistics, unlike its bourgeois version, had mass character 大众化. In other words, it ran through every sector of the economy and society:

Not only did it include all branches of the national economy, but also culture, education, and health care; its content was extremely extensive, and the procurement of materials, the collation of numbers, depended on a comprehensive and unified reporting system.107

A unified reporting system also meant that each person and each element within the system was absolutely critical. Even if one person’s work suffered from a fault 毛病, then the entirety of statistical work was at fault.

Accordingly, a second characteristic was the system’s ‘completeness’ 整体性.

This completeness was a corollary to the extensiveness in that it sought to conceive of the statistical work of every unit as an integral part of a unified statistical apparatus. The key to achieving such completeness required standardization of forms and types of data so

107 Wang 1952: 13-14.

that comprehensive and comparable data were easily available. Equally important was the creation of norms for entering data on forms and for their accompanying textual explanations. 108 Since the integrity of such an extensive and complete system could easily be threatened by a kink in any one level, it was imperative to create a unified structure of reporting that permitted data (i.e. forms and tables) to be sent up and down different levels of the statistical and administrative system.109

Such extensiveness and completeness were the basis for the third feature of socialist statistics, namely its ‘objectivity’ 客观性. Unlike bourgeois statistics, which was predicated on profit and meant that statistical data were frequently distorted 歪曲 to serve the interests of the few and protect their business secrets, the New Statistics was unafraid of exposing weakness. Quite to the contrary, it welcomed criticism and even took part in self-criticism in the pursuit of objectivity and truth 真实性. 110

None of the above three ideals could be achieved without a disciplined cadre of statistical workers who could calculate and analyze the data. Thus Wang noted that discipline undergirded the entire edifice of statistical work. Discipline meant that numbers were diligently recorded, that reporting deadlines were strictly observed, and somewhat oddly, since Wang had just accused capitalists of the same failing, that all statistical numbers were guarded like secrets.111

108 Wang 1952: 14.

109 Wang 1952: 14.

110 Wang 1952: 14.

111 Of course, unlike private secrets, these were the public’s secrets, i.e. the nation’s secrets, to be kept safe from the prying of the bourgeois nations.

These features—extensiveness, completeness, objectivity—exemplified what was deemed scientific by Wang because only by counting everything could an objective sense of social reality be achieved.112 Such exhaustive counting in turn was the basis of the socialist corrective to bourgeois statistics. Indeed, the implication was that a system that was socialist was by definition also scientific, and therefore, correct. The three terms— socialist, scientific, and correct—had become mutually interchangeable. Several years later in 1957, the Indian statistician P.C. Mahalanobis would be impressed by these very qualities of the by then national statistical system, noting that it was “appreciably better than that in India in respect of coverage, availability, and accuracy of data required for purposes of planning and current policy decisions.” 113 The system described by Wang, and one that was first attempted in the Northeast was no doubt an idealization. On the ground, a variety of challenges existed: some on account of local variation across China, some as problems of calculation, and others because of incentives and disincentives that were systemic in nature. Many of these will be explored in later chapters. For now, let us turn to an alternate method of ascertaining social fact, which drew its authority directly from Mao Zedong himself.

112 For a brief but insightful discussion on the ‘reality’ of statistics, especially through an analysis of its practitioners, see: Alain Desrosières. “How Real Are Statistics? Four Possible Attitudes.” Social Research 68, no. 2, (Summer 2001): 339-355.

113 P.C. Mahalanobis. Some Impressions of a Visit to China 19 June – 11 July 1957, p. 12-13. Unpublished typed manuscript in the P.C. Mahalanobis Memorial Museum and Archives (PCMMA), Kolkata.

Maoist Social Investigation as Method

Mao is known to have personally conducted at least two investigations into social conditions in the countryside, the first in 1927 and the second in 1930.114 The reports of these investigations, along with other writings such as his essay “Oppose Book Worship”

(1930) and “On Practice” (1937)115, served to articulate a form of social investigation 社

会调查 that was at once at odds with not only the bourgeois ‘social survey movement’ of the 1930s but also with established Marxist methods of social science research, in China and elsewhere.116 For Mao, the only way of understanding a situation was to carry out several “thorough investigations” 周密的调查 of selected cities and villages and use

Marxism-Leninist class analysis. “Only in this way,” he contended, was it possible “to gain the most fundamental understanding of China’s social problems.”117

Mao’s principal method was to carry out ‘investigation meetings’ 调查会. This meant gathering a small select group of informants from whom he would gather not just knowledge about the situation on the ground, but also information about local group- dynamics, solidarities, and inter-personal tension, all with a view to deploying such

114 In 1927 he spent 32 days, from January 4 to February 5, in five counties in Hunan. His report was published as the famous Report on an Investigation of the Peasant Movement in Hunan. A few years later, in May of 1930, he conducted a survey in the town of Xunwu, which lay in southern China where the provinces of Jiangxi, Fujian, and Guangdong met. See: Mao Tse-Tung. Report on an Investigation of the Peasant Movement in Hunan. Peking: Foreign Language Press, 1965 (3rd edition; first edition 1953). Translated from the Chinese original 湖南农民运动考察报告 in: Selected Works of Mao Tse-Tung, Volume I. Peking: People’s Publishing House, 1952), pp. prefatory material, 1. On the Xunwu investigation, see: Roger Thompson, tr. Report from Xunwu. Stanford: Stanford University Press, 1990.

115 On the controversy over when exactly this essay and the essay on contradiction were actually composed, see: Nick Knight. “Mao ZeDong's on Contradiction and on Practice: Pre-Liberation Texts.” The China Quarterly, no. 84 (Dec., 1980): 641-668.

116 On the former, see Lam 2012; and on the latter, see: Chiang 2001.

117 Mao Zedong 毛泽东 1941.《农村调查的序言和跋》. See: Selected Works of Mao Tse-Tung, Volume III. Beijing: People’s Press, 1966, 747.

information to achieve specific political ends. This method valorized the personal experience of conducting such interviews. It discounted methodological issues such as representativeness and comprehensiveness in favor of the ability to effectively create change and bring about consensus.118 The method’s importance and its impact before and after 1949 have been the subject of earlier work, most notably that of Thomas Mullaney,

Yung-chen Chiang, and Roger Thompson.119 While Chiang and Mullaney disagree on the degree to which Mao was self-reflexive about the methodological strengths or weaknesses of his method, both insist that the method retained its primacy in the years after 1949.120 While that may indeed have been true in some spheres, in the realm of statistics this method of social investigation was very much a source of tension for its practitioners and theoreticians.

Two things within Mao’s method were particularly at odds with the socialist statistics espoused by Wang Sihua. The first was that Mao placed tremendous value on individual experience—drawing authority from one’s presence in the field, personally

118 For Mao, the stakes were admittedly different: an interest in equity, self-reliance, decentralization, anatagonism to bureaucratism and elitism, collective aims and discipline, and the continuing desire for mass struggle. For a discussion of these themes, see: Mao Zedong. A Critique of Soviet Economics by Mao Tsetung (translated by Moss Roberts; annotated by Richard Levy, with an introduction by James Peck). London: Monthly Review Press, 1977, 13, 14, and 17-18. Also see Xue Muqiao 1957: 13, where the method is listed as one among a set of supplementary methods used by statisticians.

119 For detailed treatments of Mao’s social-ethnographic method that explore some of these themes, see: Chiang 2001:150-58; Mullaney 2010: 97-101; and Thompson, tr., 1990.

120 As Chiang notes: “The power and prestige of Mao were such that his vision and approach dictated the way social investigations were conducted in the People’s Republic until the 1970s” adding, “his group fact- finding sessions became the ubiquitous format according to which all social investigations have been conducted in China.” To which Mullaney is in general agreement: “Following 1949, this mode of field research—one in which the boundary between information and transformation was porous by design— continued to be employed.” See: Chiang 2001: 156, 157, Mullaney 2010: 99.

carrying out a survey. 121 Personal subjectivity was the basis for the pursuit of objective reality. The theoretical basis of this position was best articulated in his essay “On

Practice.” In it, he noted that “the peoples with real personal knowledge are those engaged in practice the wide world over.”122

If you want to know a certain thing or a certain class of things directly, you must personally participate in the practical struggle to change reality, to change that thing or class of things…Only through personal participation in the practical struggle to change reality can you uncover the essence of that thing or class of things and comprehend them…All genuine knowledge originates in direct experience… [and] there can be no knowledge apart from practice.123

Mao exemplified this position in practice in his Hunan report. From the reports very first sentence, he placed himself squarely in the middle of the investigation:

“During my recent visit to Hunan I made a first-hand investigation of conditions in the five counties…I called together fact-finding conferences in villages and county towns, which were attended by experienced peasants and by comrades working in the peasant movement, and I listened attentively to their reports and collected a great deal of material.”124 [Emphasis added]

He was present ‘first-hand,’ ‘fact-finding,’ and ‘listened attentively.’ The local informants he preferred to meet were those who had a good understanding of local

121 For a discussion of how during the Republican years personal experience (and hardship) during the conduct of a social survey was a key source of one’s authority, see: Tong Lam 2012. Also, Mullaney 2011: 96-97.

122 “Leaving aside their genius, the reason why Marx, Engels, Lenin and Stalin could work out their theories was mainly that they personally took part in the practice of the class struggle and the scientific experimentation of their time; lacking this condition, no genius could have succeeded…” Mao Tse-Tung. “On Practice.” In Selected Works of Mao Tse-Tung. Volume I. Peking: People’s Publishing House, 1952. Accessed from: https://www.marxists.org/reference/archive/mao/selected-works/volume-1/mswv1_16.htm

123 Mao Zedong 1952 (“On Practice”). This ideological position would also lead to the famous dictum: “no investigation, no right to speak.”

124 Mao 1965: 1.

conditions.125 But he would add that picking up information less systematically was just as important: “I met all sorts of people and picked up a good deal of gossip.”126 Later in the report he would note, for instance: “When I was in the countryside, I heard the gentry say…”127 Mao’s authority to speak also drew from his subjective position of being personally present in the field, which allowed him not only to interview informants but to also witness events as they unfolded: “I myself saw two [evil gentry] captured in

Changsha…128 Personal testimony was augmented by direct quotes from the people he met and interviewed, thereby adding the kind of local flavor that statistical data collected through a periodical report system could never hope to attain: “Whenever their conversation turns to the heads of the tu and tuan, the peasants say angrily, ‘That bunch!

They are finished!’”129 It was such experience that Mao would extoll but also claim was imperative because it allowed you to gather information and make assessments based on that.130 And the process was not without excitement of impending revolutionary change:

125 For somewhat contrasting discussions of Mao’s group-interview process and the selection of interview candidates, see Mullaney 2010: 97; Chiang 2001: 155-156.

126 Mao 1965: 7; also see another instance on p. 6.

127 Mao 1965: 13.

128 Mao 1965: 29.

129 Mao 1965: 35; also see: Mao 1965: 16-17, 34, 41 for other examples. In most cases these were representative quotes, and not the specific sayings of any one person that he was reporting.

130 On Mao’s understanding of cognition and knowledge, it is useful again to consult “On Practice,” where he notes: “…the dependence of rational knowledge upon perceptual knowledge. Anyone who thinks that rational knowledge need not be derived from perceptual knowledge is an idealist. In the history of philosophy there is the “rationalist” school that admits the reality only of reason and not of experience, believing that reason alone is reliable while perceptual experience is not; this school errs by turning things upside down. The rational is reliable precisely because it has its source in sense perceptions, otherwise it would be like water without a source, a tree without roots, subjective, self-engendered and unreliable. As to the sequence in the process of cognition, perceptual experience comes first…” See: Mao Zedong. 1952 (“On Practice”).

“If you revolutionary viewpoint is firmly established and if you have been to the villages and looked around, you will undoubtedly feel thrilled as never before.131

Mao’s enthusiasm for such investigative strategies was accompanied by a lack of precision where actual numbers were concerned. Throughout the Hunan report he frequently dealt in whole numbers or expressed ranges that could, at best, only be taken as having token representative value. Thus, in discussing the number of spears among spear corps in the five counties he would offer numbers such as 100,000 in one county

( Hsianghsiang), or ranges such as “70,000-80,000 or 50,000-60,000 or 30,000-40,000” in the remaining counties.132 The number of rifles amongst the landlord class was discussed with somewhat greater precision, but still with a propensity for neat, whole numbers: “An average of 600 rifles for each country would make a total of 45,000 rifles for all the seventy-five counties [in central Hunan]…”133 Elsewhere, in similar fashion, he noted that “a crowd of fifteen thousand peasants,” demonstrated at the houses of six evil gentry.134 Finally, it remains unclear how he arrived at the results of his surveys of

Changsha county (where he claimed that poor peasants comprised 70%, middle peasants

20%, and landlord and rich peasants the remaining 10%) and Hengshan (85% of peasant leaders are capable and hard-working, while 15% retain some bad habits).135

131 Mao 1965: 8.

132 Mao 1965: 37.

133 Mao 1965: 36.

134 Mao 1965: 27.

135 Mao 1965: 18-19, 21. On the impossibility of finding out what kind of data Mao actually collected during his investigations because the field data and notebooks are all likely lost, see Chiang 2001: 151.

The second feature of Maoist social research at odds with socialist statistics was the invocation of a telling example as a representation of a larger social reality.136 As

Mao would state in the Hunan report, “quite a few such cases have occurred. There was a case recently at Machiahao, Hsiangtan County, where…,” or “once in Hsiangtan when a district peasant association and a township peasant association could not see eye to eye.”137 In similar fashion, when discussing the effect that the execution of one landlord can have, he cited the example of his native county of Hsiangtan. Socialist statistics was intrinsically uncomfortable with such a strategy of using the part as a representative of the whole. The use of an example was acceptable for illustration, but could not be the basis of one’s knowledge about the totality of social reality.

If Mao’s ethnographic method was a way of ascertaining social fact, so indeed was socialist statistics.138 For PRC-era socialist statisticians, Mao’s method was essentially a form of typical sampling. In his July 1951 speech, Li Fuchun had spoken of methods of data collection that were supplementary to the periodical report system, pointing specifically to Mao’s method of typical investigations using investigation meetings 调查会, noting that they had been in use for well over a decade.139 But within socialist statistics such typical sampling was considered a secondary method, only to be

136 The rhetoric of example as discussed by Alain Desrosieres has a long history and remains a powerful form of social analysis to this day. Till the rise of randomized survey sampling, it provided the basic ideas behind systematized forms of purposive or typical sampling. For more on the rhetoric of example and the history of random sampling, see Part III below. Also see: Alain Desrosières 2002: 210-235 (“The Part for the Whole: Monographs or Representative Sampling”).

137 Mao 1965: 27.

138 Yang Jianbai 杨坚白. “Sulian zhuanjia bangzhu women jianli le kexue de tongji gongzuo 苏联专家帮助我们建立了科学的统计工作 (Soviet Experts have Helped us Establish Scientific Statiatical Work).” Renmin ribao (People’s Daily), February 05, 1953: 3.

139 See: DSJ: 5.

used when periodical reports and exhaustive enumeration were impossible. Wang’s “new type of standardized statistical work” described above was predicated on abstraction and the ability to collate vast amounts of data. An individual statistician or statistical cadre drew his authority and knowledge neither from his personal experience of the field nor from having participated in data collection at the village and county level, but primarily from the power of abstraction. It was not the observation of habit 习惯, the application of common sense 常识, or any other individual personal experience that was critical, but rather specialized training that allowed him to correctly record data, process it, and analyze it. The knowledge and practice that Mao spoke of were not the knowledge and practice that socialist statistics would have its statisticians specialize in.140 To further complicate matters, Mao’s method was also hardly the kind of representative typical sampling that had actually been conducted in the Northeast under Wang.

Socialist statistics remained in uncomfortable tension with the Maoist method for much of the decade of the 1950s. In fields that had putatively ceased to exist, such as sociology, the methodological vacuum was filled by the Maoist method. But in the case of statistics, the method went against Soviet experience as well as the very framing of socialist statistics. Given the latter’s dominance during the 1950s, the Maoist method served to make it harder to convince cadre, in particular grass roots cadre, of the efficacy and need for something as expansive and alienating as the periodical report system and exhaustive enumeration. It was only after 1958, when socialist statistics itself was abandoned during the Great Leap Forward that the Maoist method would come to occupy center-stage as the principal statistical method of social investigation.

140 “There can be no knowledge apart from practice.” See: “On Practice.”

Conclusion

For Wang Sihua, three features taken together exemplified the strengths and scientific character of the Northeast’s statistical system: extensiveness across economy and society, completeness of the statistical system, and objectivity about social fact. By placing the periodical report at the core of this system, he articulated a preference for exhaustive enumeration. Only in this way, by counting everything, did he and other Chinese statisticians claim to achieve an objective sense of social reality. In contradistinction to their system of high abstraction lay the personal, anecdotal, and ethnographic model of

Mao’s early social analyses. For much of the decade of the 1950s, the abstraction of the periodical report was the normative ideal in the world of PRC statistics. That is not to say that its alienating tendencies did not generate tensions (as will be discussed below in Part

II), but this tension was largely kept in abeyance. It was only with the launching of the

Great Leap Forward in the summer of 1958 that Mao’s method would be rebadged and elevated above other methods of statistical data gathering. We shall return to that in Part

III.

CHAPTER III

Lies, Damned Lies, and (Bourgeois) Statistics: Ascertaining Social Fact

Where did the inspiration for and confidence in exhaustive enumeration via a periodical report system come from? In the most concrete terms it came from the desire to emulate the success of the Soviet Union. To the Chinese (and many other socialist states after

1945), the Soviet Union “provided the only historical model for industrializing an economically backward country under socialist political auspices.”141 The Chinese

Communist Party’s (CCP) attitude toward such ‘advanced experience’ 先進經驗 had been articulated by none other than Mao Zedong 毛泽东 himself. In 1949, he had stated:

“The Communist Party of the Soviet Union is our best teacher and we must learn from it.”142 The Soviet Union had responded with enthusiasm. A principal element of Soviet aid was sharing technical knowledge. By 1960, when the Sino-Soviet split became final, as many as 18,000 Soviet and Eastern European Experts had spent time in China sharing their expertise.143 A handful among them were statistical experts—consultants, administrators, and teachers—whose advice was instrumental in the organization of the

Northeast Statistics Bureau and later, the State Statistics Bureau.

141 Maurice Meisner. Mao’s China and After: A History of the People’s Republic. 3rd Edition. New York: The Free Press, 1999, 109. Mao had articulated much the same point in his essay “On Practice,” where he noted: “The struggle of the proletariat and the revolutionary people to change the world comprises the fulfillment of the following tasks: to change the objective world and, at the same time, their own subjective world—to change their cognitive ability and change the relations between the subjective and the objective world. Such a change has already come about in one part of the globe, in the Soviet Union. There the people are pushing forward this process of change. The people of China and the rest of the world either are going through, or will go through, such a process.” See: Mao Tse-Tung. “On Practice.” In Selected Works of Mao Tse-Tung. Volume I. Peking: People’s Publishing House, 1952. Accessed from: https://www.marxists.org/reference/archive/mao/selected-works/volume-1/mswv1_16.htm

142 Mao Zedong (Mao Tse-tung). “The People's Democratic Dictatorship.” Speech delivered “In Commemoration of the 28th Anniversary of the Communist Party of China, June 30, 1949,” in Selected Works, Vol. 5. New York: International Publishers, n.d., 411-423. This was also the speech in which Mao used the phrase “leaning to one side” to describe China’s policy. On the speech and this policy, see: Odd Arne Westad, ed. Brothers in Arms: The Rise and Fall of the Sino-Soviet Alliance, 1945-1963. Stanford: Stanford University Press, 1998, 247-250. For other similar statements extolling the virtues of the Soviet Union, also see Thomas P. Bernstein and Hua-yu Li, eds. China Learns from the Soviet Union, 1949- Present. New York: Lexington Books, 2010, 1.

143 Shen Zhihua. Sulian zhuanjia zai zhongguo (1948-1960) 苏联专家在中国 1948-1960 (Soviet Experts in China). Beijing: Xinhua Press, 2003.

The confidence in exhaustive enumeration was not, however, a mere artifact of emulation (or practicality); rather, embedded in this choice, was a deep-seated and sophisticated theoretical criticism of statistics as it had evolved and was evolving at that time. Put differently, the confidence in exhaustive enumeration came from a particular understanding of how to count and ascertain social reality, and the place of mathematics in that process. Within this theoretical framework, exhaustive enumeration came to serve as the basis for the socialist corrective to what came to then be labeled “bourgeois” statistics. What socialist statistics and its preference for exhaustive enumeration did was to deny the universal applicability of mathematics in general and mathematical statistics in particular.

The purpose of this chapter, then, is two-fold: first, to uncover and understand the socialist critique of statistics; and second, to investigate the Soviet statistical experts who spent time in China and who were instrumental in socialist statistics’ rise to a position of epistemological and administrative dominance. The chapter proceeds with a discussion of the 1950s (or post-1945, more accurately) as a period where the imperative to ascertain social fact took on added urgency across the world. There existed, however, competing approaches to ascertaining social fact and the following sections dwell on the rise of socialist statistics, in particular its rise in the USSR, and contrast it with other approaches to statistics. I then turn toward an exploration of the Soviet experts who spent extended periods of time in the PRC, examining the variety of ways—teaching, translation of textbooks, and consultation—through which their expertise was mobilized by the Chinese

State Statistics Bureau as it sought to disseminate the correct understanding and implementation of socialist statistics.

Ascertaining Social Fact in the 1950s—A Global Imperative

Over the last two to three decades the history of quantification, probability, counting, and facticity in the early modern and pre-twentieth century world has garnered increasing amounts of attention. The rich studies that have emerged have helped establish quantification and statistics as independent subjects of inquiry, situating them at the very heart of the modern and post-modern experience.144 Among the many insights that have been offered by this rich literature is the significant role played by the state and by the study of the social world (in addition to the natural world), particularly through the expansion of bureaucracies and administrative duties, in the realization of what Ted

Porter has called the ‘statistical perspective.’145

If these were essentially nineteenth century developments, largely limited within the boundaries of nation states, or in certain cases, imperial formations, the twentieth century presented challenges of a rather new nature. The experiences of World War I led to the desire for new international/global institutions of arbitration and peace. The most well-known example of this is, of course, the founding of the League of Nations. One of the activities the League was involved in was the design and organization of commensurable economic and financial data.146 At the same time, the duties of the state

144 See, for instance, Porter 1986, 1995; Hacking 1990; Poovey 1998; Kruger, et al 1987; Patriarca 1996; and Desrosiers 1998.

145 “…what we might call the statistical perspective owed more to the work of census bureaus and statistical offices concerned with health, crime, trade, and education than to the pure theory of mathematical probability.” Theodore M. Porter. “Chance Subdued by Science.” Poetics Today 15, no. 3 (Fall 1994): 474. Elsewhere, Michael Gordin has spoken of “the importance of statistics as a lens for examining the rise of Soviet governance,” a claim that would apply more generally to the rise of governance as well. Michael D. Gordin. “Statitique et revolution en Russie: Un compromisie impossible (1880-1930), and: L’anarchie bureaucratique: Statistique et pouvoir sous Staline.” Kritika: Explorations in Russian and Eurasian History 5, no. 4 (Fall 2004) (New Series): 803-810.

146 As early as August 1919 the League convened meetings to discuss international cooperation in statistics. It mooted the establishment of an International Statistical Commission, which met for the first time in Paris

also expanded far beyond their nineteenth century norms. Fueled in part by the rise of

Communism and by the experience of the Great Depression, which, in turn, facilitated the rise of Keynesian economics, the state began to take on ever expanding responsibility in regulating the economy and providing services to its citizens.147 From a technological standpoint, each of the World Wars also played their role, as they provided opportunities for the design, experimentation and deployment of advanced techniques of data collection and analysis. Operations Research, for instance, was a field that grew substantially on the basis of war time exigencies.148 After World War II, the earlier efforts of the League of

Nations were renewed through the United Nations, which established various institutions for global order and communication, including a Statistical Commission in 1947.

Finally, the collapse of empires that gathered pace in the wake of WWII heralded the arrival of several new nation states, each with its own modernizing mission. A principal way to be modern, and demonstrate one’s modernity to others, was to know oneself and at the same time exhibit this self-knowledge to others. A statistical system that gathered and disseminated such knowledge using numbers, and which employed the latest methods, became the hallmark of a modern nation state.149 For a long time, a

in October of the following year. For more, see: J. W. Nixon. A History of the International Statistical Institute 1885-1960. The Hague: International Statistical Institute, 1960, 28-29, 42.

147 For more on this, see for instance, Desrosieres 1998. Mention can be made here of the rise of National Income Accounting, unemployment statistics, macroeconomic theory with its focus on employment, economic planning, and so on.

148 On the wartime histories of operations research in the United Kingdom and the United States, see: Maurice W. Kirby. Operations Research in War and Peace: The British Experience from the 1930s to 1970. London: Imperial College Press, 2003; Randall Thomas Wakelam. The Science of Bombing: Operational Research in RAF Bomber Command. Toronto: University of Toronto Press, 2009; and Charles S. Schrader. History of Operations Research in the United States Army, Vol. 1 1942-1962. Washington, D.C.: US Army, 2006..

149 In his discussion of planning in post-colonial India, for instance, Partha Chatterjee has shown how the act of planning for a new nation was formulated through the logic of numbers. Statistical indicators were

standard imperial argument had been that locals did not even possess an accurate and scientific means of ascertaining (social) fact. Empire’s civilizing mission, thus, included this ‘gift’ of fact and numerical representation. In the Chinese case, the discourse about the lack of facts was keenly felt. Almost on par with the nineteenth century Western imperial assertion that China lacked a rational legal system, the lack of factual self- knowledge came to inform a discourse of national humiliation. As Tong Lam has shown in a recent study, the response was the rise of a “social survey movement,” whose principal goal was to assess what China was—to make her legible, knowable, and thereby, eventually changeable.150

By the end of World War II, these developments, and the desire for a new world

[order], contributed in part to what might be termed a more globalized epoch of statistics driven numerical positivism.151 For, each of these developments depended upon and demanded new levels of data collection and analysis. This desire for data, for facts, was summed up by the UN Secretary General Trygvie Lie in 1947. Addressing the inaugural session of the International Statistical Institute’s twenthieth meeting in Washington, D.C.,

Lie observed:

“The free exchange of information on economic and social affairs among all countries in the world is absolutely necessary to economic and social advancement. We cannot cure our troubles unless we know in the first place what those troubles are. Likewise we cannot achieve international precisely the metrics through which progress could be measured and contrasted with colonial under- development. See: Partha Chatterjee. “Development Planning and the Indian State.” In The State Development Planning and Liberalisation in India, edited by T J Byres, 82-103. New Delhi: Oxford University Press, Delhi, 1998.

150 Lam 2012.

151 On how some of these tendencies played out in the articulation of a particular ‘Cold-War rationality’ in the United States, see: Paul Erickson, Judy L. Klein, Lorraine Daston, Rebecca Lemov, Thomas Sturm, and Michael Gordin [Hereafter Erickson, et al]. How Reason Almost Lost its Mind—The Strange Career of Cold War Rationality. Chicago: University of Chicago Press, 2013.

understanding, which is the basis of advancement, unless the peoples of the world are given the facts about each other…There is no substitute for facts, for clear and systematically organized facts. They alone can be relied upon to measure resources and potentialities for progress and to direct policies and actions designed to achieve the objectives of all civilized people.”152 [emphasis added]

Speaking, as he was, at the pre-eminent gathering of international statisticians, the desire for ‘clear and systematically organized facts’ was predicated on a form of numerical positivism—numbers alone could help understand ground realities, design solutions to problems, and plan for the future. The postwar years were, as a result, a time when several statistical innovations achieved tremendous legitimacy, the attendant positivism informing actions in all spheres of human activity.153 Among these innovations can be listed large scale sample surveying (something which we shall return to in Part III), operations research, development theory (most notably W.W. Rostow’s theory about the

Stages of Economic Growth), demographic transition theory, random sampling methods such as Monte Carlo, and the spread of economic planning.

152 Trygve Lie. Opening Address to the International Statistical Institute’s 1947 Meeting. Introduction— International Statistical Conferences, September 6-18, 1947, Washington, D.C. Proceedings of the International Statistical Conference, Volume I. Calcutta: Alka Press, 1951, 151 At the same meeting, discussions were also held to set up an International Association for Research in Income and Wealth. It was felt that “the solution to many important problems in the field could be sought most effectively by the close cooperation of those most familiar with the income and wealth statistics in the various countries…[and that] such an organization would serve to bring the scholars in the field together; to develop greater familiarity on the part of members with work in other countries; and to provide a continuous stimulus to the solution of ever present problems of definition and analysis and to the formulation of new approaches.” See Volume I: 308-311.

153 In 1986, Ted Porter had observed that statistical tools used in modern sciences had been worked out over the previous 100 years or so. The phase of invention was mostly in the nineteenth century and the phase of development during the late nineteenth and early twentieth centuries. Extending his periodization, we can label the rest of the twentieth century starting with WWII as the period of application and refinement. See: Porter 1986: 3.

A Universal Science

Most of these innovations were made possible because of a specific shift in the understanding of statistics, quantification, and probability that occurred during the nineteenth century. Until the late nineteenth century, statistical activities broadly construed progressed along two non-intersecting tracks, one descriptive and the other analytic. The descriptive traced its origins to the seventeenth century “political arithmetic” of William Petty among others, the German school of descriptive statistics (Statistik, from which the modern word originates), and through them to a much more fundamental and age-old global phenomenon—the desire of the state to collect information about its constituents, its land, its produce, and so on. The analytic, which can be traced at least as far back as the sixteenth and seventeenth centuries, was principally concerned with concepts of chance, probability, and estimating and controlling error, primarily in an abstract, mathematical sense.154 The shift that occurred in the 19th century involved a new understanding of error not as a reflection of chaos but rather as variations in a given natural order. As Ted Porter notes, "the acceptance of indeterminism constitutes one of the most striking changes of modern scientific thought. With few exceptions, scientists and philosophers previous to the late nineteenth century would have agreed with

Augustus De Morgan that to say an event occurs by chance is to say that it occurs for no

154 This is no doubt a gross simplification of an era where ideas about probability were in tremendous flux. As Lorraine Daston has shown, a key feature of this period was that probabilists did not differentiate between objective and subjective aspects of probability, the former a feature of the world and the latter based on a degree of judgment or certainty. For more, see Daston 1988. Also see: Poovey 1998.

reason at all.”155 Similarly, Ian Hacking has shown how the erosion of determinism made possible “autonomous laws of chance.”

The idea of human nature was displaced by a model of normal people with laws of dispersion. These two transformations were parallel and fed into each other. Chance made the world seem less capricious: it was legitimated because it brought order out of chaos. The greater the level of indeterminism in our conception of the world and of people, the higher the expected level of control.156

For Hacking and Porter, these developments owed as much to the study of natural phenomena as they did to social phenomena.157 It was not error analysis per se, but rather the use of probability as a modeling tool to analyze variation in nature and society, a mutual cross-pollination across disciplines and intellectual pursuits, which contributed to the evolution of statistics as a distinctly mathematical field of inquiry.

This, in brief, is the liberal story in which statistics emerges as a universal science, capable of studying and discerning patterns in both society and nature. But statistics also has an alternate history, which, at least for a significant portion of the 20th century, enjoyed great legitimacy and widespread use.

155 Porter 1986: 149-150. Porter continues: “Still, this story is not simply one of a new appreciation of the empire of chance. Randomness first attained real standing in scientific thought not as a source of massive uncertainty, but as a small-scale component of an over-arching order. The recognition of chance stemmed not from the weakness of science, but from its strength--or rather, its aggressive imperialism, the drive to extend scientific determinism into a domain that had previously been seen by most as the realm of inscrutable whimsy."

156 Hacking 1990: vii.

157 As Porter notes in Part II of his book: “Quetelet's insight concerning distribution of traits in human populations confirms the more general argument of this book: role of social science in the development of statistical thinking.” See: Porter 1986. In a related way, Michel Foucault too analyzed this shift, though he located it to more to the bourgeois drive for self-identification and the molding of society.

A Critique of Universality

Lenin is famously said to have observed that “socialism is first of all accounting.”158

Within the former USSR he enjoyed fame as a skilled statistician who used numbers to expose the oppression and injustices of imperialist Tsarist rule and, after 1918, as someone who vigorously promoted the development of state statistical institutions.159

Stalin, too, is known to have remarked on the fundamental importance of statistical activities and data.160 In accepting the centrality of statistics as a tool for understanding and shaping economic phenomena, Marxist and Soviet thinkers and practitioners shared much in common with thinkers in other parts of the world. 161 But in the selection of theory and methods to pursue these goals, they drew a sharp line between themselves and thinkers in a liberal tradition. While Marx was likely cognizant and appreciative of developments in mathematics and statistics made possible by the works of such notable figures as Quetelet, Poisson, and Laplace, Soviet and Chinese statisticians were quick to point out that he restricted his own discussions on the efficacy of statistics solely to the realm of political economy.162

158 V. Lenin. Collected Works, Vol. XXII. Moscow: Progress Publishers, 1960 (Reprinted 1977-1978), 50.

159 On Lenin as a statistician see: S. a. E. S. Kotz. “Lenin as Statistician: A Non-Soviet View.” Journal of the Royal Statistical Society. Series A (Statistics in Society) 153, no. 1 (1990): 73-94.

160 “Comrades, no constructive work, no state activity, no planning is conceivable without proper accounting. And accounting is inconceivable without statistics. Without statistics, accounting cannot advance one inch.” Joseph Stalin. “Organisational Report of the Central Committee.” May 24 1924. Accessed: http://www.marxists.org/reference/archive/stalin/works/1924/05/24.htm

161 “In the U.S.S.R., statistics cannot be separated from economics either in the field of theory, in the establishment of the methodology of study of new economic phenomena, or in practical work in the accounting an analysis of primary statistical data.” See: A. Yugow. “Economic Statistics in the U.S.S.R.” The Review of Economics and Statistics 29, no. 4 (November 1947): 243. Yugow is identified as a “Writer on Russian Economics, New York City,” in the paratext to the journal issue.

162 Xu Qian. “Tongji xueshuo fazhan bianhua de lunkuo 统计学说发展变化的轮廓 (An Outline of the Developments and Changes in Statsitical Theory).” In Tongjixue yuanli ziliao huibian 统计学原理资料汇

For Marx and Engels, and subsequently Lenin, Stalin and other Soviet thinkers, there were no universal laws of economics and statistics. 163 Rather, different historical epochs and different social systems generated different statistical and economic laws.

Accordingly, as A. Yugow noted in a 1947 review of Soviet statistics, it was felt that:

the statistical methods applied in the accounting of the results of private economic activity, and the patterns established for economies in which the chief economic regulating forces are the laws of supply and demand, have proven inapplicable to statistical accounting in a planned state economy.164

The nature of the economy and social organization of a particular country, which, in turn, reflected a particular stage in its history, required its own sets of statistical and economic laws. The distinction between the laws that govern society, and thus appropriately belonged to social science, and those that govern nature, was one that could not be surmounted. On numerous occasions Marx identified statistics as a weapon that could be used in service to the proletariat classes, especially by providing trenchant critiques of existing political-economic conditions and as a tool for the realization of workers’ statistics. In such a formulation, statistics was fundamentally political arithmetic and

Marx regarded William Petty, the originator of the phrase, as the founder of both statistics and political economy.165

编 (Compilation of Materials on Statistical Theory). Beijing: Zhongyang guangbo dianshi daxue chubanshe, 1983, 24.

163 “…Every special historical mode of production has its own special laws of population, historically valid within its limits alone. A single Law of Population properly so called, exists for plants and animals only, and only in so far man has not interfered with them.” See: Karl Marx. Capital, Volume 1: A Critique of Political Economy (Translated by Samuel Moore, Edward Bibbins Aveling, and Ernest Untermann , edited by Frederick Engels). London: Modern Library, 1906, 645.

164 Yugow 1947: 243.

165 “William Petty, the father of Political Economy, and to some extent the founder of Statistics…” See Marx 1906: 179; Accessed online: http://www.marxists.org/archive/marx/works/download/pdf/Capital- Volume-I.pdf). For Marx, William Petty symbolized a leap in quality as far as statistical activities were concerned. See: Xu Qian 1983: 18-21. On Marx and his views on William Petty, also see: Ted McCormick.

While Marx may not have enjoyed the real political power necessary to put many of these ideas about the distinctions between the natural and social world into action, his intellectual heirs were in positions to do so. They were able to apply statistics to statecraft: to use it as a tool not just for analysis and critique but also for administering change. The manner in which they chose to do this had implications for the evolution of statistics. For instance, in picking up Marx’s mantle, Lenin also gave primacy to statistics as a practical tool, going to the extent even of de-emphasizing academically oriented activities among statisticians.166 He was especially keen on making the Soviet Central Statistical Board the primary source for all data relating to society and economy. In a letter to the Board, Lenin noted that “the Central Statistical Board, which lags behind an unofficial group of writers, is a model bureaucratic institution. In about two years’ time it may provide a heap of data for research, but that is not what we want.” He proceeded to demand that “for our practical work we must have figures and the Central Statistical Board must have them before anybody else.”167 [Emphasis in original]

The result, initially in the Soviet Union and subsequently in the places where its influence extended, was to create a distinction in statistical practice and, eventually, in statistics as a discipline. Fueled by such an understanding of human activity and how to

William Petty: And the Ambitions of Political Arithmetic. Oxford: Oxford University Press, 2009. The epilogue, in particular, discusses what McCormick calls Marx’s resurrection of Petty as the founder of modern political economy.

166 In a letter dated May 26 1921 to his friend, the economist, planner and then chief of GOSPLAN, G. M. Krzhizhanovsky, Lenin observed: “The Central Statistical Board should be made into an organisation that does analysis for us, current, not “scientific” analysis…Statisticians must be our practical assistants, not engage in scholastics.” [Emphasis in original] See: Lenin Collected Works, Volume 35: 497-498. Source: http://www.marxists.org/archive/lenin/works/1921/may/26gmk.htm. The latter half of the citation above, about practical assistants, is also quoted in Kotz 1990: 87.

167 V.I. Lenin. “To The Manager of the Central Statistical Board,” dated August 16 1921. See: Lenin’s Collected Works, Volume 33: 30-35. Accessed from: http://www.marxists.org/archive/lenin/works/1921/aug/16.htm

account for it, an understanding of statistics that was specific, context-based, and non- universal, had begun to attain legitimacy by the 1930s.

Statistics [thus] became an instrument for planning the national economy. Consequently, its basis is the Marx-Lenin political economy; it represents a social science or, in other words, a class science. The law of large numbers, the idea of random deviations, and everything else belonging to the mathematical theory of statistics were swept away as the constituent elements of the false universal theory of statistical science.”168

Prior to the 1930s, Russian and Soviet statisticians had enjoyed a fair degree of freedom and saw themselves as participants in the evolving universal discipline of statistics, the bedrock of which was mathematics itself. Keenly aware of global developments in the field, they also actively sought engagement with methods devised outside the Soviet

Union. The application of theory to practical problems was a hallmark of this period.169

But this was also the period during which numbers as measures of performance became absolutely critical. As Michael D. Gordin has observed in his review of two studies on

Russian and Stalinist statistics: “For the Soviet state, given its self-professed character as a scientific state, these numbers become crucial for measures of how much socialism was an improvement over tsarist capitalism…”170 As a result, statisticians came under a

168 S.S. Zarkovic. “Note on the History of Sampling Methods in Russia.” Journal of the Royal Statistical Society. Series A (General) 119, no. 3 (1956): 336-338. See 338 for quote.

169 For recent histories of statistics during Tsarist Russia and the early Soviet Union see: Martine Mespoulet. Statistique et révolution en Russie: Un compromise impossible (1880–1930) [Statistics and Revolution in Russia: An ImpossibleCompromise (1880–1930)]. Rennes: Presses Universitaires de Rennes, 2001; and Alain Blum and Martine Mespoulet. L’anarchie bureaucratique: Statistique et pouvoir sous Staline [Bureaucratic Anarchy: Statistics and Power under Stalin]. Paris: La Découverte, 2003. A useful summary and review of both volumes is: Gordin 2004. For contemporary accounts of early Soviet statistical work and organization, also see: A. Ezhov. Soviet Statistics. Translated by V. Shneerson. Moscow : Foreign Languages Publishing House, 1957; and A. Ezhov. Organisation of Statsitics in the U.S.S.R. Moscow: Progress Publishers, 1967.

170 Gordin 2004: 807.

tremendous amount of pressure, both in terms of how they understood their discipline and also in terms of the actual numbers they produced and analyzed.171

By the late 1930s, Vestnik Statistiki, the leading statistics journal in the Soviet

Union, had begun to reject submissions that used mathematical approaches in dealing with statistical problems. As a result, over the course of the 1930s and 1940s, large numbers of statisticians withdrew from statistical work, finding a relative safe haven in universities or research institutes where they could continue their statistical research without identifying it as such.172 The practitioners of statistics, in turn, grew increasingly ignorant of the latest statistical methods and their potential applications to statistical work.173 Indeed, by the 1930s, the “most persistent difficulty for these statisticians stemmed from their dual task: they were supposed to provide accurate numbers for the five-year plans, but they also had to suppress real economic and demographic data that

171 For instance, Pavel Il’ich Popov, the first director of the Soviet Statistics Bureau had argued in a letter to Stalin on 22 December 1925 that there was no difference between socialist and bourgeois statistics. He was subsequently removed from his post the following year. Gordin 2004: 809, citing Blum and Mespoulet.

172 Symptomatic is the case of someone like Andrey Kolmogorov (1903-1987), recognized as among the leading probabalists of his time, but who was rarely of interest to statisticians within the USSR. It is indeed possible that a general disciplinary demarcation may have also had the benefit of freeing up the harder sciences, most notably nuclear research, which could then use advanced mathematical statistics without falling prey to ideological battles. Such an analysis was offered by the commentator (identified as J.M.) to an essay on the 1954 Soviet meeting on statistics: “One very practical reason for this distinction [the bifurcation of statistics and mathematical statistics] is the need to legitimize statistical work in technology and the natural sciences, which was questioned by some attitudes of Marxist orthodoxy expressed in the published discussion. Now that such work is labelled [sic] mathematical statistics and thus distinguished from socio-economic statistics, it can be pursued with freedom from doubts induced by high theory.” On Kolmogorov, see: D.G. Kendall 1991. "Andrei Nikolaevich Kolmogorov, 25 April 1903-20 October 1987". Biographical Memoirs of Fellows of the Royal Society 37 (1991): 300–26; on Kolmogorov and Soviet statistics, see: P.C. Mahalanobis. Some Impressions of a Visit to China 19 June – 11 July 1957: 5. PCMMA, Kolkata; and on freeing up certain technical sciences, see: J.M. and K.V. Ostrovitianov 1955. “The Discussion on Statistics Summed Up.” Soviet Studies 6, no. 3 (Jan. 1955): 321-331.. See 321 for quote.

173 Zarkovic 1956.

contradicted the official picture.”174 In Blum and Mespoulet’s formulation, “in the hybrid that was Soviet Statistics, the Soviet triumphed over the Statistics.”175 This dichotomy extended into the 1950s and was clearly visible to outside observers. During his 1957 visit to the PRC, the Indian statistician Prasanta Chandra Mahalanobis (more on this visit in Part III), informed his Chinese hosts that ever since 1947 the Soviet Union had not attended any sample surveys activities organized by the International Statistical Institute, adding that they just did not attach importance to mathematical statistics.176 The separation of mathematical statistics from statistical work, of theory from practical application, was complete.

Confusion over the true definition and status of statistics, however, lingered until it was officially resolved in 1954. It was only through a national conference held in

Moscow from 16 to 26 March that year that what had for the most part already occurred in practice was also given an official and final imprimatur.177 Organized by the Soviet

Academy of Sciences, the Central Statistical Board, and the Ministry of Higher Education, the conference’s purpose was “to discuss matters in dispute and to examine the subject and method of statistical science.”178 In all, 760 invitations were issued across the various

174 Gordin 2004: 809.

175 Gordin 2004: 808.

176 WJB 105-00530-06: 15

177 Ostrovitianov 1955: 321. This article is a translation from the Russian original, which was published in Vestnik Akademii Nauk SSSR, 1954 No.8. K.V. Ostrovitianov, the author of the report, is identified as “the chief organizer of Soviet economists.” In his commentary to the article, J.M. noted that “Ostrovitianov’s main point, broadly stated, is that, at the present time at any rate, the name statistics should be limited to the discipline and practice of which handle the quantitative aspects of socio-economic phenomena; that this is the only statistical discipline deserving the status of science in its own right; and that statistical work on natural and technological phenomena is most conveniently regarded as a branch of mathematics, namely mathematical statistics.”

178 Ostrovitianov 1955: 323.

Soviet Republics.179 The majority of those invited were academic and practicing statisticians, and they were joined by smaller numbers of economists, engineers, mathematicians, philosophers and medical men. In all, sixty people spoke at the conference and twenty sent in written contributions. Three principal points of view emerged during the conference. The first held that statistics was a universal science that studies phenomena of society and nature. The second posited that statistics was a social science concerned with method, a system of principles for collecting quantitative information in order to characterize social phenomenon. The final view was that statistics was a social science that only studies social productive relations, that is, the economy.180

Over the conference’s ten days, the first two views were discussed and eventually summarily rejected. Proponents of the first claim, which was no different from views in the “bourgeois liberal” West, were accused of a “gross error in attempting, with the help of a single science of statistics, to study by the same methods the phenomena both of nature and of social science.”181 This error was not just a scientific one but also had moral dimensions, for, as Ostrovitianov declared in his report on the conference:

The comrades who regard statistics as a universal science which studies nature and society make of it some kind of science over and above the [socio-economic] classes, coldly indifferent to good and to evil, without any preference at all as between [socio-economic] classes and between social structures.182

179 Invitations went out to (name of modern republic in parenthesis): Leningrad (Russia), Kiev (Ukraine), Minsk (Belarus), Baku (Azerbaijian), Tashkent (Uzbekistan), Tbilisi (Georgia), Riga (Latvia), Erivan (Armenia), Alma-Ata (Kazakhistan), Kharkov (Ukraine), Sverdlovsk (Ukraine), and Irkutsk (Russia).

180 Ostrovitianov 1955: 323-324.

181 Ostrovitianov 1955: 324.

182 Ostrovitianov 1955: 326 .

In support of this claim, Ostrovitianov offered a criticism of the false comparison made by other participants who drew upon similarities between the luminosity of stars and the grouping (categorization) of the peasantry.

What can there be in common between the grouping of stars according to their luminosity and, let us say, Lenin’s grouping of the peasantry according to their [socio-economic] class characteristics? Nothing, other than certain technical statistical devices used in the study of these totally different kinds of phenomena. The grouping of stars according to luminosity requires no class analysis at all, whereas in investigating the differentiation of the peasantry we must make a [socio-economic] class analysis of this phenomenon, proceeding from the postulates of historical materialism and political economy. Statistical devices and methods of investigation, the choice of characteristics for grouping, the combining of groups, were entirely without exception subordinated by Lenin to the task of [socio-economic] class analysis of the rural population.183 [Emphasis added]

Ostrovitianov conceded that quantitative relations in their pure form—equally applicable to “inorganic nature, to organic nature and to the domain of the social sciences”—could be studied by mathematics, more specifically by the subdomain of mathematical statistics.

But apart from this branch of mathematical statistics, and the socio-economic science of statistics, there did not exist a third “non-mathematical but nevertheless universal

‘general’ statistics.”184 This distinction formed the basis of the rebuttal of the second claim, that statistics was a science of method. “Every social science,” Ostrovitianov explained, “has not only its own subject, but also its own method of investigation.”

Furthermore, he noted that in political economy, the method employed was abstraction

(historical materialism in the sphere of economic relations), and not, quoting Marx, a

183 Ostrovitianov 1955: 326.

184 Ostrovitianov 1955: 326. “Thus, there is no scientific grounds for a universal statistics as an independent discipline which studies both nature and society. It would be an extremely meagre science made up of some statistical devices which are used in all branches of knowledge, and would hang somewhere between socio- economic and mathematical statistics.”

microscope or chemical reagents.185 Thus, it was decided that only in the field of the study of social phenomena did statistics reach the level of an independent science, with its own subject and methods of investigation.186

And so, following almost two weeks of deliberation and debate, the conference was in a position to offer a definition of statistics.

“Statistics is an independent social science. It studies the quantitative aspects of mass social phenomena in full awareness of and therefore within their qualitative aspect; it investigates the quantitative expression of the laws of social development in the specific conditions of place and time.”187

Statistics was thus to serve, in the words of Lenin, as the “rock of exact and indisputable facts,” the very foundation on which rested not just political economy but all other social sciences as well. The effect, in the end, was to ratify and reify a distinction that was already fairly well entrenched: that between statistics and mathematical statistics.188

Chance Tamed; Chance Rejected

Trygvie Lie’s desire for “clear and systematically organized facts” thus found an echo in

Lenin’s evocation of statistics as the “rock of exact and indisputable facts.” But while such a desire united the universalist claim and its Marxist critique, the contrasting understandings of statistics within each positioned them at loggerheads when it came to

185 It is indeed not clear how abstraction, a verb, can be equated to nouns such as microscope and chemical reagents in ascribing to them all the category of method.

186 Ostrovitianov 1955: 327.

187 Ostrovitianov 1955: 328.

188 It is worth noting that such a distinction might also have had the benefit of freeing up the harder sciences, most notably research related to nuclear and missile technology, which could then freely use advanced mathematical statistics without falling prey to ideological imperatives. See the introduction to Ostrovitianov’s report by J.M.: Ostrovitianov 1955: 321. On a related discussion concerning the status of other mathematical fields such as game theory (whose applications were also proscribed) and cybernetics (which flourished), see Erickson et al 2013: 17-20.

questions of theory and method. The contradictory claims regarding statistics’ universality outlined above were undergirded by two fundamentally different ontological positions on how social reality could be ascertained via quantification. The first, as we have seen, by embracing uncertainty, accepted its own limits. Chance and error were seen to be part of the natural order of things. Not everything could be counted, at least not accurately, and even more to the point, all attempts at counting generated errors. The goalposts accordingly shifted: Intellectual energy was no longer channeled towards devising methods to count exhaustively, but instead, to devising methods for studying variation, and eventually toward how to accurately determine, control, and minimize the error that resulted during any count. Probability, the law of large numbers, and general mathematical reasoning, thus, became crucial tools in devising new methods of counting and assessment.

Table 1 Statistics Socialist Statistics

Statistics is a Universal Science Statistics is a Social Science

Chance exists, randomness part of both No chance in the social world (only laws that natural and social order of things can be ascertained)

All counting generates errors; Can count everything accurately; can devise methods to ascertain and control the best count is the complete count error of any count

Key Methods: Key Methods:

Law of Large Numbers Exhaustive Enumeration

Probability Theory Typical Sampling

Variance, Correlation, Regression

Random Sampling

This was in stark contrast to the second approach, which, fueled by the teleological inevitability of Marxist historical progression and by the confidence offered by dialectical materialism, was predicated on the absence of chance and uncertainty in the social world. As long as particular attention was paid to time, place, and stage of historical evolution, not only could everything be accurately counted, but it must be counted. Within such an ontological world-view there was no place for probability, the law of large numbers, or any kind of abstract mathematical thought. The social world had no chaos, only laws of production, population, and relations, all of which could be discovered via measurement. The result was an emphasis on complete enumeration—the counting of everything. In other words, just as the liberal tradition pivoted on the appreciation and eventual acceptance of uncertainty, chance, probability, and error in the world and in counting, so did the Socialist-Marxist critique base itself on the comfort of certainty, and the desire to enumerate and account for everything.

In taming and rejecting chance, both traditions were interested in increasing overall control. But they had very different expectations about what they could control.

And each approach taken to its extreme could also lead to perversions in social outcomes.

Within neoliberal economics, this emphasis on error, accuracy, and calculability would lead to a stark distinction between uncertainty and risk. Unlike the former, the latter was calculable, and therefore, could be used as the basis for decision-making. For this very reason, uncertainty came to be discounted in the design of economic models, which over time contributed to an increasing disengagement with real-world actions.189 The most

189 The distinction between uncertainty and risk in the sphere of economic activity was articulated by Keynes as early as the 1920s. For how uncertainty, at least in the American Economics Academy, has largely been written out in favor of risk, which is calculable, see: Peter Katzenstein and Stephen Nelson. “Worlds in Collision: Uncertainty and Risk in Hard Times.” In Politics in the New Hard Times: The Great

recent instance of this is the 2008 financial crisis, with its basis in increasingly sophisticated mathematical models that were supremely confident in their ability to ascertain and account for risk.190

As we shall see in Part II, socialist statistics’ fetishization of complete enumeration would also lead to certain systemic cul-de-sacs. But the illusion of control this method offered was extremely attractive. As A. Yugow noted in his review of Soviet statistics in 1947:

Accounting of all economic phenomena on a national scale and the possibility of statistically following all economic phenomena in their complete economic cycle give Soviet statistics such great advantages in comparison with the statistics of other countries, that the forms, methods, and the results of the work of the statistical organs of the U.S.S.R. are highly instructive to all who work in the field of the theory and practice of economic statistics, even in cases when these data are not entirely exact and commensurable in particulars. The gathering of primary materials, the formulation of problems, and the effect of the plan upon economic factors in the U.S.S.R. provide an almost “laboratory” picture; in any case, they reflect the existing facts far more clearly and fully than in other countries, where the statistical study of national problems and those of individual branches is not only greatly complicated, in view of the technical incompleteness of the data, but is also deliberately distorted because of private interests, trade secrets, competition, etc.191

From a planning perspective, the promise of achieving data collection on a national scale, and thereby (in theory at least) ensuring that all manner of economic activity was accounted for was hard to resist. It was also the kind of comprehensiveness that non-

Recession in Comparative Perspective, edited by Miles Kahler and David Lake, 233-52. Ithaca: Cornell University Press, 2013.

190 For a fascinating study of the role of risk and uncertainty management in precipitating two earlier financial crashes (the stock market crash of 1987 and the market turmoil that engulfed the hedge fund Long-Term Capital Management in 1998), see: Donald A. MacKenzie. An Engine, Not a Camera: How Financial Models Shape Markets. Cambridge: MIT Press, 2006.

191 Yugow 1947: 244.

socialist countries with their decentralized statistical apparatus found impossible to achieve.

In many parts of the world, and especially in much of the future global south, these strengths of the Soviet system were seen as the basis for its economic performance.

For instance, the Indian industrialists who drafted the Bombay Plan in 1944 wrote with admiration and aspiration:

This [growth target for India] might appear to be too modest a goal for a planned economy to achieve, especially in view of the fact that in the U.S.S.R., within a period of 12 years since the beginning of the first Five- Year Plan, the national income is reported to have increased from 25 billion rubles to 125 billion rubles, i.e., five-fold.192

Even so, in India, such high regard for planned economic growth did not extend to an adoption of Soviet socialist statistics. In China, however, adoption of Soviet socialist statistics in the years after 1949 was enthusiastic and comprehensive. Even as late as

1957, by which time tensions between China and the Soviet Union had become increasingly evident, Xue Muqiao, the director of the State Statistics Bureau noted in an editorial:

[We must] seriously study the Soviet Union’s advanced experience in constructing and managing a socialist economy, especially its experience in the planned management of the people’s economy and in the establishment of planning and statistical work.193

192 Memorandum Outlining a Plan of Economic Development for India. Penguin Edition, 1944, 29. This was given as the reason for hoping for a 200% jump in national income in 15 years. Also quoted in: Seymour E. Harris 1947. “Appraisals of Russian Economic Statistics—Introduction.” The Review of Economics and Statistics 29, no. 4 (November 1947): 213. This is not to say that Indian statisticians and economists accepted Soviet statistical theory, but that they were no doubt impressed with what the Soviet Union had achieved in the three decades since its inception. See Part III for more on Indian statistics during this period.

193 Xue Muqiao. “我国的计划工作和统计工作必须向苏联学习 Woguo de jihua gongzuo he tongji gonzuo bixu xiang sulian xuexi (Our Nation’s Planning and Statistics Work must learn from the Soviet Union).” RMRB, November 12 1957, 7.

In order to achieve this, Xue and the SSB relied heavily on Soviet statistical experts and the expertise and experience to which they laid claim. It was upon their advice that the

PRC’s comprehensive statistical system was envisioned and its construction attempted.

Soviet Experts in China—the General Story

In his magisterial 1987 study of the PRC’s economic development, Carl Riskin observed that Mao’s phrase of “leaning to one side” anticipated, in a sense, the long-term unsustainability of Sino-Soviet cooperation. This did not, Riskin clarified, mean that

Soviet contributions to Chinese development were insignificant.194 Rather, it serves to caution us against assuming that the PRC blindly emulated the Soviet model. As much recent scholarship (discussed below) has shown, the real story is far more complex.

The general facts of Sino-Soviet cooperation during the 1950s are now fairly well known.195 In February 1950, the two nations signed the thirty-year Sino-Soviet Treaty of

Friendship, Alliance, and Mutual Aid.196 In the event, Soviet financial aid was rather limited. For instance, of the total investment in the first Five Year Plan (1953-57), the

Soviet share was as little as 3%.197 But the aid took non-financial, non-materiel forms as well. From August 1952 to May 1953 Zhou Enlai led a Chinese delegation to the Soviet

Union with the express purpose of drawing up the country’s first Five Year Plan with the

194 Carl Riskin. China’s Political Economy—The Quest for Development since 1949. Oxford: Oxford University Press, 1987, 60.

195 Soviet aid as sharing of technical knowledge was discussed fairly early by Barry Naughton. In more recent years, this aid has been understood from a variety of perspectives and sources. See discussion below

196 For an interesting analysis of the negotiations involving the signing of the treaty, during what was Mao’s first foreign tour, see Yang Kuisong 杨奎松. “The Sino-Soviet Alliance and Nationalism: A Contradiction.” Parallel History Project on NATO and the Warsaw Pact. The Cold War History of Sino-Soviet Relations, June 2005. (Originally published in Social Sciences in China (English Edition), Spring 2005.)

197 Meisner 1999: 112-113. Also see: Riskin 1987: 74-77.

help of Soviet experts.198 Indeed, equally significant, if not more so, was aid that took the form of imparting technical expertise and experience.199

Until recently, this aid had remained a relatively understudied subject and it is only in the last fifteen years or so that it has begun to receive significant scholarly attention. The opening of archives, both in the former Soviet republics as well as in China, has allowed scholars to explore this relationship in much greater detail, focusing on the men, materiel, and technology that formed its basis.200 These studies generally conclude that from 1949 to 1956 China emulated, if not always uncritically, the Soviet experience.

Chinese conviction in following the Soviet Union’s ‘advanced experience’ 先進經驗 began to weaken in 1956, and by 1960, with the Sino-Soviet split, stood entirely repudiated.201

Soviet advisors and specialists in China have also received focused attention in recent years. Deborah Kaple was among the first scholars to address this subject by cobbling together scattered and frequently fragmentary Soviet archival materials and combining them with oral interviews of several men and women who had served as

198 Bernstein and Li 2010: 12. 156 plants were also delivered as part of Soviet aid during the plan years.

199 Meisner 1999: 113. Meisner notes that as many as 12,000 Soviet and East European experts were sent to China in the 1950s (a more recent estimate is discussed below). In return some 6,000 students trained in modern science and technology in Soviet universities and 7,000 workers gained experience in factories. Another estimate is provided by Izabella Goikhman, who suggests as many as 20,000 Chinese students, technicians and scholars studied in or visited the Soviet Union in the 1950s. See Izabella Goikhman. “Soviet-Chinese Academic Interactions in the 1950s.”In Bernstein and Li 2010: 275-302. Shen Zhihua estimates an even higher number: 38,000. See Shen 2003: 405.

200 Among the first in this new wave to explore these questions was the edited volume Westad 1998. The latest and most comprehensive work that covers not just the 1950s but brings the story up to the present is the Bernstein and Li edited volume published in 2010. The Sino-Soviet split has also received a fresh treatment in Lorenz Luthi’s excellent multi-archival work; see: Lorenz M. Lüthi. The Sino-Soviet Split: Cold War in the Communist World. Princeton: Princeton University Press, 2008.

201 Bernstein and Li 2010: 3.

specialists in China.202 Some of her findings were that the specialists were predominantly men, that they stayed from as little as a few months to a year, and that they were not well supervised once they arrived in China. Most of the specialists had a hard time establishing personal friendships with locals. Conversely, the Chinese frequently complained about the quality of the specialists being sent to them. Building and substantially expanding on Kaple’s work, in 2003 Shen Zhihua published what remains the most detailed study of Soviet specialists in China.203

According to Shen, the twelve years from 1948 to 1960 can be divided into four distinct periods.204 The first, from 1948-49 was marked by a Soviet policy of wait and watch. The years from 1949 to 1953 witnessed the beginning of exchanges, but one where the experts were found predominantly either in military affairs or in specific industrial sectors. Agreeing with this assessment of a limited engagement during this phase, Izabella Goikhman has observed that until 1953 interactions were largely characterized by the exchange of publications. Even so, according to Shen’s findings, about 5,000 Soviet experts, including those working with the armed forces, were in China during this period. It is only after 1953 that academic exchanges were institutionalized and visiting personnel increased in number.205 The late 1952 slogan “completely study the Soviet Union”全面學習蘇聯 marked the beginning of this period for Shen. The four years starting from 1954 were the high tide 高潮 of Soviet experts in China. As many as

202 Deborah A. Kaple. “Soviet Advisors in China in the 1950s.” In Westad 1998: 117-140.

203 Shen 2003. Shen made extensive use of archives in both the Soviet Union and the People’s Republic and was able to augment his findings with several oral interviews.

204 For a general summary of this periodization, see Shen 2003: 405-410.

205 Goikhman 2010: 282.

11,000 visited and provided their technical expertise and experience. For the visiting experts this was also a period when the work environment was at its best, when Chinese eagerness to learn was at its most earnest. This was also, however, the period when finding China’s own path began to be openly discussed in certain areas. By 1956, the first doubts about the Soviet model had begun to be articulated. Even so, in the field of science and technology, studying the Soviet Union’s methods was still the pre-eminent strategy.206 In the final phase, 1958-60, Shen finds a gradual decline and then the sudden withdrawal of all experts in 1960. He calculates 2,000 Soviet experts were in China during these final years, leaving us with a total of 18,000 Soviet experts over the course of the 12 year period.

Soviet Statistical Experts

In spite of the varied and contingent nature of Chinese engagement with Soviet aid and expertise, statistics was one area in which Soviet impact was seminal and long-lasting.

While it is difficult to assess how many out of Shen Zhihua’s estimated 18,000 Soviet experts were statisticians, we can identify several key figures who were assigned to important positions as consultants at the SSB or at major educational and cadre training institutions such as People’s University 人民大学.207 The first mention of Soviet statistical expertise is found in Wang Sihua’s speech of 1950. In it he observed that the

Soviet Central Statistics Bureau was personally overseen by Molotov 莫洛托夫, a fact

206 Shen 2003: 18.

207 Such information can likely be obtained from materials in the Archives of the State Statistics Bureau, but I was unable to gain access to the SSB Archives.

which indicated the importance that the Soviets attached to statistics.208 He was therefore appreciative of the help offered by Soviet Statistical advisors as the Northeast Statistics

Bureau (NSB) went about organizing and carrying out statistical work.209 Some years later, Yang Jianbai 杨坚白, Wang’s deputy at the NSB, reiterated the importance of

Soviet expertise for the NSB’s statistical work. In an article in the People’s Daily he reaffirmed the need to take the Soviet Union as their teacher 以俄为师, adding: “The

Soviet Union’s advanced statistical experience is abundant and completely correct, and we must study well her experience.”210

The Soviet statistical advisors who assisted the NSB (and later the SSB) operated solidly within a tradition of socialist statistics. Their own experience in the Soviet Union had primarily focused on statistical work and not research in any other discipline. And they brought their specific understanding of statistics to the PRC. They belonged to a group that would have whole-heartedly agreed with the conclusions drawn at the 1954 meeting in Moscow. As Xue Muqiao 薛暮桥, the first director of the State Statistics

Bureau, would recall in his memoirs, the Soviet experts were decidedly un-enthusiastic about the use of sampling of any kind. In his words, “Soviet statistical experts did not

208 Wang Sihua 1952: 8. This is very likely a reference to Vyacheslav Mikhailovich Molotov в й ов о отов (1890-1986), of the 1939 Molotov-Ribbentrop Pact of non-aggression between the USSR and the Third Reich. A major politician, politburo member, and diplomat, Molotov had served till 1949 as the Soviet Union's Minister for Foreign Affairs. Oddly enough, the paragraph that mentions Molotov was excised from the version of the speech that was reproduced in the 1986 compilation of Wang’s writings on statistics. Cf. Wang 1986: 3.

209 Wang Sihua 1953: 8 and Wang Sihua 1986: 45. For some examples of how Soviet advice and experience was put into practice in the Northeast, see: Wang 1953: 8 and Wang Sihua 1986: 45-46.

210 Yang 1953: 3. Yang was the Secretary-General and Deputy Director of the Northeast Statistics Bureau from 1951 to 1954, when he was transferred to the main offices of the State Statistics Bureau in Beijing. At the SSB, Yang specialized in National Income Accounting. In 1956 he was deputized/transferred to the State Planning Commission.

promote the use of surveys, [but rather] advocated making comprehensive arrangements for statistical reports.”211 At one point, he and Li Fuchun had attempted to have V.N.

Starovskiy’s book Sample Surveys 《抽样调查》issued for study by the Planning

Commission and the SSB, but they were blocked by the Soviet statistical experts present in Beijing.212 Thus, from the very beginning, the advice the Chinese received, and which they valued highly, was about the primacy of socialist statistics, and the concomitant irrelevance of abstract mathematics.

The first among a series of high-level Soviet statistical experts deputized by the

Soviet Central Statistical Administration to serve in the PRC was A.I. Ezhov (A. E. Ежов;

叶若夫).213 Ezhov arrived in the PRC in early 1950 and returned to the Soviet Union in

February 1952. In the Soviet Union, Ezhov was a Deputy Director of the Central

Statistical Administration and as such brought with him the prestige of his high office.214

As the first Soviet statistical expert, Ezhov quite possibly had the greatest impact in

211 Xue 2006: 174.

212 斯塔洛夫斯基 о ов т о в й Vladimir Nikanovich Starovskiy (1905—1975), was the head the Soviet Central Statistical Administration from 1940 to 1975. Xue claimed in his memoirs, that Starovskiy had personally presented him with a copy of this book during his 1954 visit to Moscow. See: Xue 2006: 174.

213 A few different versions of the name in Chinese exist: in the DSJ he is identified as 叶诺夫 [yenuofu], in a Beijing Statistics Bureau document located in the Beijing Municipal Archives he is called 叶若夫 [yeruofu]; and in an NSB publication he is listed as 叶饶夫. The last transliteration is the closest approximation of his actual name. Cf. DSJ: 2; BMA 002-020-00969: 6; and Tongji gongzuo—disan xuanji 统计工作—第三选辑: table of contents. On the general problems with transliterations of Russian names during the 1950s, see fn 217 below.

214 Ezhov also played a key role in disseminating information about Soviet statistics to a wider audience, both in Russian and in other languages. Among his works that were translated into English are: (1) Ezhov 1957; (2) A. Ezhov. Industrial Statisitcs [Ru.: STATISTIKA PROMYSHLENNOSTI]. Calcutta: Statistical Publishing House, 1958; and (3) Ezhov 1967.

shaping statistical work in the PRC.215 During his two years in the PRC, he consulted with the Statistical Office of the National Finance Committee 中财委统计总处, played a critical role in helping the NSB design statistical work, and was instrumental in offering advice and general guidelines for the establishment of a national statistical system.216

Ezhov was followed at the SSB by four statistical experts in succession (See table

2 below). Sergei Krotevich (S. К от в ; 克拉特维奇), who spent a little over three years in the PRC, was the first among them. In addition to acting as a consultant on financial economics, he also played a key role in designing and carrying out the 1953 national census. An expert by the name of Semodzhefgin (Ru: о ж фг (?); Ch:谢木杰夫金) arrived on 22 July 1955.217 A specialist in infrastructure statistics 基建统计, he stayed in the PRC for about three years. The next expert to arrive was Maximov (Ru: ов; Ch:

马克西莫夫), who stayed until1957. The last expert [that I could locate] arrived on 11

December 1957 but spent less than a year in the PRC, returning to the Soviet Union on 11

June of the following year.

215 In addition to appearing with regularity in published and archival materials, Ezhov was also the Soviet expert whose name was recalled most frequently by several of my interviewees. The possibility remains that this is a result of the fact that his name was relatively easier to pronounce, and therefore, to remember.

216 On the latter, see the discussion in Part II, Chapter I.

217 The transliteration of Russian names during this period appears not to have followed a consistent method of concordance between the Cyrillic alphabet and Chinese characters. As a result, I have been unable to ascertain some of the names, for which the source was only the transliterated Chinese name. Semodzhefgin (谢木杰夫金 in the Chinese) is one such case. Another is Suparin (素包里 in the Chinese). In both cases, handbooks of Russian-Chinese names, which I consulted to arrive at guesses of the names in Cyrillic, have proven less instructive than hoped. The input of Russian and Chinese speakers has also, thus far, not offered any resolutions. Cf. Xin Hua 辛华, ed. Eyu xingming yiming shouce 俄语姓名译名手册 (Handbook of Translations of Russian Names). Beijing: Shangwu yinshuguan 商务印书馆 (Commercial Press), 1982. See in particular Appendix 5 which is a Russian-Chinese Transliteration Table 俄汉译音表.

218 Table 2: Soviet Statistical Experts at the SSB Name in Dates in the PRC Chinese Name in Pinyin Cyrillic Name in English Early 1950 - 叶诺夫219 Ye nuo fu A. E. Ежов A. I. Ezhov February 1952 February 1952 - 克拉特维奇 Ke la te wei qi S. К от в Sergei Krotevich July 1955 July 22 1954 - (for 谢木杰夫金 Xie mu jie fu jin о ж фг (?) Semodzhefgin (?) about 3 years) August 22 1955 - 马克西莫夫 Ma ke xi mu fu ов Maximov (for about 2 years) December 11 1957 索包里 Su bao li уп (?) Suparin (?) - June 11 1958

Broadly speaking, Soviet statistical experts such as these five were directly or indirectly involved in three types of activities: consulting on the design and execution of statistical work, aiding the publication of translations from Russian into Chinese, and directly teaching statistical cadres.

Consultation

The most basic form of consultation involved introducing elements of the Soviet statistical system and offering advice to Chinese statisticians on how to proceed with setting up their own statistical system. Thus, for example, in late 1950 Ezhov introduced the structure and key tasks of the Soviet system to statistical cadres in Beijing.220 Such discussions usually involved a wide-ranging array of questions and topics, including how to distinguish between agriculture and industry 工业与农业怎样划分, whether handicrafts industries should be regarded as private industries, how statistical survey

218 Table compiled from: DSJ: 2, 6, 30, 39, 55-56.

219 Elsewhere listed as 叶若夫 Yeruofu. Cf. BMA 002-020-00969: 6

220 See Ezhov’s December 1950 speech at a meeting on statistical work in Beijing. BMA 002-020-00969: 5-6.

work should be organized, how analysis of data should be conducted, the importance of original records 原始记录, and why statistics was a science of party spirit 党性的科

学.221

Soviet advisors were also direct participants in specific survey projects that were initiated by the State Statistics Bureau or served as advisors at particular enterprises. For instance, Sergei Krotevich offered key advice as the SSB carried out the first complete modern census in China’s history in 1953.222 Data was collected across only four parameters: name, age, sex, and nationality.223 Two and a half million central and local government workers took part in the actual recording, and they were in turn aided by volunteer census workers in the localities.224 Final enumerations and tabulations were compiled by June of 1954, and on 1 November 1954, the SSB reported that on the

221 BMA 004-011-00019: 2-12, 13-18. These materials were based on notes from a lecture delivered by Krotevich; the accompanying handwritten cover letter praised his exegesis and requested that the materials be widely disseminated. See p. 1 for the cover letter.

222 The Qing government had attempted a complete enumeration in 1908, but that project had floundered in the face of political exigencies and the eventual demise of the dynasty. The four decades leading up to the 50s had witnessed some surveys conducted by the Guomindang regime in the regions they controlled, but nothing on a national scale. During the decade of the 1940s, scholars in Republican China had urged for the administration of a census and the establishment of a census bureau. Foreign scholars too had noted the need for a standardized census. But the civil war between the Guomindang and the Communists had blocked the implementation of any of these recommendations. Since the establishment of the People’s Republic of China, the Communists had made several attempts at deriving a new population total for the country by collecting reports on the population through administrative channels. These efforts had resulted in population totals of around 566 million. On the 1908 census, see Lam 2012; on foreign view of pre-1949 demographic activity, see: A.J. Jaffe. “A Review of the Censuses and Demographic Statistics of China.” Population Studies 1, no. 3 (Feb. 1948): 308-337. (Jaffe noted that “at no time did the Government appear to realize the value of accurate population data as such, and the data actually collected have been very largely the by-products of inquires with more immediate administrative aims.”); on early PRC administrative attempts at population reckoning, see: Chung Lin and Hsia Lu. “How We Made the Census. China Reconstructs, no. 4 (December 1955): 9-11.

223 Chung Lin and Hsia Lu 1955: 9; on the question of nationality see: Mullaney 2011.

224 Chung Lin and Hsiao Lu 1955: 10.

midnight of 30 June 1953, the population on the Chinese mainland was 583.6 million.225

This number surprised many observers used to the League of Nations 1933 estimate of around 450 million for Republican China.226 Krotevich himeself wrote a report on the census in Russian published in Vestnik Statistiki in 1955. A Chinese language translation was published the following year by the Statistical Press.227

In similar fashion, Soviet consultants also spent time at designated factories and enterprises, helping them set up or rationalize statistical work. For instance, in an essay in the journal Statistical Work the Anshan Iron and Steel Company’s Planning and Statistics

Office thanked the Soviet statistical expert 切布尔尼赫 [Qie bu er ni he; бу ;

Cheburnikh?] for his wide-ranging assistance. The essay explained that his with his help they were able to appreciate what was distinctive about socialist statistics; improve original records 原始记录; simplify forms; establish monthly, ten-day, and daily reports; and initiate research and analysis.228

225 China State Statistical Bureau. “Communique of the Major Statistics from the first national population census as related by the State Statistical Bureau of the People's Republic of China (November 1, 1954).” Chinese Sociology and Anthropology 16, no. 3-4 (Spring-Summer 1984): 62-5.

226 For a sense of the surprise and skepticism these results generated see John S. Aird. “China’s Population, Census and Vital Statistics.” Journal of the American Statistical Association 56 (1961): 443; John S. Aird. “Estimating China’s Population.” The Annals of the American Academy of Political and Social Science 369 (1967): 67-72; John S. Aird. “Sources of Demographic Data on Mainland China.” American Sociological Review 24, no. 5 (1959): 623-630; Ansley J. Coale. Rapid Population Change in China 1952-1982. Washington, D.C.: National Academy Press, 1984; and G.B. Cressey. “The 1953 Census of China.” The Far Eastern Quarterly 14, no. 3 (1955): 387-388. It is possible that none were aware of the early 1950s estimates of 566 million noted above. Many may also have assumed that the disruptions caused by the civil war may have served as a check on population growth.

227 See: Sergei K. Krotevich. “Vsekitayskaya perepis’ naseleniya 1953 g. (The All-China Population Census of 1953).” Vestnik Statistiki (Journal of Statistics), no. 5 (September-October 1955): 31-50; and Sergei K. Krotevich. Zhongguo 1953 nian quanguo renkou diaocha 中国 1953 年全国人口调查 (The All- China Population Census of 1953). Beijing: Statistical Press, 1956.

228 Statistical Office of the Anshan Iron and Steel Company’s Planning Department 鞍钢计划处统计科 1957. See: “Ganxie sulian zhuanjia dui women de bangzhu 感谢苏联专家对我们的帮助 (Grateful Thanks to the Soviet Experts for their Help).” Tongji gongzuo 统计工作 (Statistical Work), 1957.21: 31-32.

Another important consulting task was the inspection of local statistical work.

From November to December 1955 the Soviet specialists Maximov 马克西莫夫 and Xie

Mu Te Jin 谢木特金 were accompanied by SSB Deputy Director Sun Yefang 孙冶方 on a tour of Hubei, Hunan, Guangdong, Zhejiang, Jiangsu, and Shanghai, with the goal of understanding city- and province-level statistical work. Principal questions included how to strengthen local statistical work, especially under the conditions brought about by the high tide of cooperativization, and how to strengthen county, district, and village/township level statistical work. On 18 January 1956 the team produced a report that contained three key parts: First, it was severely critical of the work done by the various provincial bureaus, noting that no matter how well the local or national bureaus worked, the provincial bureaus did not show enough leadership. Their organization was incomplete and their materials were either lacking in quality or late. Second, it offered suggestions on how to strengthen province, district, and village level statistical work.

And third, it offered advice on how to set up a united and scientific agricultural survey scheme.229

Publications

A second way in which Soviet experts contributed to statistical work in the PRC was by aiding the translation of Soviet textbooks, organizing curricula at universities and training centers, and compiling collections of special articles.230 The earliest compilations of the

229 DSJ: 43.

230 For more on Soviet advisors at major universities such as Renmin, Tsinghua, Peking, and so on, see: (1) Yu Shengqi 俞圣祺. “Sulian zhuanjia bangzhu xia de zhongguo renmin daxue 在苏联专家帮助下的中国人民大学 (The People’s University under the aid of Soviet Experts).” RMRB 1954 November 8, 2;

Northeast Statistics Bureau’s Statistical Work frequently contained a section entitled

“Introducing Soviet Experience”苏联经验介绍, which carried a few translated articles.

For most workers in the Northeast (and in the rest of the country) this was the first opportunity to read about Soviet methods through the writings of Soviet statisticians. For instance, a relatively early critique of bourgeois statistics can be found in an article by the author N. 廖佐夫 (H. озов?; N. Liao zuo fu). 231 Each of the NSB’s compilations from

1950 to 1952 carried articles by Soviet statisticians, including in some cases by experts who were present in the PRC as consultants.232 Subjects covered ranged widely, including the meaning, role, and organization of economic statistics in the Soviet Union; survey methods in the Soviet Union; the Soviet experience in unifying and simplifying tables and reports; and a discussion of statistics in People’s Democracies. At the same time, advice on how to deal with local statistical problems was also offered, and dwelt on the role of city and village statistics, the methods of statistical and economic analysis, and so on. By 1953, when several compilations had been produced, Soviet authors had largely been replaced by Chinese authors and the experience discussed was primarily Chinese experience of statistical work.

(2) Chen Jun 陈俊. “ Sulian zhuanjia zai Beijing daxue 苏联专家在北京大学 (Soviet Experts at Beijing University).” RMRB 1953 September 16, 3; (3) Bai Sheng 柏生. “Sulian zhuanjia zai qinghua daxue 苏联专家在清华大学 (Soviet Experts at Qinghua University).” RMRB 1953 November 5, 3; (4) Ge Jiali 葛家理. “Zai sulian zhuanjia de jiaodao xia 在苏联专家的教导下 (Under the Guidance of Soviet Experts).” RMRB 1954 November 6, 2.

231 H. 廖佐夫. “Xin zhongguo tongji de gaikuang yu tongji gongzuozhe de renwu 新中国统计的概况与统计工作者的任务 (Summary of Statistics in New China and the Tasks of Statistical Workers).” Tongji gongzuo—disan xuanji 统计工作—第三选辑. Northeast Statistical Press (January 1952): 7-9, 10.

232 For instance, the 1951.2 compilation had articles by 叶饶夫 (叶若夫; Ezhov?); 巴巴勺夫; 杜伯维亦基; 索亲斯基; and 柴尔林合著; the 1952.1 by 叶饶夫 (叶若夫; Ezhov?); H. 廖佐夫; 沙拉保夫; and 赖绍多夫; and the 1952.10 by 梅尔然诺夫 and 苏什诺夫.

With the establishment of the national State Statistics Bureau in August 1952, much of the translation and editing work began to focus on textbooks.233 By September

1957 over 100 different items had been translated. Their scope was extremely wide: from general outlines of principles of statistics to every kind of specialized statistics such as industry, agriculture, capital construction, trade cooperation, supplies, transportation, finance and economics, health and education, and so on. In addition to advanced texts for specialists, several books were targeted toward the basic statistical worker and were therefore at the middle-school or high-school level. Examples of such books, whose publication runs were upwards of 20,000 or 30,000, included Statistical Theory 《统计理

论》by the Soviet State Statistics Bureau, Petrov’s (彼得洛夫 П т ов] Guide to

Economic Statistics 《经济统计学教程》, and Shaensky’s (沙文斯基 ш ы ) Guide to

Industrial Statistics 《工业统计学教程》. Starting in 1956, the Soviet and the Chinese

Departments of Culture began a collaboration, as a result of which many additional books were received from the Soviet Union. The Chinese were also able to directly approach many authors and request them to write forewords for the Chinese editions of their books.234

Finally, starting in 1956, the Statistical Press also began producing compendia of translated essays and articles. In 1957, it published a book containing transcripts of selected talks by Soviet experts such as Maximov, Semodzhefgin, and others. The book

233 The establishment of the SSB is discussed in Part II Chapter I.

234 See: Anon. “Banian lai woguo fanyi chuban le daliang sulian tongji xuezhezuo—woguo tongji chubanshe kaishi yu sulian tongji xuejie jianli youyi de lianxi 八年来我国翻译出版了大量苏联统计学著作—我国统计出版社开始与苏联统计学界建立友谊的联系 (We Have Translated a Large Number of Works by Soviet Statisticians in the Past Eight Years).” Tongji gongzuo 统计工作 (1957.20): 33.

touched upon various standard topics of statistical work, ranging across the various sectors of the economy and discussing methods of collection and analysis.235 Also begun in 1956 was a series titled Collected Translations in Statistics 《统计译文传记》， which had at least 14 issues (see table 3 below). In its first year, each issue covered a wide variety of topics and was intended for wide circulation. By 1957, each issue had come to focus on a specialized theme, with reduced circulation numbers suggesting a more targeted audience. In its final year, 1958, a couple of issues (numbers 10 and 11) were also devoted to introducing materials from ‘bourgeois’ countries, most notably the

United States and the United Kingdom.236

Table 3: Collected Translations in Statistics 《统计译文专辑》 Beijing: Statistical Press

Series No. Theme Publication Print run Essays 1 General 1956 May 7100 13 2 General 1956 June 7150 9 3 General 1956 September 7140 14 4 General 1956 December 5060 8 5 ? ? ? ? 6 Agricultural Labor Productivity 1957 May 2070 15 7 Bourgeois Statistics 1957 July 1570 8 8 Accumulation and Consumption 1957 July 1570 7 9 Sample Surveying 1957 August 1460 10 History of the Development and Use of 10 Sampling Theory 1958 February 1260 4 The Use of Sample Surveying in Population 11 Research 1958 April 1470 6 12 Theory of Index Numbers 1957 December 1460 6 13 ? ? ? ? Bulletin of Meeting of All Soviet Statistical 237 14 Workers 1958 March 1480 13

235 See: Zhu Hong’en 朱鸿恩. “Du ‘sulian tongji zhuanjia tanhua jilu xuanbian’ 读‘苏联统计专家谈话记录选编’ (Reading ‘Selected Lectures of Soviet Statistical Experts).” Tongji gongzuo 统计工作 (1957. 21): 32-33.

236 Essays in Issue 10 were drawn from The Journal of the Royal Statistical Society and essays in Issue 11 were drawn from The Journal of the American Statistical Association.

237 The meeting that took place in Moscow from June 4 - 8 1957.

Teaching

A final method was that of direct teaching via lecture and discussion. During oral history interviews, the statisticians Ni Jiaxun 倪加勋 and Lin Fude 林富德 recalled attending lectures organized at People’s University and delivered by Soviet experts, in particular

Ezhov.238 Both Ni and Lin remarked that issues of translation and access made these experiences rather episodic and of limited value: The Soviet expert would lecture via an interpreter. Attendees were not permitted to ask any questions during the session. If the attendees had questions, they were required to write them down and hand them to the expert’s local liaison 接待员. The liaison would then select and translate a handful of these and deliver them to the expert, who would then prepare responses to them. These responses would then be delivered at the start of the following lecture.239 Such a discontinuous learning process with no classroom interaction arguably had limited pedagogical value for the attendees themselves. But the lectures may have had a wider impact, since they were typed-up and circulated as reference materials for study 学习 sessions.240 Finally, from time to time, larger seminars were also organized by the SSB or

238 Based on interviews with the statisticians Lin Fude 林富德 and Ni Jiaxun 倪加勋, who were both active in statistical work and education in the 1950s. I met Lin on 22 April 2011 in Beijing and had two meetings with Ni, both in Beijing, on 29 March 2011 and 7 September 2011. A statistical worker during the early 1950s, Ni attended Peoples University, where he obtained a degree in 1959 (verify). He subsequently took up a faculty position at the university. It was through him that I learned of the identity of the two statisticians who visited India later in the year. [provide details of interview with Ni Jiaxun and Lin Fude]

239 This description is based, in particular, on Lin Fude’s recollections.

240 Krotevich’s lecture from August 1952 appears to be an instance of such a process; the cover letter to his lecture states that “his report is very good, and can serve as reference material for cadre in agricultural and industrial statistics during their vocational learning.” See BMA 004-011-00019: 1, 2-12. In another instance, Semodzhefgin’s 谢木杰夫金 May 28 1956 lecture at the Xi’an Statistics School was published in Statistical Work. See: Semodzhefgin’s 谢木杰夫金. “He qingnian tongzhi tantan tongji xuexi—sulian tongji zhuanjia xiemujiefujin 1956 nian wu yue 28 ri zai xi’an tongji xuexiao de bagao 和青年同志谈谈统计学习—苏联统计专家谢木杰夫金同志 1956 年 5 月 28 日在西安统计学校的报告.” Tongji gongzuo 统计工作 (1956.14): 13-15. Reproduction of lectures could include the works of more hallowed figures as

one of the universities. For instance, on May 12 1955, the SSB organized a seminar 讲座 in Beijing where Soviet experts and professors from People’s University delivered lectures 讲课 to statistical cadre from all the departments within the State Council 国务

院.241 Such exercises, however, appear not to have been a frequent feature of a statistical expert’s work.

A fourth way of drawing upon Soviet expertise also existed, though its site was not the PRC but the Soviet Union. While it is not clear how many Chinese students studied statistics in the Soviet Union, at least two senior Chinese statisticians visited the

Soviet Union. From December 1954 to January 1955 Xue Muqiao visited the Soviet

Union as part of a Sino-Soviet Science and Technology Cooperation delegation. During his time in Moscow, Xue met with the director and deputy director (the recently returned

Ezhov) of the Soviet Central Statistical Bureau. He also met with members of the

Planning Commission.242 A little over a year later in July of 1956 Sun Yefang 孙冶方, one of the SSB’s deputy directors, led a seven man team of statisticians to the Soviet

Union. They met various officials, visited various bureaus, and studied various statistical and planning activities. Altogether the team spent over 30 days and discussed a variety of subjects including the calculation of total output value 总产值, constant prices 不变价格, and the reporting structure of the Soviet Union.243

well. One of Lenin’s letters to the Soviet Central Statistics Bureau was translated and circulated in June 1953, to be used as reference material in study sessions. See: BMA 123-001-00416: 50-56.

241 DSJ: 37.

242 DSJ: 34; Xue 2006: 174.

243 DSJ: 47.

Conclusion

Li Fuchun’s dismissal in 1951 of statistics as it had been practiced heretofore indicated a crisis on two levels, one ontological and the other epistemological. What categories should the world be divided into? And what forms of knowledge can be deployed to ascertain these categories? In line with the discourse within the Soviet Union, as summarized by Ostrovitianov in 1954, the point of departure was the premise that there was no single universal reality for the known world; rather, different theories and methods were applicable to these different divisions—the social and the natural—of the world. These different theories and methods, in turn, promoted very different interests and led to very different social and economic outcomes in society. For Chinese thinkers, drawing upon Soviet and Marxist views, the natural and the social world were not the same and could not be understood using the same tools. They gained these views not just from reading Marx and Lenin, but especially through Soviet statistical experts who were present in the PRC and through a vast array of translated material. The overall effect was to engender an overarching theoretical approach to all knowledge that prefigured what statistics could be. The practical impact of this theoretical distinction between the natural and social world was to idealize and attempt to construct a vast national network of statistical offices, from Beijing down to the county level, with the aim of capturing all necessary data. In Part II we shall explore the implications of this ideal. For now, let us turn to how socialist statistics gained primacy and what its contents were.

CHAPTER IV

No “Mean” Solution: The Reformulation of Statistical Science

In February 1957, professor of statistics and one-time banker Jin Guobao 金国宝 published a self-criticism in the journal Statistical Work 统计工作.244 The focus of the self-criticism was his successful textbook, “An Outline of Statistics” 《统计学大纲》, the most recent revised edition of which had been published in 1950.245 An abridged version published under the title Statistics 《统计学》in 1935 had also become popular as a textbook in vocational schools. In his opening paragraph to the self-criticism Jin explained that the ancients’ 古人 possessed an important principle: getting rid of hackneyed phrases 陈言务去. By that principle alone, Jin confessed, “An Outline of

Statistics” was beyond redemption. Copied entirely from Anglo-American bourgeois statistics, the book was full of stale words 陈言; not one sentence was the product of new thinking or composed of original ideas. The article was thus an attempt to atone for the failures of the book and for his failure to correct his thinking in a timely fashion.

Such strong self-criticism is not entirely surprising, especially in 1957 when intellectuals became particular targets for thought reform.246 Nor was Jin’s the only self- criticism written by a statistician in the mid-1950s.247 These self-criticisms were

244 Jin Guobao 金国宝. “Dui ‘Tongjixue dagang’ de ziwo pipan 对“统计学大纲”的自我批判 (Self- Criticism on ‘An Outline of Statistics’).” Tongji gongzuo 统计工作, 1957.3: 28-32.

245 Jin Guobao 金国宝. Tongjixue dagang 统计学大纲 (An Outline of Statistics). Shanghai: Commercial Press, 1934.

246 On thought reform of intellectuals during the 1950s more generally, see: Theodore Hsi-en Chen. “The Thought Reform of Intellectuals.” In Government of Communist China, edited by George P. Jan, 477-485. San Francisco: Chandler Publishing Company, 1966. For a recent work that focuses on thought reform campaigns conducted against vagrants and the lumpen proletariat, see: Aminda M. Smith. Though Reform and China’s Dangerous Classes—Reeducation, Resistance, and the People. New York: Rowman and Littlefield Publishers, Inc., 2013.

247 At least two other self-criticisms were published in 1956 and 1957: (i) Zou Yiren 邹依仁. “Guanyu zichanjieji tongji xueshu sixiang de ziwo pipan 关于资产阶级统计学术思想的自我批评 (Self-Criticism Regarding Bourgeois Statistical Thought).” Tongji gongzuo 统计工作 1956.8; and (ii) Wang Sili 王思立.

100

themselves part of an extended series of debates within the academy, which from 1950 onward had sought, on the one hand, to establish the new definition of statistics, and on the other, to mark out which statistical tools were appropriate for a socialist society.248

Jin’s self-criticism, in particular, offers a useful route through which to access two related strands of inquiry. The first of these has to do with understanding consensus formation within the scientific community of early PRC statisticians. As shown in Part I Chapter I, the pre-1949 world of Chinese statistics (and social science) was vibrant and methodologically and epistemologically diverse. In that chapter I made the case for the centrality of the Northeast Statistical Bureau in reshaping the nature and content of statistical work, indicating that changes in the practice of statistical work preceded their articulation within the theory and pedagogy of statistical science. Crucial to the latter transition was disciplining a vibrant and varied community of statisticians so that they came to accept and espouse statistics as a Marxist-Leninist social science in the years after 1949.

The sense that practice preceded changes in pedagogy raises questions about how we conceptualize scientific change during and after political revolution. How did

Socialist Statistics become the dominant paradigm during the 1950s? Recent work in the sociology of science has devoted much attention to studying how practitioners arrive at scientific consensus and closure. One the one hand, these studies stress ‘extra-scientific’

“Pipan wo zai tongjxue zhuzuo zhong de zichan jieji sixiang 批判我在统计学著作中的资产阶级思想 (Criticism of the Presence of Bourgeois Ideology in my Statistical Work).” Tongji gongzuo 统计工作 1957.10.

248 Articles critical of pre-1949 statistics and its practitioners were published in journals such as Tongji gongzuo tongxun 统计工作通许 (Statistical Work Bulletin),Tongji gongzuo 统计工作 (Statistical Work), Xin jianshe 新建设 (New Construction), and Zhongguo gongye 中国工业(China Industry), as well as in scientific theses written at Renmin University in Beijing. See: Zou 1956: 28.

101

factors such as culture, power and funding, politics, and personal credibility.249 Others have focused on the “social processes within core sets of practicing scientists who negotiate results,” demarcate knowledge claims, and construct boundaries.250 Yet others have explored the cultures of knowledge societies, in particular the disunity of science made visible by a focus not on “the construction of knowledge but in the construction of the machineries of knowledge construction.”251 This machinery is further broken down into empirical, technological, and social machinery; the argument being that it is combination of these that makes possible the rise of "truth effects.”252 The Chinese experience with statistics in the1950s reflects many of these forces. But none of these studies seem to account for how drastic political change, which brought with it an overarching new theoretical paradigm, was negotiated—both by the people promoting the paradigm and by those whose original positions were not in agreement with that new paradigm. The first half of this chapter explores this question by focusing on the nearly decade-long sequence of publications in various journals that sought to criticize Anglo-

American ‘bourgeois’ statistics and its practitioners.

If the first half of this chapter deals with the statistics community, the second explores how this transition affected the very content of statistics and statistical education.

In broad terms, the consequences of this transition are well known: as a social science,

249 For a brief overview of this literature, see Uri Shwed and Peter Bearman. “The Temporal Structure of Scientific Consensus Formation.” American Sociological Review 75, no. 6 (2010): 817-840.

250 Shwed and Bearman 2010: 819.

251 Karin Knorr-Certina. Epistemic Cultures—How the Sciences Make Knowledge. Cambridge, London: Harvard University Press, 1999, 3 Knorr-Certina’s work itself builds on earlier work that focused on the work of scientists as actors; most notably the works of Bruno Latour, Peter Galison, and Lorrain Daston.

252 Knorr-Certina 1999: 11-13.

102

statistics dealt with the social world and not the natural. It was, as a result, bifurcated from what was seen as the abstract theorizing of mathematical statistics, which was, in turn, banished to departments of mathematics.253 But the distinctions ran deeper. To merely reject the law of large number 大数法则 and general probability theory 几率论 was far from enough. Even fundamental concepts such as measures of central tendency

(especially the mean), indicators 指数, and time series 时间数列, would come under the microscope, as earnest attempts were made to excise the last vestiges of any polluting bourgeois influence. A related and critical element in this process was the production of new textbooks aimed at both the future professional as well as the common man.

Comparing these textbooks with Jin’s An Outline of Statistics will permit us to see what from the older textbooks was retained and what was rejected. At the same time such a comparison also indicates the extent to which the newest global developments in statistical science were incorporated in 1950s China.

Jin Guobao and the Road to (Self-) Criticism254

In the immediate years after 1949 most Chinese statisticians were under increasing pressure to accord to a new overarching theoretical paradigm, described in the previous two chapters, which stood in fundamental contradistinction to the one they were familiar

253 See: “Chouyang de shuli tongjixue 抽样的数理统计学 (Sampling and Mathematical Statistics).” Caijing yanjiu 财经研究 1958 (03): 57.

254 Jin’s biographical details are primarily drawn from: (i) Shi Tai 施泰. “Jin Guobao 金国宝 (1893-1963).” Tongji yu yuce 统计与预测, 1994.06: 60-61; (ii) “Zhongguo jindai tongji xue dianjiren zhiyi—jin guobao 中国近代统计学奠基人之一—金国宝 (Among the Founders of Modern Statistics in China—Jin Guobao).” In Minguo renwu dacidian 民国人物大辞典 (Biographical Dictionary of Republican China), by Xu Youchun 徐友春. Hebei People’s Press, 1991; and (iii) “Jin guobao 金国宝.” In Shijie tongji mingren zhuanji 世界统计名人传记 (Biography of Famous Statisticians of the World), by Gong Jianyao 龚鉴尧, 402-407. Beijing: Statistical Press, 2000.

103

with. In order to attempt an answer to the question of how this happened, it is important to first disaggregate the community of statisticians.

During oral history interviews with the statistician Lin Fude 林富德 (1925-) and the demographer Wu Cangping 鄔滄萍 (1922-), both provided contrasting views on how they adapted to the new statistics of the 1950s. Both had come of age and been trained in the 1940s, and as such, prior to 1949 had a relatively catholic approach to statistics. In the years that followed, Lin informed me, he and other statisticians had to recalibrate their ideas of statistics and bring them into closer alignment with the new definitions. Wu, on the other hand, while accepting that such recalibration took place, observed that the change was often only rhetorical or symbolic; the actual statistical ideas were the same.

Both pointed out that the tangible result was that methodology and statistical work were privileged over research and theory and application stood estranged from each other.255

While Lin and Wu may have successfully negotiated the transition, as the youngest generation of statisticians trained before 1949 they were hardly in positions of influence; that is, positions from where they could have an immediate and widespread impact on the state of the discipline and on the education and training of future statisticians and statistical cadres. That space was occupied by an older generation of statisticians who had been active over the preceding decades; especially the 1930s, when the social survey movement was at its peak. This generation had published textbooks that were in widespread use. Its members had served in government, in education, in banking, and also as researchers in a private capacity. In particular, it was in their role as authors of important and popular textbooks that curtailing their activities and reformulating their

255 I interviewed Lin and Wu in Beijing; Wu on 11 March 2011 and Lin on 22 April 2011.

104

views became crucial to the disciplinary reformulation of statistics. Prominent among that generation was Jin Guobao.

Today, Jin Guobao is widely regarded as one of the founders 奠基人 of modern statistics in China.256 Such recognition is based to a large extent on his activities prior to

1949. A professor of statistics and one-time banker and senior government official, in the decades before the CCPs’ victory he had become arguably China’s most famous statistician as well as its most prominent teacher of statistics. His text, An Outline of

Statistics 《统计学大纲》, went through multiple editions and as the standard textbook for college level statistics it had become “the most influential statistics textbook in the country during that time.”257 It was precisely this perceived influence that made him one of the targets of concerted criticism in the early PRC.

Jin Guobao was born Oct. 12 1893 in Tongli Town in Wujiang county, Jiangsu province 江苏省吴江市同里镇. His early education was at Shanghai’s Fudan Gongxue

复旦公学. After graduating from school, he served as an English teacher at the Wujiang

Middle School. During this time he frequently contributed essays to the journals New

Bulletin of Current Affairs 《时事新报》and Eastern Miscellany 《东方杂志》. Using the nom de plume Jin Lüqin 金侣琴 he was the first person to translate Lenin into the

Chinese language, publishing Lenin’s “Bolshevik's Statement of Demands” 鲍尔雪维克

256 See, for instance, the entry for Jin in Xu 1991.

257 See: “Jin Guobao 金国宝 (1893-1963). Tongji yu yuce 统计与预测 1994.06: 61. By 1948, the book was in its ninth edition. I haven’t so far been able to locate actual publication details for the 1950s, but a 13th ‘revised’ edition was published in 1961, which suggests three possible editions during the 1950s. The book appears to have been re-published in 1992, this time by the Shanghai Press 上海书店.

105

之所要求与排斥 in the first issue (September 1919) of the journal Liberation and

Reform 《解放与改造》.258

A few years later in 1921 he traveled to America and studied statistics at

Columbia University on a scholarship from the South Seas Tobacco Company 南洋烟草

公司. Jin received a Master’s degree in 1923, upon which he returned to China and successively taught statistics at such institutions as the Shanghai Chinese Public School

上海中国公学, Fudan University, and the University of Commerce 商科大学. During this time, he was also a Section Chief in the Ministry of Finance. In 1928, on the orders of the President of the newly founded Academia Sinica 大学院, Cai Yuanpei 蔡元培

(1868-1940), he traveled to various countries in Europe and the Americas to study their statistical activities. During well over a year of travel he gathered a variety of statistical materials.

Upon his return to China, he became directly involved in government, banking, and statistical education. In May 1929 he was appointed the Director of the Nanjing

Finance Bureau and served as a special member of the Capital Construction Committee.

The following year, he was among the people who joined D.K. Lieu in setting up a

Statistical Society, which was conceived as a forum for the promotion of private academic discussions. In April of 1930 he traveled abroad again, this time to Japan, to study local finance and to attend the International Statistical Institute’s nineteenth

Biennial Conference in Tokyo.259 Upon returning from Japan in July of that year, he moved to Shanghai and reentered the world of finance. Over the next several years he

258 This journal apparently only existed till 1922; Liang Qichao was its onetime editor.

259 Other attendees included D.K. Lieu and Zhu Junyi.

106

held various managerial, auditing, accounting, and research positions at the

Communications Bank 交通银行 and at the Central Bank 中央银行.260 He would eventually rise to senior positions such as Director of Accounting and Chairman of the

Board at several of these institutions. After 1937, Jin also held professorial appointments at Shanghai University of Commerce and at Jinan University.

Of Jin’s various professional identities, the one that would persist into the 1950s following the Communist takeover was that of respected professor and prolific author.

Early in the decade he was appointed a professor at Fudan University, where he also served as the director of special training courses in statistics and trade. During these years he also concurrently held professorships at the Shanghai Institute of Finance and

Economics 上海财政学院 and at the Institute of Economics of the Shanghai Academy of

Social Sciences. Jin was a prolific author and had already produced several monographs and textbooks for the study of statistics and economics prior to 1949. In addition to the aforementioned An Outline of Statistics 《统计学大纲》, he had published tracts on subjects such as England’s tax structure, currency reform, index numbers, industrial and economic problems, and Keynesian economics. He continued writing after 1949, and authored at least four books on statistics, including an abridged version of the Outline of

Statistics, titled Statistics 《统计学》.261 He was 69 years old when he passed away in

Shanghai on 15 February 1963.

260 See Jin’s entry in Xu 1991.

261 The other three were: (i) Gaoji tongjixue 高級統計學 (Advanced Statistics). Shanghai: Lixin kuaiji tushu yongpinshu, 1951; (ii) Gongye tongjixue yuanli 工业统计学原理 (Principles of Industrial Statistics). Shanghai: Lixin kuaiji tushu yongpinshu, 1952. (Republished in 1953); and (iii) Shehui zhuyi zhidu xia de guomin shouru yanjiu 社会主义制度下的国民收入研究 (Analysis of National Income under the Socialist System). Shanghai: Shanghai Finance Institute, 1956.

107

It was Jin’s An Outline of Statistics, which, because of its content and continuing and widespread use after 1949, became one of the early targets of criticism during the

1950s. An Outline of Statistics was published in 1935 by two different presses: firstly as part of a series by the National School of Commerce Press 国立商学院出版社, and secondly by the Commercial Press. By 1950, it had gone through 13 editions.262 Jin’s views on statistics and its links to mathematics, however, were in evidence as early as

1925. In the introduction to the New Discussion/Theory of Statistics 《统计新论》, his first textbook on statistics, published in 1925 shortly after he returned from attending

Columbia University, he wrote:

“The study of statistics is subtle and broad. If one is without mathematical awareness and a scientific mind, he cannot speak about it. And yet, it is used frequently and very widely; there is no one who doesn’t use it. Indeed, scholars who care not to learn it are immediately struck by dizziness when they come across its charts and tables; and the less said about [their reaction to] advanced mathematical principles and refined and meticulous analysis, the better.”263

Ten years later, the first sentence in Jin’s An Outline of Statistics was more direct:

Statistics is the study of using counting or estimation to represent the dynamic-ness and static-ness of social or natural phenomena and to carry out the analysis of the relationship between numbers.264 [emphasis added]

In thinking about statistics as fundamentally mathematical in nature and universally applicable to the social and natural worlds, his views were very much a reflection of the prevailing wisdom of the time. Like Jin, the significant number of Chinese statisticians

262 See: Song Tao 宋涛 ed. 20 shiji zhongguo xueshu dadian: jingjixue—xiace 20 世纪中国学术大点：经济学—下册 (Major Themes in Twentieth Century Chinese Academics: Economics—Volume 2). Fujian Education Press 福建教育出版社, 2005, 785.

263 Quoted in: Shi Ta 1994: 60.

264 Jin Guobao 1934: 1.

108

(and economists and social scientists) active in the 1930s had been trained in the United

States or Western Europe or were students of those that had been trained there.265 They had served in government and were in many instances active participants in what Tong

Lam has called the Social Survey Movement of the 1930s.266 At this time, a well-defined socialist critique of statistics had yet to be articulated in China or elsewhere. The espousal of statistics as a universal science by Jin and his contemporaries, many of whom were authors of prominent textbooks as well, had no coherent counter-discourse. This view was reflected in several textbooks they published shortly after 1949, all of which enjoyed widespread popularity: the Jilin based professor of statistics Wang Sili 王思立 published

The General Theory of Statistics 《统计学通论》in 1951; Zhang Zhihong 张志鸿 published Agricultural Statistics 《农业统计学》in 1953, and Zou Yiren 邹依仁 published Industrial Statistics 《工业统计》in 1952.267 By publishing books that promoted what was to quickly become a theoretically unsound understanding of statistics after 1949, these authors drew the immediate and targeted attention of New China’s new

265 For the role of Anglo-American training and the larger influence of the Anglo-American Academy in shaping disciplines such as Demography, Economics, and Sociology in Republican China, see: Tong Lam 2012; Paul B. Trescott 2007. Jingji Xue—The History of the Introduction of Western Economic Ideas into China, 1850-1950. Hong Kong: The Chinese University Press; and Yung-Chen Chiang 2001. Social Engineering and the Social Sciences in China, 1919-1949. Cambridge: Cambridge University Press. According to Trescott, the Columbia Economics Department, which produced the largest number of Chinese graduates in the social sciences during the first half of the 20th century, was particularly distinguished for its heavy emphasis on statistics. Trescott 2007: 69. For a discussion of Marxist social science during this period see Chiang 2001, Chapter 3 and 4: “Genesis of a Marxist Social Science Enterprise in the Early 1930s,” pp. 136-183. For an alternate view that considers developments in economic thought during the Republican years through the rubric of the totalizing governmentalizing logic of capitalism, see: Malcolm Thompson. “The Birth of Chinese Population: A Study in the History of Governmental Logics.”University of British Columbia, PhD Dissertation, 2013.

266 Lam 2012.

267 Song, Xuguang 宋旭光. “Xin zhongguo jingji tongji fazhan de lishi kaocha 新中国经济统计发展的历史考察 (Historical Investigation of the development of Economic Statistics in New China).” Caijing wenti yanjiu 财经问题研究 (Research on Issues in Finance and Economics), no. 3 (2013): 21-26.

109

socialist statisticians. As one of these socialist statisticians would go on to explain, the hazards such books posed were indeed quite grave:

… [They] will not only impede the progress of statistical science, but will also cause ideological confusion among statistical workers, which will lead statistical work to suffer losses; they therefore [eventually] hinder the development of the enterprise of socialist construction. 268

And so, in very short order after 1949, Jin Guobao, many of his fellow statisticians, and the textbooks they had composed, all came to represent the most obdurate of foes to those seeking to enforce a new understanding of statistics.

Patterns of Critique and Self-Critique

Obdurate foes Jin and his fellow statisticians may well have been, but they were also the type of higher intellectual who the CCP prized tremendously. For the CCP, their value lay in possessing skills and knowledge that were sorely needed. But such groups were also the least susceptible to dogma. As Harriet C. Mills noted in a 1959 article,

Communist policy toward this strategic group [of higher intellectuals] in the past decade has consistently aimed at securing most effective utilization of its knowledge. Zigzagging steadily toward this goal, the Communists have now attacked, now united, now criticized. Meanwhile, they are recruiting their own Red intellectuals, but have not yet had time to train a new group of Red and expert.”269

This then was the challenge: to not wholly discredit the statisticians but to reform them so that they could continue to contribute from within the system and be held up as

268 Gui Shizuo 桂世祚. “Ping Zou Yiren, Jin Guobao liangwei xiansheng de gongye tongjixue zhuzuo 评邹依仁、金国宝两位先生的工业统计学著作 (Criticisms of Zou Yiren and Jin Guobao’s Works on Industrial Statistics).” Tongji gongzuo 统计工作 1956: 24

269 Harriet C. Mills. “Thought Reform: Ideological Remolding in China.” In Government of Communist China, edited by George P. Jan, 486-498, 495-96 San Francisco: Chandler Publishing Company, 1966.

110

exemplars of the potential of remolding.270 A lack of adequate numbers of Red experts in statistics remained a challenge through the decade, meaning that this older generation of statisticians never lost its value.

The criticisms of the 1950s can also be understood as a younger generation’s attempts to supplant an older generation. The people responsible for ‘zigzagging’ toward the goal of effectively utilizing statisticians belonged, for the most part, to a middle generation, between the Jin Guobaos on the one side, and the Lin Fudes and Wu

Canpings on the other. While it is unclear if either Lin or Wu directly participated in the criticisms (Lin co-authored a textbook in the mid-1950s along socialist lines), many of their colleagues did, and they all shared certain general characteristics. Most were born in the 1910s or 1920s, had been trained in economics or statistics (sometimes abroad), and by the late 1930s had thrown in their lot with the CCP. Many of them had received further training in Yan’an and at party schools elsewhere and, as such, were particularly receptive toward socialist statistics as it became increasingly clearly articulated by the late 1930s. As they took on administrative and pedagogical positions at key institutions following 1949—in some cases they had already been involved in administering CCP controlled areas prior to that—they turned their enthusiasm towards propagation. Notable among these institutions were the Northeast Statistics Bureau (e.g. Yang Jianbai 杨坚白,

Wang Sihua 王思华, Wang Jianzhen 王建真, others?), the State Statistics Bureau (Xue

270 On thinking about the Chinese revolution as a redistribution of skills and knowledge, see: Jacob Eyferth. Eating Rice from Bamboo Roots—The Social History of a Community of Handicraft Papermakers in Rural Sichuan, 1920–2000. Cambridge: Harvard University Press, 2008.

111

Muqiao 薛暮桥, Wang Sihua 王思华, Wang Jianzhen 王建真), People’s University (Lin

Fude 林富德), and Beijing University (Hu Daiguang 胡代光).271

Newly established journals of the PRC offered the principal channels through which this younger generation attacked and cajoled Jin Guobao and his colleagues. Of particular prominence were journals that focused specifically on statistics, most notably the Northeast Statistics Bureau’s Statistical Work 《统计工作》 (1949-1954) and the

State Statistics Bureau’s Statistical Work Bulletin 《统计工作通讯》 (1953-1959, 1979-

).272 Other venues where statistics was debated included the journals New Construction

《新建设》, the Journal of Finance and Economics 《财经研究》 (1956-1960, 1980- ).

By the middle of the decade student theses written at Renmin University had also begun to tackle the subject of socialist and bourgeois statistics.273

The pattern of these criticisms and self-criticisms did not therefore follow the larger patterns of national thought reform campaigns, in which newspapers were a key element. Harriet Mills described such national campaigns by noting that they generally began “with a series of articles and editorials in newspapers.”

Since newspaper reading is a political obligation in China and items of the day are often taken up in study groups, a subject which has received more than usual attention will begin to be discussed. Thus, a demand is created for further study, for which materials and instructions are soon

271 The senior statistical leadership, people like Xue Muqiao, Li Fuchun, Sun Yuefang, Wang Sihua, etc., did not themselves participate in the criticism, though they must have sanctioned them. For example, Li Fuchun clarified the PRC’s new statistics in his 1951 speech, distinguishing it from ‘Anglo-American’ and ‘bourgeois’ statistics. See Part I Chapter I for Li’s actual words.

272 The SSB’s Tongji gongzuo 《统计工作》(Statistical Work) was a monthly from 1953 to 1955; became a fortnightly 1956 onward, and ceased publication after 15 issues in 1959. It was revived in the late 1970s, and these days has the name Zhongguo tongji 《中国统计》(China Statistics).

273 Zou 1956: 28.

112

forthcoming. The campaign, which may last several months, is launched.”274

As noted, most criticisms of bourgeois statistics, however, were published not in newspapers but in specialized journals, which were not available to the general public.

The readership and target audience for study sessions, therefore, was more narrowly conceived: the statistical cadre, student, and teacher. In addition, unlike nation-wide campaigns such as the three-antis and the five-antis or the anti-rightist campaign that usually had a clear start and end date, the critique of statistics persisted through the decade, gathering steam along the way.275 The interlocutors in these debates no doubt took some of their cues from national developments, but they were also subject to the internal and international dynamics of the world of statistics.

Based on these general observations, we can divide the years from 1949 to 1958 into four periods (See Table 1 below). 276 The relative paucity of critical articles during the first phase, lasting from 1949 to 1951, suggests that attention was devoted primarily to establishing a basis for statistical work in the PRC. Once this foundation was in place in the Northeast, a sustained analysis of existing statistical theory and pedagogy could be undertaken. Indeed, through the end of 1951, criticisms of ‘Anglo-American’ or

274 Mills 1966: 492.

275 On the slippage between the normative start and end dates of campaigns and policies and their actual practice/implementation, see: Neil J. Diamant “Policy Blending, Fuzzy Chronology, and Local Understandings of National Initiatives in Early 1950s China.” Frontiers of History in China 9, no. 1 (2014): 83-101.

276 The periodization here is reconstructed from the following materials: (1) Liu and Zhang 2000: 119-127; (2) Various articles from State Statistics Bureau’s Tongji gongzuo tongxun 《统计工作通讯》and Tongji gongzuo 《统计工作》; (3) Compilations of the Northeast Statistics Bureau’s serial anthologies of materials culled from the monthly journal Tongji gongzuo 《统计工作》from 1950-1953 (The NESB’s journal was discontinued by the middle of 1954); and (4) various other journals, such as Xin jianshe 《新建设》, Jingji zhoubao 《经济周报》, Caijing yanjiu 《财经研究》, etc. It should be pointed out that this is not a comprehensive listing of all publications attacking ‘bourgeois statistics;’ it does, however, reflect some of the general trends of the decade.

113

‘bourgeois’ statistics were rather limited. On a national level, the first book to be criticized was Gou Shisheng’s 勾适生 Principles of Statistics《统计学原理》. Essays critical of Gou were published in 1950 by He Sheng 荷生 and in 1951 by Tie Hua 铁华 and Liu Xin 刘新.277 These initial criticisms appear not have been very detailed. On a more regional level, some criticisms were published in the Northeast Statistical Bureau’s

Statistical Work《统计工作》 starting in 1951, especially following the start of the three-antis campaigns late in that year.278

Table1. Frequency of Articles Criticizing "Bourgeois" Statistics, 1949-1958 (including self-criticisms) 1949-1951 1952-1954 1955-1957 1958 Number of Articles 2 6 10 1

Developments in late 1951 and during 1952 provided a clearer focus for the critique, which expanded both its targets and its frequency. The first of these developments was the publication of new editions of textbooks by Jin Guobao, Zou Yiren, and other statisticians. If these new editions offered a significant challenge to the rise of socialist statistics, then the preparations for and establishment of the State Statistics

Bureau (1952) and selection of People’s University as the main training site for the

(social science) of statistics (1952) provided the twin loci through which this challenge was met. Both these developments, as well as the nation-wide reorganization of educational institutions 院系调整, encouraged a more sustained analysis of the state of

277 He Sheng 荷生. “Ping “tongjixue yuanli--diyijuan” 评《统计学原理》(第一卷) (Review of Principles of Statistics (Volume I)”).” Jingji zhoubao 经济周报 (Economic Weekly) 12, no. 3 (1950); Tie Hua 铁华 and Liu Xin 刘新. “Lun tongjixue de duixiang ji qi kexue jichu—ping Gou Shisheng zhu 论统计学的对象及其科学基础—评勾适生著<统计学原理> (Discussion of the Object and Theoretical Basis of Statistics—A Review of Gou Shisheng’s Principles of Statistical).” Xin jianshe 《新建设》 (New Construction) 4, no. 5 (1951).

278 In particular, a series of articles by Wang Jianzhen 王建真, reprinted in the compilations of the NESB’s Tongji gongzuo 《统计工作》 in 1952.01, 1952.10, and 1953.01.

114

textbooks and academic discourse. Not surprisingly, therefore, 1952 witnessed a larger number of articles published on the subject, including those by Yang Jianbai and Wang

Jianzhen in the Northeast Statistics Bureau’s Statistical Work《统计工作》.279 That same year Guo Gengji 过庚吉 published an essay criticizing Jin Guobao’s 金国宝

Principles of Industrial Statistics 《工业统计学原理》 and Zou Yiren’s 邹依仁

Industrial Statistics 《工业统计》.280 In his essay, Guo claimed that in spite of claiming to have changed their texts to reflect Soviet statistical theory, these authors were guilty of still effectively following Anglo-American statistical theory. Additional essays on the hypocrisy of capitalist index numbers and the dangers of bourgeois statistical thought were also published during 1952.281

In early 1953, Xu Qian 徐前 and Liu Xin 刘新 published what was up to that point the most sustained and detailed criticism and analysis of the state of statistical theory in the PRC. Their essay, entitled “Criticism Regarding Bourgeois Statistical

Theory,” appeared in the first and second number of that year’s New Construction 《新

建设》, and expanded on the earlier critiques of several statisticians, including Gou

279 For Yang Jianbai’s 杨坚白 essays, see for instance the Tongji gongzuo 《统计工作》1952.01, 1952.2, 1952.3 and 4.

280 Guo Gengji 过庚吉. “Fandui sanbu zichan jieji tongji sixiang dusu—ping jinguobao zhu he zou yi ren zhu 反对散布资产阶级统计思想毒素——评金国宝著《工业统计学原理》和邹依仁著《工业统计》(Combat the Spread of the Toxin that is Bourgeois Statistical Thought—A Review of Jin Guobao’s Principles of Industrial Statistics and Zou Yiren’s Industrial Statistics).” 1952.

281 Fang Bingzhu 方秉铸 published (1952.10): “Cong zhishu shang kan zichan jieji tongji sixiang de xuwei benzhi 从指数上看资产阶级统计思想的虚伪本质 (A Look at the Hypocrisy of Bourgeois Statistics from the Perspective of Index Numbers)”; and Feng Jiqing 冯杞靖 published (1952.10): “Xijing tongji sixiang zhong de zichan jieji yidu 洗净统计思想中的资产阶级遗毒 (Wash Clean the Bourgeois Legacy from Statistical Thought)”.

115

Shisheng, Jin Guobao, and Zou Yiren. Liu and Xu offered specific examples to demonstrate the problems with bourgeois statistics (discussed in the following section).282

While one would have expected the article to have served as a clarion call for additional critiques, 1954 was an oddly quiet year. One possible reason was the focus on the first five-year plan (1953-1957).

The third period, from 1955 to 1957, is marked by not just a quantitative jump in the number of articles, but also by a qualitative shift in their content. Three features, in particular, characterize this period: the impact of the 1954 Soviet Resolution, the emergence of self-criticisms, and the role of the Shanghai Finance and Economics

Institute. Let us consider the 1954 resolution first. The meeting in Moscow in Spring

1954, which had categorically adjudicated in favor of statistics’ definition as a social science, had a signal influence on the wider PRC statistical community. News of the resolution and the now Soviet state-sanctioned definition of statistics, however, took some time to filter through; a complete Chinese translation of the deliberations was not published until early 1955.283 But its impact was evident in the more confident tone and tenor of the critiques that started emerging in 1955.284 Close on the heels of the Soviet resolution’s publication, Xu Qian 徐前 and Liu Xin 刘新 published an updated version of their 1953 article. It retained its earlier title, but this time was published in the SSB’s

282 Xu Qian 徐前 and Liu Xin 刘新. “Guanyu zichan jieji tongji lilun de pipan 关于资产阶级统计理论的批判 (Criticisms of Bourgeois Statistical Theory).” Xin jianshe 《新建设》, no. 1 and 2 (1953). The authors noted that the article was drafted by them in December 1952.

283 See: “Sulian tongji kexue huiyi jueyi 苏联统计学科学会议决议 (Resolution of the Soviet Meeting on Statistical Science).” Tongji gongzuo tongxun 《统计工作通讯》(1955.02): 1-4, 45

284 This is an impression based on the interviews I conducted with, among others, the Renmin University professor of statistics Yuan Wei 袁卫.

116

journal Statistical Work 《统计工作》. A brief editorial comment noted that it had been updated based on the conclusions arrived at during the 1954 Moscow Conference.285 In the ensuing months, additional essays were published by Hu Daiguang 胡代光 and Wang

Jianzhen 王健真; the former demanding the eradication of the pernicious influence of bourgeois statistics and the latter offering an analysis of the failings of bourgeois theories of index numbers.286

1956 was a particularly fruitful year for criticisms. Early in the year, Li Zhenzhou

李振周 published an essay criticizing agricultural statistics for their rightist conservative thought and demanded that such problems be overcome.287 Li was followed by Wang

Feixian 王飞宪 who critiqued the capitalist thought present in educational materials published at the start of the 50s and by Shi Qing 石青 who focused his attention on the

285 Xu Qian 徐前 and Liu Xin 刘新. “Guanyu zichan jieji tongji lilun de pipan 关于资产阶级统计理论的批判 (Criticisms of Bourgeois Statistical Theory).” Tongji gongzuo tongxun 统计工作通讯 (Statistical Work Bulletin), (1955.08).

286 Hu Daiguang 胡代光. “Ganqing zichan jieji tongji sixiang de liudu 肃清资产阶级统计思想的流毒 (Eliminate the Pernicious Influence of Bourgeois Statistical Ideology).” Tongji gongzuo 《统计工作》 (1955.10); and Wang Jianzhen 王健真. “Dui zichan jieji zhishu ‘lilun’ de pipan 对资产阶级指数‘理论’的批判 (Criticisms of the Bourgeois Theory of Index Numbers).” Tongji gongzuo 《统计工作》(1955.11).

287 Li Zhenzhou 李振周. “Chedi pipan yu kefu women nongye tongji gongzuo zhong de youqing baoshou sixiang 彻底批判与客服我们农业统计工作中的右倾保守思想 (Thoroughly Critique and Overcome Rightist Conservative Ideas in our Agricultural Statistics Work).” Tongji gongzuo 《统计工作》(1956.4).

117

statistics of infrastructure.288 Zhang Zhihong 张志鸿 and his Agricultural Statistics were critiqued by Zhang Minru 张敏如.289

1956 is also significant for witnessing the first self-criticism by a statistician.

Early in the year, Zou Yiren, who had been one of the subjects of targeted criticisms earlier in the decade, published a self-criticism.290 Gui Shizuo 桂世祚 responded at the end of the year with an essay in which he approved of and welcomed Zou’s contrition, but felt that in many instances the self-criticism did not go far enough. For good measure, in the same piece he also returned to Jin Guobao’s Principles of Industrial Statistics and offered additional criticisms.291 Perhaps in response to the increased stridency during the previous year, 1957 witnessed additional self-criticisms. Jin Guobao published a self- criticism early in 1957. As noted in the introduction above, he did not discuss his book on industrial statistics or any other works, and focused instead on his most successful textbook, Outline of Statistics 《统计学大纲》Tongjixue dagang.292 A third self-

288 Wang Feixian 王飞宪 . “Guanyu jiben tongjixue yuanli jiaocai de yixie yijian 关于几本统计学原理教材的一些意见 (A Few Recommendations for Textbooks of Basic Statistics).” Tongji gongzuo 《统计工作》 (1956.16); and Shi Qing 石青. “Pipan jiben jianshe tongji gongzuo zhong tuoli shiji de xianxiang 批判基本建设统计工作中脱离实际的现 (Criticisms Regarding Divorcing from Reality the Statistical Work in Capital Construction).” Tongji gongzuo 《统计工作》(1956.19).

289 Zhang Minru 张敏如.“ Dui ‘nongye tongjixue’ yishu de jidian yijian 对‘农业统计学’一书的几点意见 (Some Suggestions Concerning the Book “Agricultural Statistics”).” Tongji gongzuo《统计工作》 (1956.12).

290 Zou 1956.8 remains the first self-criticism that I can locate. Zou (1908-1994) was a Paris University trained mathematician and statistician who had taught at Fudan, Chongqing, and the Shanghai Business University. In 1952, he was transferred to the Shanghai Finance and Economics Institute. In addition he also held appointments at the Shanghai Academy of Social Science and Jiaotong University. For more on Zou, see: Shi Tai 施泰. “Zou Yiren 邹依仁.” Tongji yu yuce 《统计与预测》, no. 6, (2001): 52-53.

291 Gui Shizuo 桂世祚. “Ping Zou Yiren, Jin Guobao liangwei xiansheng de gongye tongjixue zhuzuo 评邹依仁、金国宝两位先生的工业统计学著作 (Criticisms of Zou Yiren and Jin Guobao’s Works on Industrial Statistics).” Tongji gongzuo 统计工作, (1956: 24-27).

292 Jin Guobao 1957.

118

criticism was published a few months later by Wang Sili 王思立, the author of books such as A General Theory of Statistics 《统计学通论》.293 Late in the year, Zou Yiren also published a short history of the evolution of statistics, and in line with his own self- criticism from the previous year, described how statistics was, after all, a social science.294

These self-criticisms followed a common pattern. Each author began by observing that he had been trained within the Anglo-American system, in some cases had spent time abroad, and was completely in thrall to bourgeois vices like formalism 形式主义 and idealism 唯心主义. After 1949, even though he may have paid lip service to the slogan of learning from the Soviet Union, he still regarded Marxist-Leninist statistical theory as being too simplistic. Only after extended study of Marxist-Leninist thought over the following five to six years had he come to realize statistics’ true object (the social world) and theoretical basis (Dialectical Materialism 辩证唯物主义).295 The process can be summed-up via the following three-point schema:296

1. He now understood that statistics was a social science that researched large- scale social phenomena. It was different from all the natural sciences. 2. Statistics studied the quantitative aspects of these large-scale social phenomena. Therefore, it was also different from all other social sciences

293 Wang Sili 王思立. “Pipan wo zai tongjixue zhuzuo zhong de zichan jieji sixiang 批判我在统计学著作中的资产阶级思想 (Criticism of Bourgeois Ideology in my Statistical Works). ” Tongji gongzuo《统计工作》(1957.10).

294 Zou Yiren 邹依仁. “Lun tongji shi yimen shehui kexue 论统计是一门社会科学 (Discussion of Statistics as a Social Science).” Caijing yanjiu 《财经研究》(1957.04).

295 Jin Guobao 1957: 28.

296 Wang Sili 1957: 27.

119

3. Finally, statistics took into account the qualitative aspects of numerous social phenomena in order to study their quantitative aspects; i.e. it took into account the interconnectedness of society. In so doing it was again different from bourgeois statistics. A final feature that distinguishes the 1955-57 years is the role that the Shanghai

Institute for Finance and Economics 上海财政经济学院 (SIFE) came to play, both as a site to which several recalcitrant ‘bourgeois’ statisticians had been relocated, as well as the venue where those same statisticians could perform their contrition and restitution via structured discussion and criticism sessions. The SIFE’s origin can be traced to Nanjing’s

Higher Normal School 南京高等师范学校, which had been established in 1917.

Departments dealing with commerce and business 商科 had migrated to Shanghai in

1921 and by 1932 they had become the National Shanghai School of Business 国立上海

商学院. In August of 1950 its name was changed to the Shanghai Institute of Finance and

Economics. Over the following three years the business and finance departments of 20 schools in eastern China were merged into it.297 As a result, by the mid-1950s, it was home to many of the major pre-1949 figures from the world of statistics, some of whom we have already encountered in this chapter and in earlier chapters: Zhu Junyi (the last

Director of Statistics under the Guomindang), Jin Guobao, and Zou Yiren, among many others.

By the mid-1950s, many of the statisticians at SIFE began organizing informal discussions and meetings that focused on the varying foibles of bourgeois statistics. One such meeting was organized in 1956 and engaged in criticism of bourgeois index

297 http://www.shufe.edu.cn/structure/shangcaigk/shangcaijj.htm, and http://www.shufe.edu.cn/structure/shangcaigk/lishiyg.htm; accessed May 05 2014.

120

numbers. Organized by the Statistics department of the SIFE, the meeting and its outcome were originally described in a report published in the first issue of The Journal of Finance and Economics 《财经研究》 during the second half of the year.298 It included the views of statisticians such as Zou Yiren (the journal’s first editor and whose self-criticism had just been published), Zhu Junyi, and Jin Guobao. The report was reprinted in abridged form later in the year in the SSB’s Statistical Work 《统计工

作》.299 The group at SIFE stayed active into 1958, when, motivated by the tumult of the

Anti-Rightist movement, they published another lengthy report on an informal meeting where the subject was the status and use of mathematical statistics within a socialist society. Among the many involved in this discussion were, what should by now appear to us, the usual suspects: Wang Sili, Jin Guobao, and Zou Yiren.300

The different periods through which we have viewed the growing intensity and changing nature of criticisms permit some generalization. What is evident is that even with a top-down dominant paradigm, consensus formation was a process that had to be negotiated. For its target audience—the statistical cadre and students—this negotiation took place through the pages of various statistics and economics journals, thereby producing the veneer of a community of scientists in pursuit of “true knowledge,” and

298 See: “Zichan jieji tongji zhishu lilun pipan 资产阶级统计指数理论批判 (Criticism of the Bourgeois Theory of Index Numbers).” Caijing yanjiu 《财经研究》(1956.01): 13-19.

299 See: “Pipan zichan jieji tongji zhishu de fan kexuexing he fandong benzhi 批判资产阶级统计指数的反科学性和反动本质 (Critique of the Reactionary and Anti-Scientific Nature of Bourgeois Statistical Indicators).” Tongji gongzuo 《统计工作》(1956.17).

300 See:: “Guanyu shuli tongji zai shehui jingji tongji kexue zhong de diwei yu zuoyong wenti de taolun 关于数理统计在社会经济统计科学中的地位与作用问题的讨论 (Discussion Regarding the Status and Use of Mathematical Statistics in Statistical Work in a Socialist Economy).” Caijing yanjiu 《财经研究》 (1958.03).

121

arriving at a consensus about it through debate and discussion—that is, approximating the idealized workings of a general scientific community. The testing of boundaries was in clearer evidence during the early 1950s, when none of those being criticized had felt it necessary to respond in print. As the space for alternate views shrunk to near nonexistence after 1955, staying above the fracas was no longer an option. As more textbooks in the new paradigm were published, especially by the SSB and by Peoples

University Press, and as more people were trained in socialist statistics, self-criticism offered the only route to salvaging one’s reputation and position. While it is impossible for us to parse an individual’s stated acceptance of the new paradigm from his true views, the whole exercise very quickly became as much about performance as it was about belief. In this way, by 1956 the burden of performing criticism had shifted from New

China’s socialist statisticians to those they sought to reform—the likes of Jin Guobao,

Wang Sili, and Zou Yiren.

Before we proceed to a discussion of how these processes affected the actual content of statistics, it is worth asking if any obvious candidates for criticism were successful at evading it. Perhaps the most famous such case is that of Chen Da 陈达

(1892-1975), the famous demographer and sociologist. Chen’s career can be neatly divided into two parts. Chen received his doctorate from Columbia in Sociology in 1923 and from that year until 1951 he taught at Qinghua University 清华大学 in Beijing and at

South West Associated Universities 西南联合大学. During that time, he led several sociological studies, authored numerous books and articles on labor, migration and population, and held a number of important administrative positions. He was also a vocal advocate of improved demographic data, in particular the establishment of a national

122

census.301 From 1938 to 1948 he directed the Institute of Census Research and developed a census program and carried out surveys in and around the southwestern city of

Kunming 昆明. After 1949, Chen’s activities become virtually untraceable and stand in stark isolation compared to his activities up to that point.302 And yet, as a practitioner who used statistical methods but did not write textbooks or seek to teach statistics he remained uninteresting to the exponents of socialist statistics.303 Thus, even though Chen was targeted during the anti-rightist campaign of 1958, he escaped the attention of socialist statisticians.

The Content of Critique

Unlike the nature and timing of the self-criticisms, which leaves open the interpretation that they may have been, in the end, about performance and survival, the impact of the imposition of socialist statistics on the content of statistical science is more readily discernable. At the general level, criticisms of bourgeois statistics were repetitions of arguments already encountered in the previous chapter. Bourgeois statistics conflated the natural and social world as a common unit of analysis.304 It lacked comprehensiveness

301 See for example: Chen, Da. Population in Modern China. Chicago: University of Chicago Press, 1946.

302 In 1951, he lost his position at Tsinghua as sociology was scrapped as a discipline. In 1957, he participated in a conference on population research in Beijing; his paper was the principal paper at the conference and was titled “New China’s Population Census of 1953 and its Relation to National Reconstruction and Demographic Research.” So far, I have been unable to track this paper down. For a mention of the paper, see: Howard L. Boorman & Richard C. Howard, eds. Biographical dictionary of Republican China, 2 Vols. New York: Columbia University Press, 1967, 238.

303 See for instance: Gong Jinyao 2000: 377-381. Gong’s biographical essay devotes only one paragraph to Chen’s post-1949 activities, noting that he delivered one paper in 1957 using data from the 1953 national census. The same skewed pattern can be observed in Boorman & Howard 1967: 235-239. Chen’s case is, of course, symptomatic in many ways of the side-lining of several major pre-1949 figures.

304 Liu Xin and Xu Qian 1955: 30.

123

and as a consequence lacked objectivity.305 It existed to exploit workers and helped capitalists and imperialists increase their profits.306 Among such broader grievances, perhaps the most trenchant and specific charge was that it placed formal mathematical doctrine 形式数学 at the very center of statistics. By doing so, it failed to distinguish between the causal mechanisms behind functions (common to the natural world) and the multifaceted relations captured by correlation (common to the social world).307 As a result, it did not accurately take into account the economic content and reality of what was being studied; nor did it adequately account for the relatedness of events.308 Or, put another way, it failed to adequately take into consideration the inter-relatedness of the quantitative 量 and qualitative 质 aspects of social phenomena.309 As Jin Guobao explained in his self-criticism: probability theory was based on the independence of events 几个事件各各独立, such as the coin toss or the throw of the dice, but “events in society are inter-related and mutually constrained and definitely not independent [of each other].”310 And so, many critiques concluded dismissively: by detaching itself from the real world, bourgeois statistics was nothing but a numbers game 数字游戏.311

305 Liu Xin & Xu Qian 1955: 32; Zou Yiren 1956: 26; “资产阶级统计指数理论批判” Caijing yanjiu 《财经研究》(1956.01): 16; Hu Daiguang 1955: 23

306 Zou Yiren 1956: 25.

307 Wang Sili 1957: 29.

308 Wang Jianzhen 1955: 40.

309 Zhang Zhiji 张知几 and Zou Yiren 邹依仁, eds. Dazhong tongji 大众统计 (People’s Statistics). Shanghai: Xin zhishi chubanshe 新知识出版社 (New Knowledge Press), 1953, 3-4.

310 Jin Guobao 1957: 29. Also see: Wang Jianzhen 1955: 40.

311 Liu Xin and Xu Qian 1955: 21, 30; Hu Daiguang 1955: 25; Jin Guobao 1957: 31.

124

These broader critiques filtered into two specific areas where their impact is clearly discernible: textbook content and the evaluation of specific statistical concepts. In the very early years, form 1949 to 1953, new educational materials including textbooks were produced either by the Northeast Statistical Bureau or by the Statistics Department of People’s University. These materials usually consisted of compilations of reports， speeches, and lectures by leaders such as Wang Sihua and Di Chaobai 狄超白 as well as

Soviet experts. By 1951, the compilations had taken the shape of sector- or activity- specific handbooks. For instance, in 1951 the first in a series of reference materials was published for trade and industrial statistics.312 Over the following two years, reference materials were published on statistical theory, agriculture, the Northeast’s statistical work, national income, and much else.313

By 1953, the Statistical Press 国家统计局出版社, the publishing division of the

State Statistics Bureau, had become active and would rapidly become a major producer of didactic material. In addition to general textbooks, the Press published books that discussed specific methods in various fields, shared experience gained, and so on.314

Provincial and city level statistics bureaus established their own presses and followed suit.

By the mid-1950s, general introductory textbooks were being produced for all levels of

312 Maoyi tongjixue cankao ziliao—yi 贸易统计学参考资料—I (Reference Materials for the Study of Trade Statistics). Beijing: People’s University Statistics Department, 1951; Gongye tongjixue cankao ziliao 工业统计学参考资料 (Reference Materials for the Study of Industrial Statistics). Beijing: People’s University Statistics Department, 1951.

313 Tongji lilun zhong de jige wenti 统计理论中的几个问题 (Some Questions in Statistical Theory). 1952; Nongye tongjixue cankao ziliao 农业统计学参考资料 (Reference Materials for the Study of Agricultural Statisitcs); 1952; Tongjixue cankao ziliao—yi 统计学参考资料—I (Introducing the Northeast’s Statistical Work—Vol. 1); 1952; and Jingji tongjixue cankao ziliao—yi 经济统计学参考资料—I (Reference Materials for the Study of Economic Statistics—Vol. 1); 1953.

314 In addition to volumes focusing on industry, agriculture, capital construction, etc., two other genres are prominent: relating experience gained and reprints(?) of lectures.

125

students by several different organizations. Representative examples of such books targeting beginners and non-specialist cadre are listed in Table 2 below.315

Table 2. Introductory Statistics Textbooks Title Authors/Editors/Compilers Publisher Date Comments

Zhang Zhiji Revised editions May 1954 and New 张知几编 January 1956. Self-Study material for the People’s Statistics316 Knowledge Zou Yiren ed 1953 common cadre; Middle-School Press 邹依仁校 level This was the 4th edition (first Lectures on Statistical People’s University People’s edition was in 1952) but the 9th Principles—Vol. 1 & August Department of Statistical University printing; including this print run, a 2317 1956 Theory Press total of 60038 copies had been

produced Common Problems in Lin Fude, Liu Xin, Jiang Statistical March First edition; By January 1957 six Statistical Theory318 Zhao, Xu Qian, Zheng Press 1956 print runs, total 59,910 copies Yao, Zhou Fugong

These new volumes differed from pre-1949 textbooks with regard to their content and their focus. In his 1954 report on Soviet statistics, Ostrovitianov had noted that the controversy over the true nature of statistics had prompted some Soviet “economists and statisticians to be afraid of making use of mathematical methods in investigating social phenomena. This, evidently, is why the methods of mathematical statistics are almost entirely neglected in the drafts of the textbook prepared by the Moscow Institute of

Economics and Statistics.”319 A similar pattern is discernable in the production of

315 More specialized textbooks and translations of Soviet textbooks also became increasingly available through the course of the decade. See: Zu Ting’an 祖廷安. “Siben gongye tongji shuji 四本工业统计书籍 (Four Books on Industrial Statistics).” Tongji gongzuo 《统计工作》(1958.01): 25.

316 Zhang Zhiji 张知几 and Zou Yiren 邹依仁, eds. Dazhong tongji 大众统计 (People’s Statistics).Shanghai: Xin zhishi chubanshe 新知识出版社 (New Knowledge Press), 1953.

317 Renmin University. Tongjixue yuanli jiangyi—shangxia ce 统计学原理讲义—上下册 (Lectures on Statistical Principles—Volumes 1 and 2). Beijing: People’s University Press, 1956.

318 State Statistics Bureau. Tongji lilun yiban wenti jianghua 统计理论一般问题讲话 (Common Problems in Statistical Theory). Beijing: China Statistical Press, 1956.

319 Ostrovitianov 1955: 329.

126

textbooks and related educational material in China. As an illustration, Table 3 below compares the tables of contents of the above three books against that of Jin’s An Outline of Statistics.

127

Table 3. Comparison of Statistics Textbooks (n.b.: the shading identifies similar content across books) Lectures on Statistical Common Problems in Title An Outline of Statistics People's Statistics Principles—Vol. 1 & 2 Statistical Theory first ed 1952, 4th edition in Year 1934 September 1953 1956 1956 Lin Fude, Liu Xin, Jiang People's University Statistics Author Jin Guobao Zhang Zhiji Zhao, Xu Qian, Zheng Yao, Department Zhou Fugong Publisher Commercial Press New Knowledge Press People's University Press Statistical Press Print run ? ? 60,000 (upto 4th Ed) 59,910 (upto 6th ed) Chapter

The Idea of Statistics The Object and Method of The Object and Method of 1 Introduction (Introduction) Statistics Statistics

The Principle Method The Organization of 2 Statistics Tables of Collecting Data: Statistical Surveys Statistics in the PRC Statistical Reports

A Subsidiary Method of 3 Statistical Charts Collecting Data: Statistical Grouping Statistical Observation Specialized Surveys The Organization and Birds of a Feather: Technology of Statistical The Aggregation and 4 Average Numbers Grouping and Sorting Summaries and Statistical Grouping of Statistical Data Tables The Organization and Making Comparisons: Technology of Summary 5 Variance Comprehensive Indices Relative Numbers Statistics: The Statistical Table Probability and Error What is the Average?-- Absolute and Relative 6 Dynamic Series normal curve Average Numbers Numbers

Changing because of Average Numbers in 7 Skewness and torque Indices/Indicators Time: Dynamic Series Statistics

Did the prices rise or 8 Index Numbers Sample Surveys Dynamic Series fall?--Price Indices Did Production Rise or Important index numbers The Graphical Representation of 9 Fall?--Product Volume Indices/Index Numbers in 5 countries Statistical Data Indices Even More Striking are The Development and Recent The Graphical 10 Correlation Diagrams: Statistical Organization of Statistics in the Representation of Statistical Illustration USSR and New China Data

11 Long-term trends

12 Seasonal variation

13 Cyclical Changes

14 Linking of Time Series

Non-linear 15 systems/correlation?

Other Types of Relations 16 (?) Partial Relations? 偏系 17 联 18 响应

19 Business Forecasting

Collection and Collation 20 of Statistical Data

128

The first striking difference is the overwhelming emphasis on the collection, grouping, and aggregation of data in the PRC era textbooks, where chapters devoted to these activities usually follow the introductory chapter (Chapters 2, 3, 4 in two cases;

Chapters 3, 4, 5 in one). In contrast, An Outline of Statistics does have early chapters discussing statistical tables and charts, but leaves a discussion of actual collection and collation to a single chapter (20) at the very end. This inversion and change of emphasis is indicative of the much more practical orientation of the PRC era textbooks. The purpose of statistics was to produce data that could inform National Construction.

Statistical work, therefore, was the principal task and enjoyed preeminence over anything else. In general, chapters in the three PRC era textbooks can be divided into four categories: (1) Introduction; (2) Methods of collecting, collating (grouping), sorting, and summarizing data; (3) exposition of key analytical tools (average numbers, indices, etc.); and (4) graphical representation of statistical results. Even the chapters on analytical tools focused only on various forms of average and index numbers (chapters 5-9; 5-8; and 6-9 respectively), which were the principal tools used to process rates of change, ascertain the level of plan implementation, and carry out related analysis.

The most salient change, however, is the absence in the PRC volumes of any chapters or content focusing on ‘mathematical’ statistics. In contrast, Jin devoted individual chapters in An Outline of Statistics to topics such as Variance (chapter 5),

Probability and Errors (chapter 6), Correlation (chapter 10), and Time Series (chapters

11-14).320 Most of these topics were predicated on the acceptance and employment of probability theory and the law of large numbers. Their elision in the PRC textbooks raises

320 In addition, by the 1950s, random sampling had become a major statistical method globally and was being widely taught. For more on this, see Part III.

129

the question of what exactly was the status of mathematics within the new socialist statistics. Chinese statisticians were quick to point out that they did not altogether reject mathematics. Rather, they insisted that what was wrong was to take mathematical statistics as the basis of all statistics; the use of mathematics to pursue statistical calculations itself was perfectly reasonable.321 Thus, it was left to decide what particular methods were best suited for use within a socialist society and economy. These debates usually centered on four concepts: Measures of Central Tendency 平均数 (Average numbers), Indices and Index numbers 指标与指数; Time Series 时间数列, and Sample

Surveys 抽样调查. Of these, the most fractious were the first two; let us turn to them next.322

Measures of Central Tendency

Among the most important concepts that came up for debate and discussion were measures of central tendency. Of the five principal measures of central tendency—the mode 众数, the median 中位, the arithmetic mean 算术平均数, the geometric mean 几何

平均数, and the harmonic mean 调和平均数/倒数平均数—it was the last three that

321 “关于数理统计在社会经济统计科学中的地位与作用问题的讨论.” Caijing yanjiu 《财经研究》 1958.3: 57. Ostrovitianov had included a similar exhortation in his 1954 report: “However, the use of the best mathematical methods, where possible and appropriate, in studying social phenomena, is neither shameful nor un-marxist. What should not be done is simply to substitute abstract mathematical formulae for politico-economic analysis.” See: Ostrovitianov 1955: 329.

322 The question of sample surveys is addressed in Part III. One additional topic that came up for sporadic discussion was the study of business cycles 商情预测, a prominent topic but one that was easy to dismiss in a socialist setting since it studied market forces, was predicated on speculation, and used the tools of probability analysis. See: Jin Guobao 1957: 31-32.

130

generated the greatest amount of debate.323 Among the key points of interest were the importance of grouping 分组 and problems of bias 偏误 and of weights 权数.324

The problem of grouping brought together the socialist concern for melding theory with economic realities on the ground. In his self-criticism, Jin Guobao cited an example from Marx’s own writings to illustrate how average numbers could easily mislead:

Imagine a situation where ten workers each worker earns 2 shillings every week, five workers each earn 5 shillings per week, and another five earn 11 shillings per week. Thus, the total weekly income of these 20 workers is 100 shillings, or 5 pounds. Now, imagine if total wages rose by 20%, so that now the total weekly wage stood at 6 pounds. If we were to calculate the new average wage, we would expect to see everyone’s wages rising by 20%. But in reality the wage hike only affected the two five-man teams, one of which saw weekly wages go up to 6 shilling, and the other to 14 shilling. Thus, half of the total workforce saw no improvement in their income, even though average income for all had indeed increased.325

As this example indicated, calculating the average wage hike actually served to distort class distinctions. A similar analysis was made of American per capita GDP calculations, which did not take into account the differences between the bourgeoisie and the proletariat, thereby covering up 掩盖 the class contradictions of capitalism.326 In textbooks, the importance of grouping was framed in a more positive light: attention to group averages permitted the identification of advanced 先进 workers as well as those lagging behind 落后.327 In this way, Chinese statisticians sought to stress that it was

323 Wang Sili 1957: 28-29.

324 Bias could be caused by weights 权偏误 or by the model itself 型偏误. Wang Jianzhen 1955: 41.

325 Jin Guobao 1957: 29.

326 Xu Qian and Liu Xin 1955: 33.

327 Common Problems in Statistical Theory 1957: 97-98.

131

absolutely critical to calculate not just total averages 总平均数 but also group averages

组平均数.328

The problem of bias 偏误 was used to interrogate which type of average was best suited for making economic calculations. Most discussions of this subject initially offered a simple example to illustrate that for any set of numbers the arithmetic mean (AM) was larger than the geometric mean (GM), which was in turn larger than the harmonic mean

(HM). For example, given the three numbers 400, 300, 200, the corresponding mean values would be329:

AM = (400+300+200)/3 = 300

GM = √ = 286

HM = 3/(1/400 + 1/300 + 1/200) = 277

But as Wang Sili explained, such numbers did not represent any real situation on the ground and therefore this example merely served to highlight the different results obtained via the use of the three different methods. Given such an example, therefore, each type of mean could theoretically be correct. If, however, examples with actual economic content were considered, then it would become apparent that in one instance it was the AM that was the only correct choice, and in another, the HM. Consider the following two cases that were discussed by Wang Jianzhen330:

328 Jin Guobao 1957: 30; Xu and Liu 1955: 30, 32-33; Wang Sili 1957: 28 also emphasized the importance of grouping.

329 Example drawn from Wang Jianzhen 1955: 41; other instances of the relationship between AM, GM, and HM were also listed in “资产阶级统计指数理论批判” 《财经研究》1956.01, pp. 13; Jin 1957: xx

330 Example from: Wang Jianzhen: 1955: 41.

132

Case I: Three types of food products 甲、乙、丙 have the following production and pricing characteristics.

A B C D Quantity Price/ton Value Product (tons 吨) (元) (%) 甲 5,00 400 20 乙 1,000 300 30 丙 2,500 200 50

If all products were considered together, what would be the combined average price of each ton be? According to Wang Jianzhen, the only suitable method to arrive at such a price was to use the arithmetic mean:

Average Price = (500x400 + 1000x300 + 2500x200) / (500+1000+2500)

= 250 元

Case II: If, on the other hand, the only information available was the price of each product and their relative value in monetary terms, that is columns C and D, then to answer the same question only the method of the harmonic mean would suffice:

Average Price = (20 + 30 + 50) / ( )

= 250 元

In neither case was the geometric mean the appropriate choice. And yet, that is precisely what bourgeois statistics would have preferred and what had been preferred by the various bureaus of the Nationalist Government prior to 1949.331 As Wang Jianzhen and other Chinese statisticians explained, bourgeois statistics considered the GM superior to the AM and the HM because it lay between the other two, and therefore was putatively

331 See Li Deyin’s 厉德寅 statement in: “Zichan jieji tongji zhishu lilun pipan 资产阶级统计指数理论批判.” Caijing yanjiu 《财政研究》 (1956.01): 17. Also see discussion of index numbers below.

133

less biased.332 Indeed, in a paper read at the eighty-second (1920) annual meeting of the

American Statistical Association, the noted mathematical economist Irving Fisher had observed:

As to bias, it can be shown that the arithmetic average has an upward bias, i.e., has in the very nature of the arithmetic process a natural tendency to give results too large, that the harmonic has a downward bias, that the geometric, median, and mode have no inherent bias in either direction.333

The tendency in the PRC, therefore, was to ignore the GM. For instance, the chapter on average numbers in People’s Statistics did not even discuss GMs, listing only the four other measures of central tendency and devoting the most space to the AM.334

But the GM was not entirely dismissed. The SSB’s Common Problems in Statistical

Theory discussed it, but explained that its use was extremely limited.335 As Hu Daiguang explained elsewhere, the appropriate use of the GM to calculate rates of growth ought to be restricted to only those instances where the individual rates represented unidirectional change (consistently up or consistently down) over the entire time-period under study.336

In all other instances, the GM was nothing but an example of mere mathematical formalism.337 And it was in this form and this form only that we see the GM being used

332 Wang Jianzhen 1955: 41; Jin Guobao 1957: 30; “Zichan jieji tongji zhishu lilun pipan 资产阶级统计指数理论批判.” Caijing yanjiu 《财经研究》(1956.01): 13.

333 Fisher, Irving. “The Best Form of Index Number.” Quarterly Publications of the American Statistical Association 17, no. 133 (March 1921): 536.

334 Zhang Zhiji 1953: 41-51 (chapter 5).

335 Common Problems in Statistical Theory 1957: 113.

336 Hu Daiguang 1955: 26-27.

337 For a more detailed explanation of this line of reasoning, see: Jin Guobao 1957: 30. As Jin explained, the geometric mean of five annual growth rates A1, A2, A3, A4, A5, was calculated by taking into account only the first and last rates, A1 and A5. Therefore, it could not tell us anything about A2, A3, or A4, or the fluctuations they may represent.

134

to calculate the combined annual growth rate in industry and agriculture.338 The implication of much of this is that several growth rates and related statistical measures were calculated using measures such as the AM, which leaves open the possibility that our sense of the changes in the 1950s is based on numbers that may possess an upward bias.

Indicators and Index numbers

The mean is a specific type of indicator and the distinction drawn between its three types was itself part of a wider discussion on indicators and index numbers. Broadly speaking, an index number is “a quantity which shows by its variations the changes over time or space of a magnitude which is not susceptible to direct measurement in itself or to direct observation in practice.”339 Examples of index numbers include those for wholesale prices, labor productivity, cost, business activity, physical volume of production, cost of living, and many more. The calculation of index numbers were considered absolutely essential to socialist statistics and it is here that the antipathy toward mathematical statistics became most transparent.340 In a book on industrial statistics A. I. Ezhov, the

Deputy Director of the Central Statistical Administration of the Soviet Union and the first

Soviet statistical expert in China (1950-1952), had stressed the superiority of socialist

338 State Statistics Bureau. Weida de shinian 伟大的十年 (Ten Great Years). Beijing: People’s Press, 1959, 90. The table in question lists the average annual rate of increase in output of major products (Steel, Pig Iron, Coal, Electric Power, etc.) for three time periods: 1950-52, 1953-57, and 1950-58. For each time- period, the method used is the GM. The data used to calculate these rates is in tables on p. 84-89.

339 “Important features in the construction of an index number are its coverage, base period, weighting system and method of averaging observations. The term can also be applied to a series of values which are standardized by being referred to a basic period or area.” The Oxford Dictionary of Statistical Terms (2003): 197. A contemporary Chinese definition can be found in Xu Qian and Liu Xin 1955: 34.

340 Xu Qian and Liu Xin 1955: 34.

135

statistics by observing that: “Industrial statistics of all major capitalist countries does not, owing to the absence of a centralized governmental statistics, possess a number of important indicators and many indicators are estimated indirectly.”341

The central challenge in arriving at a good index number is to be able to account for changes in all variables; for instance in the case of price indices, this would mean accounting for changes in price and in quantity. While numerous indices had been proposed and numerous tests devised to check their quality, the problem of over- or under-estimation had remained unresolved.342 To get around this problem, in 1921 the mathematical economist Irving Fisher had proposed what he labeled the ‘two supreme tests’ that could help determine the formula for an ideal index: first, that it “should work both ways as to the two factors, prices and quantities”; and second, that “the formula should work both ways as to time.”343

Irving Fisher, who was regarded as the foremost American economist by several of his most accomplished peers, including Joseph Schumpeter, James Tobin, and Milton

Friedman, was also a brilliant statistician and among the pioneers of the subfield of econometrics.344 Among the many areas in which he made seminal contributions was the

341 A. I. Ezhov Industrial Statistics ( т т т п о ыш о т ). Calcutta: Statistical Publishing House, 1958, 2.

342 For instance, two basic price indices are the Paasche index (named after the Hermann Paasche) and the Laspeyres index (named after the Etienne Laspeyres). In the former, prices are weighted by the quantities of the given period and in the latter they are weighted by the quantities in the base period. See: The Oxford Dictionary of Statistical Terms (2003): 227-228, 299.

343 Fisher 1921: 534. The equivalent Chinese terms were: 时间互换测验 and 因子互换测验 (source?)

344 See, for instance: Joseph Schumpeter (Ten Great Economists 1952: 223), James Tobin 1987, and Milton Friedman (Money Mischief 1992: 37).

136

study of index numbers.345 Fisher also had a direct link to Chinese statistical practice during the 1920s and 1930s. During his time as a PhD student at Yale, the economist

Franklin L. Ho (He Lian 何廉, 1895-1975) had worked with Fisher in developing index numbers. Upon his graduation in 1926, Ho returned to China and took up an appointment as professor of public finance and statistics at Nankai University. At Nankai, Ho put to good use the $500 gift he had received from Fisher and organized the Committee on

Social and Economic Research. This committee would eventually become the

Department of Economics in the early 1930s. Among Ho’s first research related activities was the compilation of historical price indices calculated by applying Fisher’s ideal index to over sixty years of price and quantity data from the Chinese Maritime Customs Service.

These indices, as well as others that Ho and his colleagues calculated, were published between 1934 and 1936 in Nankai Index Numbers.346

For China’s socialist statisticians after 1949, however, in promoting the concept of an ideal index formula 理想公式347, Fisher was conflating two distinct problems: what type of mean to use and what type of weights to use. According to them, Fisher had begun by assuming that the problem was a mathematical one. He had completely ignored the economic content of the various indices. 348 Fisher thus became the embodiment for

345 See: Fisher 1922: 533–537; and Irving Fisher. The Making of Index Numbers: A Study of Their Varieties, Tests, and Reliability. Boston: Houghton Mifflin Compan, 1922.

346 H.D. Fong. “Recollections of Early Research on Chinese Economy.” The Tsing Hua Journal of Chinese Studies 清华学报 3 & 4, no. 1 (1963): 69-86, 70, 72. Also see: Trescott 2007: 263-264 (the proliferation of index number calculation across China); Chiang 2001: 84 (Franklin Ho’s obsession with calculating indices).

347 Jin Guobao 1957: 30

348 “资产阶级统计指数理论批判” 《财经研究》1956.01, pp. 19; Jin Guobao 1957: 31; Xu Qian and Liu Xin 1955: 34

137

the failings of bourgeois statistics as a whole.349 His two tests as well as “eight big formulas” 两大测验与八大公式 were discussed and their failings dissected in detail by several interlocutors.350

Fisher may have served as a productive focal point for the critique of Anglo-

American bourgeois index number theory, but for all their deliberations, socialist statisticians in China did not propose any suitable or sophisticated alternatives.351 Given such an absence, the most frequently used types of index numbers were of a much simpler kind, which focused on year-on-year change, or took one particular year as the base year and tracked nominal changes in future years based on that. The following two tables from Ten Great Years provide an example of each such case.352

349 Wang Jianzhen 1955: 41: “这些反动的观点在资产阶级统计和指数“理论”的“巨擘”费喧 [Fisher] 的著作中，表现得最为彻底和露骨.”

350 See, for instance: i) “Zichan jieji tongji zhishu lilun pipan 资产阶级统计指数理论批判.” Caijing yanjiu 《财政研究》 (1956.01): 17-18; ii) Jin Guobao 1957: 28, 30-31; and (iii) Wang Jianzhen 1955: 41.

351 Discussions incorporated a critique of a wide variety of index numbers, such as composite index 综合指标, aggregate index 总量指标, relative index 相对指标, average index 平均指标, structural index 结构指标, comparative index 比较指标, strength index 强度指标, dynamic index 动态指标, and so on.

352 Ten Great Years 1959: 16, 80.

138

The table above provides index numbers for annual Industrial and Agricultural

Production from 1950 to 1959. For each year (row), the previous year (row immediately above) is taken to be the base (i.e. 100). Thus, we can see that total industrial and agricultural output in 1957 was 107.8 or 7.8% greater than that in 1956. The primary driver of this increase was industrial growth (114 or 11.4%) and not agricultural growth

(103.5 or 3.5%). In the table below, on the other hand, all comparisons were made taking

1949 as the base year (=100). Thus, the growth in modern industry by 1957 was just over

7 times (703.1) that in 1949.

139

Conclusion

Neither of these transitions—the disciplining of a generation of statisticians and the changes in content—happened overnight, even though that was very much the idea behind the re-organization of schools and disciplines in 1952 院系调整.353 Genuine concerns about capitalism—such as the existence of incentives to dress up data to fool people and favor private interests—were conflated with the perceived weakness of

353 The readjustment of universities and faculties 高校院系调整 was carried out in an attempt to Sovietize the existing educational structure. In the humanities and social sciences the starkest change was in the de- legitimization of such disciplines as Sociology and Anthropology, both of which ceased to exist.

140

specific statistical methods. In some ways, many of these criticisms and self-criticisms may have been storms in a teapot, since the bigger and significantly more far-reaching changes were happening in the world of actual statistical work. Indeed, the far more telling rupture in statistics in the early PRC was in statistical practice, where changes in practice preceded and served as markers for the changes in scientific discourse that were to follow. In Part II of this dissertation we shall explore different elements of statistical practice in greater detail.

The narrower range of statistical theory that was taught starting in the 1950s did, however, significantly limit the abilities of several generations of practicing statisticians.

The strong boundaries established between the social science of statistics and mathematical statistics left both communities of scientists (and practitioners) mutually ignorant. Particularly revealing in this regard is the case of the statistician Xu Baolu 许宝

騄.354 Born in 1910355 in Beijing, Xu did his undergraduate work at Tsinghua University

清华大学, from where he graduated with a degree in mathematics. He then spent four years at University College, London, earning a PhD in 1938 and a ScD in 1940, both in mathematical statistics. Upon his return to Beijing, he was appointed a professor in the

Department of Mathematics at Peking University. Barring a two-year stint in the United

354 For more on Xu Baolu 许宝騄, see: (1) T.W Anderson, K.L. Chung, and E.L. Lehmann. “Pao-Lu Hsu 1909-1970.” The Annals of Statistics 7, no. 3 (1979): 467-470. (2) Jiang Xuepei 江学培. “Beijing daxue jinian—zhuming shuxuejia, tongjixuejia, Xu Baolu jiaoshou 北京大学纪念—著名数学家、统计学家许宝騄教授 (Beijing Univeristy Memorial—The Famous Mathematician and Statistician, Professor Xu Baolu).” Tongji 统计, (1981.01): 39. (3) Chen Jiading and Zheng Zhongguo. “Academic Achievements of Professor P.L. Hsu.” Review essay on the website of the International Chinese Statistical Association, 2010. Accessed online: http://www.icsa.org/awards/PLHsu2.pdf.

355 There appears to be some disagreement on his date of birth. One biography lists September 1, 1910; another 1909.

141

States from 1945 to 1947, he spent the rest of his career there. Xu’s major area of research was probability theory and mathematical statistics and he is today identified as the first person in China to engage in research in these areas:356 “He was the first Chinese who was internationally recognized in the area of probability and statistics,” and counted among his friends E.A. Pearson and Jerzy Neyman.357 These are precisely the areas that socialist statistics refused to include within its purview. And yet, none of the criticisms targeting bourgeois statistics ever touched Xu. Why?—because he was based in a department of mathematics.

This dichotomy between statistics and mathematical statistics cut both ways—it constrained what each knew of the other—and remained influential into the 1980s. In a recent conversation, the Harvard statistician Xiao-Li Meng related to me that the first time he encountered a histogram with actual data was in graduate school in the United

States in 1986. Before that, as a student of mathematics and mathematical statistics at

Fudan University in Shanghai, his curriculum had been purely theoretical and did not involve engagement with data.358

356 See: Jiang Xuepei 1981: 39.

357 Chen and Zheng 2010: 1; Anderson, Chung, and Lehmann 1979: 468. A survey of Chinese contributions to statistics noted: “It was Professor Pao-Lu Hsu (许宝騄) who first introduced modern statistics to China in the 1940s. He was also the first Chinese statistician to win a renowned international reputation in his field. ” See: Chen, X. R., K.T. Fang, and C.C. Yang, eds. The Development of Statsistics: Recent Contributions from China. Harlow, Essex: Longman Scientific and Technical, 1992, 1.

358 Discussion with Xiao-Li Meng, Professor of Statistics and Dean of GSAS, Harvard University, Cambridge, MA, 12 December 2013.

142

PART II

STATISTICS IN ACTION

143

CHAPTER V

The Nature of Statistical Work

144

Too much Civil Service work consists of circulating information that isn't relevant about subjects that don't matter to people who aren't interested.359

From a very early stage, the idealized version of statistical work that had been espoused in the writings of Wang Sihua and claimed as having been largely achieved by the

Northeast Statistical Bureau had to face up to reality of expansion across a vast, complex, and largely agrarian landscape. Through the work of the Northeast Statistics Bureau and, in particular, the writings of Wang Sihua, a model of statistical work had been articulated that valorized extensiveness (across economy and society), completeness (of the statistical system), and objectivity (about social fact). This model was predicated on the primacy of exhaustive enumeration; other methods of data collection, notably surveys, were of secondary importance and only to be used in a supplementary or as-needed basis.

As shown in Part I, that such a valuation of methods was the only correct choice had been based on a concerted and unambiguous theoretical understanding of the social world and how to account for it. By 1951, the state and party leadership were in a position to start thinking about expanding the Northeast’s system across the country and establishing a unified national statistical system.

This chapter focuses principally on the roughly six years from 1952 to 1957, a period during which a national statistical bureau and a large nationwide network of subsidiary statistical offices at provincial, city and county levels were set up. The concern here is no longer with the articulation of an ideal, but with its actual practice. As a period that includes within it China’s first five-year plan (1953-1957), it witnessed the

359 Anthony Jay and Jonathan Lynn 1989. The Complete Yes Minister. BBC Books; New Ed edition. Preface. Source: http://www.jonathanlynn.com/tv/yes_minister_series/yes_minister_episode_quotes.htm

145

inextricable linking, at least in a putative sense, of statistics and planning. In the third chapter in this part, I shall explore this linkage in greater detail. The goal of this chapter, however, is to offer a sense of the nature of statistical work and the challenges that it faced. The chapter is divided into five principal sections. The first section offers a brief institutional history of the establishment of the State Statistics Bureau and the expansion of a national network under it. The remaining sections are organized schematically around the four key tasks of statistical work in the PRC: Collection 收集,

Collation/Ordering 整理, Research 研究, and Supply 提供.360 Of these four, collection 收

集 and supply 提供 were considered the more fundamental tasks.361

The nature of statistical activity during this period not only mirrored broader bureaucratic and governance-related trends and forces frequently identified with the mid-

1950s, but by constraining them also gave them their shape.362 The first of these key trends was the increasing degree of complexity across the economy as new factories and industries were set up, agriculture was re-organized and subsequently nationalized, and a new system of remuneration put in place, all against the background of sustained

360 See for instance, Xue Muqiao’s closing remarks at the Sixth National Meeting of Statistical Work (薛暮桥局长在第六届全国统计工作会议上的总结报告(记录)). BMA 133-001-00090: 83 ( p. 3 of 7 of document). Xue had used a similar list earlier in during the same meeting, when delivering his report on experience gained in statistical work over the course of the first five-year plan. See: Xue Muqiao 薛暮桥. “Di yige wunian jihua qijian woguo tongji gongzuo de chubu jingyan he jinhou renwu 第一个五年计划期间我国统计工作的初步经验和今后任务 (The Preliminary Experience and Recent Tasks of Statistical Work during our country’s First Five-Year Plan Period).” Tongji gongzuo 统计工作 1957.21: 1-21, 1, 6. Numerous other instances of similar listings exist. For instance, in his self-criticism Zou Yiren also listed the same four tasks. See: Zou Yiren 1956: 26

361 Xue Muqiao 1957: 5. This categorization is, of course, not unique to the PRC or to socialist statistics. Jin Guobao in his An Outline of Statistics (1934) also offered the same four-fold breakdown for the procedural methods associated with statistics. See: Jin Guobao 1934: 8.

362 For a general discussion of this period, Franz Schurmann’s work still remains a useful resource. See: Schurmann 1968.

146

economic growth. For statistics, this meant an expansion of the types of data that needed to be collected as well as an expansion into the more virgin territory of performing statistical analysis and research based on the data that had been collected. Complexity also brought into focus issues of bureaucratic organization and coordination across various bureaus and organs. Another key trend through the mid-1950s was the push towards greater decentralization. Decision-making power, especially as it pertained to implementation of policies decided at the central level, was frequently devolved to the

Provinces or lower. In statistics, this broader push toward decentralization came up against the centralizing tendencies of the statistical apparatus. A resolution was attempted via the principle of ‘unified leadership, dispersed (by grade) management’ 统一领导，分

级管理, which favored allowing local party and government administration to execute the work as they saw fit and without too much oversight.363 This tension between centralization and decentralization, in turn, brings into focus the two forms of political and executive authority—horizontal and vertical—that were a feature of both the organization and prosecution of statistical work.364 Finally, a third key feature of the period was the realization that agriculture was absolutely critical to the economic fortunes and socialist goals of the PRC. Best articulated in statements that simultaneously reiterated the primacy of industry while also elevating agriculture to near equal status, such a rebalancing of emphasis forced the SSB to confront the unreliability of agricultural data and the unsuitability of the methods used to collect that data.

363 BMA 133-001-00090: 83.

364 For a general discussion of this decentralization as well as vertical and horizontal forms of governance, see Schurmann 1968.

147

Statistical work was thus carried out within the larger unfolding context of increasing complexity, changing (economic) focus, and struggles over administrative devolution. As work got more complicated and demanding, the engagement with and discussion about statistical activities grew more sophisticated—methods were expanded, analysis was undertaken, and the results of such analysis circulated amongst leaders and bureaucrats. However, as the following chapter will show, a key problem remained—the system that had been set up incentivized the over-production of reports and the state had little or no capacity to handle the resulting excess reports. This meant that in the interests of maintaining timeliness 及时性, which director of the SSB, Xue Muqiao would repeatedly state was of higher importance than correctness 正确性, there was an ever increasing disjuncture between local data and regional and national data. Whereas local data collected was often of reasonably good quality, the need to report it up it in a timely fashion meant that estimation was permitted.365 As a result, the further up it was reported, the more estimation was likely used, rendering provincial and eventually national data victim to gross inaccuracies.

Furthermore, given their reliance on the systematic collection, collation, and upward reporting of data, statisticians had no independent and systematic means available to check the accuracy of the numbers their system was generating.366 At the same time, such a system was particularly prone to delays. Archival reports of statistical work are full of complaints about factories, work units, as well as agricultural cooperatives, all falling behind reporting schedules. Delays combined with excess issuing thus created an absolutely chaotic situation by the mid-1950s, leading Wang Sihua to lament that

365 See, for instance: BMA 002-020-00746: 12.

366 The best methods were to do spot-checks or one-time surveys. See: People’s Statistics 1956: 9-13.

148

statistical work remained unable to address the needs of the Planning Commission and the demands of the nation’s leaders.367

Expansion of a National Network

Prior to September of 1952, when the State Statistics Bureau (the National Statistical agency) was set up, the only bureau 局 of statistical work in the People’s Republic was the Northeast Statistics Bureau, headquartered in Shenyang (and discussed in Part I,

Chapter 1). That does not, of course, mean that no statistical work was carried out during the first three years. Much of this work was coordinated through Finance and Economics

Committees 财政经济委员会 and statistical departments 处 and branches 科, which collected data or prepared estimates based on pre-existing data. In similar fashion, regional and provincial levels of government possessed statistics divisions, as did the various ministries. Most statistical reports were compiled through their efforts.368 The key national level coordinating institution was the Central Finance and Economics Committee, which was set up on July 21, 1949. Composed of six departments, it was led by Chen

Yun 陈云, who was in turn supported by such figures as Bo Yibo 薄一波, Ma Yinchu 马

寅初, Li Fuchun 李富春, and Xue Muqiao.369

It was this Central Finance and Economics Committee that initiated discussions about a national statistical system. In July of 1951, Li Fuchun 李富春, a Deputy Director

367 See: Wang 1986: 100-101.

368 For instance, a Work Department of the CPC Central Finance and Economy was set up in July of 1948 headed by Dong Biwu 董必武, with Xue Muqiao 薛暮桥 as the Secretary. A Department of Customs Statistics was set up in early 1950. See: DSJ: 1-2

369 See: DSJ: 2.

149

of the Committee, chaired a national level conference on Financial and Economic

Statistics to start discussing how to implement the Northeast model on a national scale.

Held in Beijing from the 16th to the 20th of the month, this conference would in later years be retroactively dubbed the first National Meeting on Statistical Work.370 Eighty-five delegates, joined by 91 observers 列席代表, gathered to address for the first time the statistical needs of the country as a whole.371 Echoing Stalin decades earlier, Li would observe in his speech that without statistical data the tasks of national construction, planning, and economic work could not be undertaken. He also went on to make clear that China would learn socialism from the Soviet Union, and take Marxism-Leninism as her guiding principle, criticizing Anglo-American capitalist statistics along the way. But the key task before the conference was the question of a national statistical system; something that even after two years the PRC did not possess. It was Wang Sihua who proposed setting up a national statistical agency, which would centrally guide statistical work across the country.372

The details of such a national system were articulated by the main Soviet

Statistical expert and advisor, A. Ezhov 叶诺夫.373 In an address delivered on the penultimate day of the conference, Ezhov pointed out that while all across at the regional

370 In all, there would be another five such annual national meetings of statistical work 全国统计工作会议, all held in Beijing: December 1952, February 1954, February 1955, February 1956, and September 1957. For details, see DSJ and the three volumes of Tongji gongzuo zhongyao wenjian huibian 《统计工作重要文件汇编》 (Compilation of Important Documents Relating to Statistical Work). Beijing: Statistical Press, 1955, 1957, 1959.

371 Discussions ranged across how to ascertain guiding principles 方针 and specific tasks 任务, what methods 方法 to employ, organizational questions 机构, and the training and cultivation of cadre 干部.

372 See: DSJ: 6; Wang 1986: 225

373 As discussed in Part I, Ezhov, a deputy director of the Soviet Statistical Bureau, arrived in China sometime in early 1950 and returned to the Soviet Union presumably sometime in February 1952.

150

level a lot of statistical work had been done, China still lacked national statistics under a unified leadership structure. Therefore, the most urgent task was to link the different statistical activities under one unified system. Yezhov identified nine key tasks. The most important of these was the absolute necessity of one unified seamless system that linked the local to the central and the ministry to the work unit, with clear lines of command and oversight. To man this system, a large body of trained cadre was needed. The Northeast

Statistics Bureau was to serve as the model for all other provincial and city level bureaus across the country as they went about setting-up such a system.374

It took nearly another year for the National Meeting’s recommendations to be put into effect. On August 7 1952, Xue Muqiao was appointed the head of the newly established State Statistics Bureau (SSB).375 The Bureau began operations the following month, on September 8. In a report delivered to the Central Finance Committee 中财委,

Xue noted that the Bureau had set up a secretariat 秘书厅, and departments for summary/comprehensive 综合 statistics, industrial statistics, agricultural statistics, trade statistics, capital construction, and communications statistics. In addition, also set up

374 The nine points were (See: DSJ: 5-6): 1. From the enterprise to the bureau/department and from the local to the central, institutions and systems should be set up as part of one seamless system. 2. Cultivation of cadre must begin with cultivation of their thought. Their training can be dispersed, but must be centrally administered. Should also consider distance learning. 3. The Center and regional administrations must produce a Statistical Work Handbook 《统计工作手册》 4. Upper levels of the apparatus must oversee the lower levels: inspect their work, enforce deadlines, and offer guidance and help. 5. Rapidly clean-up useless, chaotically issued reports. 清理无用的、乱发的报表. 6. Reports should be simplified, especially comprehensive reports. 7. Lower levels should report, upper levels should offer judgment and instructions 8. Extensively use and study the experience of the NE. 9. Extract the essence 抽取精华 from already collected statistical data and carry out analyses so that it is easy to compare [and assess] the development that has been made

375 See: DSJ: 7. Wang Sihua was appointed one of the Bureau’s Deputy Directors and put in charge of Industrial Statistics. See: Wang 1986: 225.

151

were sections devoted to labor wage statistics 劳动工资统计, distribution of goods 物资

分配, and the world economy 世界经济, for a total of ten sections.376 Departments devoted to culture and education and to social statistics 社会统计 had yet to be set up.

The plan was to employ about 250 people across these offices (excluding those working part-time).377

A few months after the establishment of the SSB, a Second National Meeting of

Statistical Work took place from 15 to 22 December in Beijing. It included an agenda- setting address by the famed military commander and recently appointed vice-chairman of the central government, Zhu De 朱德. For the most part, Zhu’s speech covered ground already trodden by Ezhov. One notable difference was in how he framed the importance of statistical work. In this, he was much more in agreement with Wang Sihua’s speeches from his time in the Northeast: statistical work was necessary because it was the basis of a planned economy. Zhu thus reiterated Wang Sihua’s call for a united statistical apparatus and his exhortation was subsequently repeated by Xue Muqiao in his own speech to the delegates.

One of the primary tasks identified at the Second National Meeting was the need to set up regional, provincial, and city level statistical offices as part of one unified national network. In a letter dated 14 April 1953 and delivered to all regions, provinces, and cities, Xue Muqiao noted that by March-end five regions and eighteen provinces and

376 See: DSJ: 7-8. For a chart depicting the seven departments in 1952, see DSJ: 492

377 In a survey conducted over the following month it was established that there were 354 statistical workers in the main offices of the six administrative regions and in Inner Mongolia. Another 766 statistical workers were in the various province level finance committees. The total thus was 1120 workers: North 98, Northeast 388, East 240, South-Central 168, Southwest 111, Northwest 95, and Inner Mongolia 20. See: DSJ: 9

152

cities had reported the successful establishment of statistical offices.378 But the status of statistical offices of the South-Central Region as well as twenty-three other provinces and cities remained unknown. He proceeded to accept the SSB’s culpability in causing some of these delays: the organizational guidelines about a national statistical network had, after all, only been disseminated in early March. Even so, he urged the recalcitrant to rapidly establish statistical offices.379

As shown in Table 1, Xue’s exhortation met with direct results. By the end of

June 1953, all but Liaoning and Qinghai had set up regional, provincial, or city-level statistical offices.380 According to the sanctioned size of different provincial bureaus at time of establishment, 1275 statistical workers were distributed across twenty-four bureaus. Average bureau size was about fifty-three workers.381 City level bureaus, however, were often larger: Shanghai had ninety-six statistical workers and Beijing sixty- five. Most bureaus were organized into Departments 处, Offices 室, Branches 科, and

Teams 组.382

378 These were: Northeast, North, East, Southwest, Northwest, and Inner Mongolia, Beijing, Tianjin, Hebei, Shaanxi, Suiyuan 绥远 (), Guangdong, Hubei, Hunan, Henan, Shanghai, Jiangsu, Anhui, Chongqing, Yunnan, Xi’an, Shanxi, Shenyang, Luda 旅大. See BMA 002-020-00670: 1.

379 BMA 002-020-00670: 1. The directive was issued on April 14 1953.

380 The Liaoning 辽宁 Statistical Bureau was formed by merging the Liaodong 辽东 and Liaoxi 辽西 bureaus in August 1954. The Liaodong and Liaoxi Statistics Bureaus had been set up separately in May and August 1950 respectively, under the aegis of the Northeast Statistics Bureau. Evidence from the directive, the first document in BMA 002-020-00670, is not always consistent with the dates published in the DSJ. For instance, the BMA document notes that Anhui had already set up a statistical bureau by March 1953. The DSJ, however, lists it as happening in June of that year.

381 These numbers and estimates are based on information culled from the DSJ: 7-35.

382 I have not found uniform guidelines delineating a department from an office and a branch; considerable variation likely existed. For instance, in 1957, the Henan Statistics Bureau made the following recommendation: cities with populations greater than 0.5 million should set up a Bureau 局, those with a population between 0.2 and 0.5 million should set up a Department 处, and those with fewer than 0.2 million should set up an Office 室. See: BMA 133-001-00091: 14.

153

Table 1. Establishment of National, Regional, Provincial, and City Level Statistics Bureaus Source: 中华人民共和国统计大事记，1949-2009 （国家统计局编）, pp. 1-35 Authorized Date set up Province/City Staff Notes

August 7 1952 National SSB Established, Xue Muqiao Director September 8 1952 National SSB Starts Operations Set up stats 局、科、处 in various bureaus and organs of the January 12 1953 National central government November 26 1954 National SSB first Bureau listed under reorganized State Council

April 4 1950 Northeast ? January of 1953 Northwest 80 Stopped funtioning on August 5 1954, when large admin February 1 1953 North China 60 regions were canceled. March 13 1953 Southwest 70 Located in Chongqing March of 1953 East 110 Located in Shanghai. Stopped work by end of year. May 1 1953 South-Central 80

The Suiyuan 绥远 Bureau was set up in March 1953, and later December of 1952 Inner Mongolia 45 incorprated into the Inner Mongolia Bureau January of 1953 Hebei 70 January of 1953 Hubei 60 Built on work done by the province's 调查统计委员会 and February 1 1953 Guangdong 57 the 华南财委计划局统计处 February 9 1953 Yunnan 55 March 4 1953 Zhejiang 65 March 12 1953 Shanxi 55 March 14 1953 Shandong 90 March 15 1953 Henan 66 April 1 1953 Shaanxi April 24 1953 Guizhou April of 1953 Heilongjiang 50 April of 1953 Jiangxi 50 Prior organ: 省财委统计室 May 25 1953 Sichuan 80 Established by abolishing May of 1953 Jilin 55 吉林省人民经济计划委员会统计处 May of 1953 Hunan 50 统计处 was changed to 统计局. Oct. 1 1955 the 新疆维吾尔自治区 was established, 统计局 accordingly May of 1953 Xinjiang 15 transferred. June 1957 started setting up 科室机构. June 9 1953 Anhui 65 Built on foundation of: 安徽省人民政府财经委员会统计处 In December 1957, became part of the 福建省计划委员会; In June of 1953 Fujian 45 October 1959 it was again separated became again June of 1953 Guangxi In 1958 it was changed to the 广西壮族自治区统计局 June of 1953 Ningxia 99 1954.9.23: Ningxia merged with Gansu July of 1953 Gansu 50 辽东 and 辽西 were merged to form 辽东. 辽东 and 辽西 August of 1954 Liaoning 124 Bureaus had been set up separately in May and August 1950 January of 1955 Qinghai 29 Previously: 省人民政府计划委员会统计处

1950? Shenyang 1953? Chongqing March 5 actually set up on the foundation of the Jaunary 30 1953 Beijing 65 北京市财委统计科 Formed by combining: (1)津市财政经济委员会统计处，(2)天津市委国营工业部统 March 5 1953 Tianjin 65 计组，and (3)天津市人民政府研究室统计科 March? 1953 Xi'an April 1 1953 Shanghai 96 154

As a result of the restructuring of government organs that took place in 1954 (in the wake of the first people’s Congress), the State Statistics Bureau was identified as the first among 20 organizations [机关 or 机构] directly subordinate 直属 to the newly formed State Council 国务院 (which had replaced the Government Administrative

Council of the Common Program era, 1949-1954).383 This is the national structure that remained in place till the summer of 1958.

By 1956, the SSB main office in Beijing had expanded and consisted of thirteen key branches. Two of these were offices 室 dedicated to administration and research. Six divisions 司 handled statistics relating to comprehensive/summary 综合, industry, agriculture, capital construction, trade, and distribution of supplies. In addition, there were five departments 处 dedicated to transportation, labor wages, culture, education, and compilation (publication).384 By the following year, the total number of statistical cadre claimed was in the region of 190,000 to 200,000.385 The SSB headquarters in Beijing itself reported about 650 dedicated staff. Outside of regional, provincial, and city level offices, statistical activity was undertaken by a mixture of full-time and part-time staff.

This was especially the case at the more local level. For instance, statisticians at the factory level frequently doubled up as accountants 会计 as well.

383 The reorganization occurred after the First People’s Congress, wherein the Common Program, in place since 1949, was set aside for a more formal governmental and administrative structure. For a copy of the actual government notice 通知, see: BMA 153-001-01221: 1-3. Also see, Schurmann 1968: 183-185, where the full list is reproduced. For Schurmann, the order of the list indicated the relative importance of the organ. I have not found anything to categorically substantiate this.

384 DSJ: 492.

385 See: State Statistics Bureau. Re’ai women de tongji gongzuo 热爱我们的统计工作 (Ardently Love our Statistical Work). Beijing: China Statistical Press, 1956, 10; and Tongji ganbu de zixue 统计干部的自学 (The Self-Study of Statistical Cadre). Beijing: Statistical Press, 1957.

155

By 1957, the four-fold description of statistical work—collection, collation, research, and supply—reflected both the de jure and the de facto situation.386 Of these four, collection, collation and supply were part of the original tasks listed by Wang Sihua when describing the work of the Northeast Statistics Bureau. Research at that time was largely aspirational. But by the end of 1957, it had come to occupy a normatively equal status. In reality, however, statistical workers perceived the importance of these activities in unequal ways. In 1957, Xue Muqiao noted that some workers preferred performing supervisory 监督检查 or research 研究分析 tasks over the more basic tasks of collection and collation, thereby failing to recognize that without these latter tasks the former were impossible to undertake. 387 As he explained:

If, in order to supervise or [carry out] research, one were to become relaxed about the collection and collation of statistical work, this was akin to one destroying the foundations of one’s own work; exactly like the scholar who having lost his pen takes a sword to a [literary] competition at a colleague’s home.388

The overemphasis on control was at least partially contingent on the early experience gained in the Northeast. In 1951, Li Fuchun had observed that statistical work should first focus on the industrial sector. Later, with additional experience and capabilities, the work could expand to cover other sectors. The Industrial sector, given its localization in space

386 See for instance, Xue Muqiao’s various speeches from the Sixth National Statistical Meeting, in BMA 133-001-00090: 83.

387 Xue Muqiao 1957: 6. A truncated version of this report is also located in BMA 133-01-00090: 66-80 (第六屆全國統計工作會議文件 (一)), (it is missing the final two sections in the published version (on 如何运用各种调查方法（略）and 统计工作的今后任务 (略)). The report was also reproduced in 《统计工作重要文件彙編—第三輯》, 1959. In similar fashion, collection (here identified as survey 调查) was described as the basis of all other statistics work in the 1956 introductory textbook People’s Statistics (see p. 9).

388 Xue Muqiao 1957: 6.

156

and its discrete nature (when compared to villages or cooperatives), permitted a focus on control 监督作用 and vertical leadership 垂直领导.389 In the following chapter, I shall look more closely at such problems of professionalization and motivation.

Collection

Collection 收集 (the term 搜集 was also frequently used) was the most basic statistical task performed by the SSB and its subsidiary organs. For Chinese statisticians and their

Soviet advisors, the best way to count was to count exhaustively. As explained in Part I,

Chapter 1, this understanding generated a hierarchy of methods used for data collection.

The most important, because it was the most fundamental, was the Complete

Enumeration Periodical Report System 全面定期统计报表制度. The periodicity of these reports ranged from ten-day (旬) cycles, to monthly, from quarterly, and annual reports. Amongst these, the Annual Report was the most important, especially because it was used most directly in arriving at planning targets and in checking their success and in summarizing the state of affairs in the economy and in society.390 Reports at shorter frequencies were more useful at the provincial level and below. Of secondary importance, but widespread use, were censuses 普查. Finally, unlike the first two methods which were based on complete enumeration, surveys 调查 which only counted a selected sample 样本 were also used, but sparingly.391

389 BMA 133-001-00090: 82.

390 Xue Muqiao 1957: 6).

391 For a similar hierarchy of methods, see: People’s Statistics 1956: 9-13

157

Over the course of the 1950s, there was a general trend in the direction of increasing complexity, both in terms of the nature of data collected, as well as in terms of the methods used to collect that data. At the start of the 1950s, the emphasis was on setting up the periodical report system and conducting basic assessments of industry. If one-time specific sets of data were needed, then the census was the chosen form to do get such data.392 For example, even before the establishment of the SSB, a census of state- owned and public-private industries had been carried out in 1950. In 1952, the year the

SSB was established, censuses of industrial production and employment were completed.

A population census in 1953 was followed by censuses of private enterprises and the handicrafts industry in 1954. Private activity, in particular, was tracked using the census method. By the mid-1950s several additional censuses were announced with this specific goal. For instance, an urgent notice issued on October 20 1955 ordered that four types of surveys were of critical importance and must be completed in the fourth quarter of the year. These were surveys of Merchants, Private workers in Key Industries, National

Employment, and Material Inventory.393 That same month, another report also issued via telegraph spoke of the completion rates across provinces and cities of surveys of Private

Businesses and Catering.394

392 Another listing of major censuses 普查 carried out in the 1950s is provided in: Wu Hui 吴辉. “Zhongguo de tongji diaocha fangshi 中国的统计调查方式 (China’s Statistical Survey Methods).” In Zhongguo shehui zhuyi tongji gongzuo de jianli yu fazhan 中国社会主义统计工作的建立与发展 (The Establishment and Development of Socialist Statistical Work in China), edited by the China State Statistical Bureau’s Institute of Statistical Science, 34-35. Beijing: China Statistical Press, 1985.

393 BMA 002-020-00746: 22-23.

394 BMA 002-020-00746: 16-22. The report observed that the actual completion rates were uneven across the country. Among those provinces and cities that had more or less completed the surveys were: Beijing, Tianjin, Shanxi, Inner Mongolia, Jilin, Heilongjiang, Rehe, Shandong, Hunan, and Jiangxi. Another group was in the midst of carrying the survey out, and included Hebei, Liaoning, Shaanxi, Qinghai, Zhejiang, Henan, Hubei, Guangdong, Guangxi, Sichuan, and Jiangsu. A smaller set had begun planning work but not

158

The key method for collecting data, however, remained the periodical report system.395 And by 1956, an integrated statistical system extending down from the SSB in

Beijing allowed for unprecedented levels of penetration. With a staff of about 650 personnel, the SSB, headquartered in Beijing, served as the apex body of this system.

Below it, at the provincial level, statistical offices coordinated work across roughly 2,200 counties. Under the counties were village administrations, which coordinated the statistical work of about 750,000 village cooperatives. Unlike in the Soviet Union, where a similar unified chain of command was free from oversight by other party and government organs, in the People’s Republic a statistical unit at any level was simultaneously also under the authority of the Party Committee at that level. In other words, a provincial bureau took its orders not just from the SSB in Beijing, but was also a constituent part of the provincial People’s Committee, which had a say in its organization.

Thus, the provincial People’s Committee could make its own requests for data to the provincial statistics bureau. A statistical worker at each level usually had more than one set of leaders to answer to.396 Such an arrangement it was hoped would make the statistical system flexible enough to meet local and national needs.397 But it also generated problems, as the following section will show.

the survey itself: Gansu, Anhui, Fujian, Guizhou, Yunnan, and Shanghai. Xinjiang was the lone region that had yet to even begin making preparations for the survey.

395 “Among all the various methods of investigation, the complete enumeration periodical report system is the basic and [most] important method.” [在各种统计调查方式里，全面定期统计报表是基本的主要的方式] See: Wu Hui 1985: 32.

396 Xue Muqiao 1957: 8.

397 A description of this system can be found in: Xu Qian and Lin Fude 1956: 18-19.

159

State Statistics Bureau (Beijing)

Provincial Bureau Provincial Bureau Provincial Bureau

County Office County Office County Office County Office County Office County Office

Vil Vil Vil Vil Vil Vil Vil Vil Vil Vil Vil Vil Vil Vil Vil Vil Vil Vil lag lag lag lag lag lag lag lag lag lag lag lag lag lag lag lag lag lag e e e e e e e e e e e e e e e e e e Co Co Co Co Co Co Co Co Co Co Co Co Co Co Co Co Co Co op op op op op op op op op op op op op op op op op op

Given the goals of extensiveness and completeness (as identified by Wang Sihua), the management of reports was a major challenge for the SSB. At the Second National

Meeting on Statistical Work in December 1952 the emphasis had been on stressing united leadership 统一领导, which gave all decision-making authority regarding forms and schedules to the SSB in Beijing. The result was the proliferation of forms such as Annual

Reports for which the State Statistics Bureau in Beijing was the sole recipient. From the center’s perspective, this was a world of neatness, simplicity, and legibility. All data would be collected via exhaustive enumeration at the level of the enterprise and received via successive rounds of upward reporting and collation as they traveled from the unit 单

位 and county level up to provincial bureau and eventually to the SSB.

160

Let us consider the example of “The Basic Annual Report for State-owned, Local

State-owned, and Private-Public Large Scale Industrial Enterprises for the year 1953,” which encapsulates this idealized vision.398 Below is the summary sheet from the Report.

Several centralizing and comprehensive features, meant to make an industrial enterprise

‘legible’ to the SSB are readily visible: its basic identifying attributes such as ID number and geographic and administrative location, industry division, key products produced, total labor input, and gross value of industrial output for 1953 and 1952 at current and fixed prices. The summary page was followed by 12 additional pages which sought further details on various activities at the industrial enterprise:

398 BMA 018-011-00054: 80-91. Basic Annual Report for State-owned, Local State-owned, and Private- Public Large Scale Industrial Enterprises for the year 1953 [一九五三年国营、地方国营及公私合营大型工业企业基本年报].

161

Page 2: Total Output Value and Commodity Output Value 总产值及商品产值 Page 3: Output (quantity) of Principle Products 主要产品产量 Page 4: Labor and Wages 职工人数及工资 Page 5: Training of new workers and improvement of workers’ techniques 新工人培养及工人技术提高情况 Page 6: Production, Allocation, and Balances of key materials 主要物资的生产、拨出与结存 Page 7: Receipt, Outlay, and Balance of Key Raw Materials (semi-manufactured goods) and Fuel 主要原材料（半成品）及燃料收拨与结存 Page 8: Consumption of Main Raw Materials (semi-finished goods) and Fuel 最主要原材料（半成品）燃料及电力的消费情况 Page 9: MISSING Page 10: Power Plant 电力设备 Page 11: Metal Cutting Machine Tools and Forge Equipment 金属切削机床及锻压设备 Page 12: Dedicated Railways and Main Transportation Tools 企业专用铁路及主要运输工具 Page 13: Housing and Educational Welfare Facilities of Workers 企业职工住宅及文教福利设施情况

By collecting Annual Reports from all industrial units in this way, the SSB hoped to capture the total reality of industrial production in the entire country.

In order to arrive at the neatness of the Annual Report described above, each industrial enterprise had to track its activities on a far more frequent level (daily, ten-day, monthly, quarterly) by using a much wider array of reports. Given the goals of extensiveness and completeness (as described above), the management of reports was thus a major challenge for the SSB. It soon became clear that the kind of focused authority called for in December 1952 had led to numerous contradictions between lower and upper levels within the statistical structure as well as across the same administrative level. Furthermore, as reported by the Henan Statistics Bureau, enthusiasm often ran high

162

during the first year of work. As a result, demands were too high, too many, too urgent, and a general impatience 急躁情绪 characterized all work. This situation was reversed during a national meeting in May 1953. A few years later the national policy of “More,

Better, Quicker, Cheaper” 多、好、快、省 put forward during the 8th Party Congress in

1956 also generated tensions since statistical work often focused on ensuring “Better” and “Quicker” at the expense of “More” and “Cheaper”.399 A similar flip-flopping between simplification and complication was also reported by the Jiangxi Statistics

Bureau. As an example, it cited the case of commercial statistics in the province, which had gone through four instances of simplification and complication during the five years between 1953 and 1957.400

It was in the field of agricultural statistics, however, that the SSB would meet its most intractable foe. For much of the decade, agriculture occupied a secondary, subsidiary position to industry within the economy. As Carl Riskin has noted, at the

“most general level, the principal themes of the [first five-year] Plan did not include agricultural production.”401 And yet, it was the surplus generated in agriculture that was an essential source of investment needed to drive growth in heavy industries. The efficacy of the periodical reporting system as a means of assessing agricultural output, and thus the surplus, was however constantly undermined by the scale of the agricultural sector, the variability of terrain and produce, seasonal differences, and a lack of

399 BMA 133-001-00091: 16.

400 BMA 133-001-00091: 21. In its own report, Henan province observed that the simplification was successful to varying degrees based on the level of administration. For instance, the simplification at the county level and above was quite successful. Below the county level, especially at the village level, the reverse was true. See: BMA 133-001-00091: 11.

401 Riskin 1987: 56.

163

manpower. Take for instance a March 1955 telegram delivered to planning and statistical bureaus across the nation in which Xue Muqiao bemoaned the quality of the 1954 Annual

Reports for agriculture. Xue observed that when compared with figures from earlier years, both the yield and area sown showed massive upward and downward fluctuations for agricultural staples such as grains, cotton, oilseeds, peanuts, and sesame. He concluded that “such large volatility 波动幅度 in area sown and in the production figures for various crops amply illustrated that statistical work in agriculture still lacked a solid foundation and was badly in need of strengthening.”402 Even the use of typical sampling

典型调查, which was meant to supplement the periodical report system, generated problems since most cadre were unfamiliar with what specific methods to use, how to select typical cases, and other basic principles.403

Part of the problem lay in a lack of personnel. Unlike in industrial units, many villages and cooperatives did not possess a dedicated statistician. And even if they did, this person, in many cases, could only work part-time 兼做. The actual duties of statistical work often fell to the village clerk 乡文书, whose regular duties became increasingly onerous as life in the countryside went through successive waves of cooperativization and eventual collectivization. In this milieu, some clerks even refused to perform statistical duties, claiming that since it was not formally listed under their responsibilities it was an ‘additional burden’ 额外负担, which they could or could not

402 BMA 002-020-00746: 9-11.

403 BMA 133-001-00091: 14.

164

perform 可做可不做.404 Already by the mid-1950s, agricultural statistics had thus become a particular source of embarrassment for the SSB.

Even as a constant battle was fought to standardize and simplify the overall numbers and types of forms, there was a growing realization that the scope of data collected had to be expanded. As a report on industrial statistics poignantly noted of the state of statistical work in1957: “thus [we are] able to reflect production, but [can only offer] a poor reflection of livelihood.”405 That same year, Xue echoed the report’s sentiments, stating: “in the past we spent more time thinking about production and construction statistics, and did not emphasize [enough] surveys of people’s lives.”406

Indeed, already by the end of 1954, this need to better understand changes (and improvements) in people’s livelihood had precipitated a turn toward systematically collecting data on rural and industrial family income and expenditure, purchasing power, wages, population, prices, and much else previously outside the purview of the SSB.407

Exhaustive enumeration, ill-suited to these tasks, had to be augmented by newer, more sophisticated representative methods. Coupled with the challenges already being faced due to the re-elevated status of agriculture, where exhaustive enumeration via the periodical report system had already proven ineffective, 1955 and 1956 thus witnessed a broader engagement with non-exhaustive forms of counting—in other words, with survey

404 BMA 133-001-00091: 20.

405 “因而能反映生产，不好反映生活” See: BMA 063-007-00088: 5.

406 See: BMA133-0010090: 85.

407 See: BMA 133-001-00091: 13-14, 17 [河南省]. On the 1954 Agricultural Household Survey, see: DSJ: 35.

165

sampling.408 Even so, per the dictum of centralized leadership and dispersed management, the SSB could only offer regulations and suggestions on how to conduct work. 409 The actual work had to be carried out by local units at the provincial, city, county, and village level and be coordinated by various bureaus at each level. 410

A household income survey that the Beijing Statistics Bureau planned to carry out starting in 1955 reflects these themes of expanded scope, new methods, and coordination.411 As a clear instance of an attempt to carry out a city-level typical sampling survey 典型调查, it is indicative of the kind of coordination that was required.

The directive called for the coordination of several different units 单位; the SSB’s role was one of providing guidelines and not of carrying out all the work, highlighting again the limits of actual SSB control. In earlier years, smaller-scale surveys exploring some of these questions had been attempted by individual bureaus. In 1951, the Labour Bureau had organized about fifty people to survey more than 200 workers families; and in 1952, the City Finance Committee had organized over 120 people to survey 700 households.

None of these were regular surveys and the surveyors’ lack of experience meant that the surveys were poorly executed and their data deemed unreliable.

In December 1954, the City Statistics Bureau offered guidelines to the City

Finance Committee concerning setting up a Resident’s Family Livelihood Survey 居民家

408 Xue Muqiao 1957: 1-2; BMA 133-001-00091: 13-14.

409 A 1953 directive had put SSB in charge of all statistical regulations. See: Xue Muqiao 1957: 9.

410 BMA 133-001-00090: 84. For an example of the SSB sending out guidelines for conducting a particular set of surveys, see: BMA 002-020-00746: 22-23.

411 Starting in 1955, there was a nationwide campaign to collect such data. This was the kind of more intrusive but also complicated statistical activity that the SSB (and the State) started to conduct during the first five year plan. See: DSJ: 35.

166

计调查 for 1955.412 The purpose of the survey was four-fold: first, to understand the material and cultural living standards of the people; second, to organize and plan the circulation of goods; third, to plan for the circulation of currency; and fourth, to prepare indices of worker’s cost of living. Regular surveys would thus measure the income and expenditure (cash and kind) of every level of society, their purchasing power, changes in the rate of consumption of major commodities, changes in disposable income and in savings.413 These new types of statistical data were to be collected across all members of society, who were classed into four groups: workers, independent workers, private industrial and commercial workers, and peasants and other grassroots people.414

If the survey had already indicated an expansion of the types of data to be collected, then the method for the survey was also a departure from the exhaustive enumeration of the periodical report. Typical or purposive sampling 典型调查 was a method of sampling that relied on the selection and exhaustive surveying of a representative handful of sample units. In order to ensure that the samples selected were indeed representative, the guidelines suggested using the method of proportional

412 BMA 004-010-00805: 1-3. In a letter dated December 31 1954, the City Economics and Finance Committee (财经委员会) agreed with the Beijing Statistics Bureau’s suggestions and began preparations for setting set up a Resident’s Family Livelihood Account 居民家计帐.

413 BMA 004-010-00805: 4-9, “关于建立居民家计账的初步意见;” the rest of the discussion on the survey draws from this document.

414 The following definitions were used: Workers: People at various Agencies and Organizations; Cultural & Public Health Organizations; State- owned enterprises; and at Private Enterprises. Independent workers: Individual Handicrafts Makers; Cycle-rickshaw drivers; Laborers; and Mental Workers [white collar]. Private Businesses: Shopkeepers; Peddlers; and other relatively stable industries or those that had already been flagged as possessing characteristics of state capitalism. Agricultural Households: Distinguished by whether they were members of a cooperative or not, and by crop grown (land, cotton, vegetables, etc.). See: BMA 004-010-00805: 5.

167

sampling. In other words, urban and rural residents were separated into various classes based on certain characteristics and on income. Once this was done, the relative weight of each class in the total population was used to establish the numbers from each class within the sample of typical households in the survey. In concrete terms, this meant the

250 households to be surveyed were divided into four groups: a little over 170 were classified as workers (Group 1), a little over twenty as independent workers (Group 2), twenty-five as private businesses (Group 3), and another twenty-five as agricultural

(Group 4).

A general lack of experience with such survey work led to a cautious, staggered implementation scheme. By January 1 1955 all preparations had to have been completed for surveying Group 1; the remaining three groups would wait till April 1.415 Furthermore, different methods were recommended for surveying urban and rural households: urban families would fill the form with the help of a surveyor, whereas rural families would be interviewed and the surveyor would himself or herself fill in the form.416 In order to preserve labor capacity 节省力量, typical households belonging to Group 1 were selected based on regional distributions; for the remaining three groups, typical households were selected at the level of the administrative area. For rural surveys, a surveyor was invariably sent to the household to conduct the survey. The survey was tasked with collecting data across four parameters: family population and property; sources and

415 If this was felt to be too onerous, then Group 4 could be delayed till July 1.

416 The problem of survey design and the need for trained interviewers is a fairly universal problem. D.K. Lieu, writing in 1948 had drawn attention to this problem and the possible solutions that had been employed. See: D.K. Lieu. “Collecting Statistics in China.” The American Statistician 2, no. 6 (December 1948): 12-13.

168

categories of monetary and in-kind income; expenditures (both commodity purchasing power and non-commodity expenses); and lastly, daily cash savings.

Given its wide ambit, the survey also required additional coordination across various government bureaus, vertically and horizontally. Ten bureaus were identified and asked to send personnel to form a central team and a survey team. The central team’s job was to guide the survey team. It was also supposed to periodically call meetings to discuss problems, and to coordinate work across bureaus, get help in selecting typical households, and in propagandizing the survey. The total combined strength of both teams was 58 persons, with their composition outlined in Table 2 below:

Table 3 City Bureau/Unit Central Survey Planning Committee 1 1 Stats Bureau 1 1 Commerce Bureau 1 12 People’s Bank (Beijing branch) 1 14 Cooperatives 1 12 Grain Bureau 1 3 Labor Bureau 1 2 Federation of Trade Unions 1 1 Women’s Federations 1 1 PSB 1 1 Total 10 48

Preparatory work for the survey had to be completed by December 1954 and involved the preparation of account books, making announcements and spreading information about the survey, selecting typical households, training surveyors, and so on. Twenty million yuan 元 were allocated for the survey, and included a special outlay for rewarding households being surveyed. The instructions noted that based on experience in other cities, such awards were necessary for the success of the survey.417

417 The 20 million was further divided into the following categories: Cost of account books: Each book estimate at: 3,000 元; Total 3,000 books; Total cost: 9 million 元 (900 万元)

169

Reconciliation/Correction 整理

By the mid-1950s, collection of statistical data had expanded to include heretofore neglected subjects and also begun to employ methods that required the development of capacity and expertise in new areas such as systematized survey sampling. Ensuring the quality of this data proved to be just as daunting a task as the actual collection itself. At each annual statistical meeting, reports on measuring quality were shared and discussed, but the SSB also intermittently approached provincial bureaus in order to obtain a more real-time sense of the problems of organization and administration in statistical work.418

For instance, a telegram from the Northern Region Statistics Bureau (华北统计局) sent to the statistics bureaus of Hebei, Shanxi, Suiyuan 绥远, Beijing, and Tianjin in June 1953 sought responses to the following three questions:

1. What kind of data did the bureaus provide to local leaders? How was this data supplied? (Report not just successful experience but also point out problems and their possible solutions).

2. What other data do local leaders ask for? What is it [currently] possible to supply?

3. Among data currently being supplied, what kinds are regular/periodic and what kinds are of a temporary 临时 nature?419

As a result of numerous such inquiries and the debate and discussion they generated, we can classify problems of quality into three inter-related categories:

Cost of awards/meetings: Per HH/per year: 30,000 元 (三万元); Total HHs: 250 Total cost: 7.5 million 元 (七五 0 万元) Other office expenses: Included preparatory work: 3.5 million 元 (三五 0 万元) See: BMA 004-010-00805: 4-9.

418 This was in addition to the reports they received during the annual National Meetings on Statistical Work, of where there were a total of six: the first in July 1951 and the last in September 1957.

419 See: BMA 002-020-00670: 32. Another telegram asked for provincial and city bureaus to report problems and their experiences in cleaning up false (illegal) reports. See BMA 002-020-00670: 30. In similar fashion, a Provision/Regulation 规定 called for the coordination of planning and statistical work. See: BMA 004-010-00255.

170

1. Over-production/Duplication/Chaotic reporting 滥发报表 2. Reconciling statistical and accounting data 统计、会计不一致 3. Delays

Overproduction/Duplication/Chaotic Reporting Overproduction and duplication of forms, which were prepared to meet the needs of various claimants, was a key difficulty faced across all levels of the statistical administration. A survey of a Glassworks and a Machine Factory carried out in 1957 by the Nanjing Statistics Bureau offers an illustration of this range and the duplication of forms and data that resulted. Table 3 shows the results of the survey.420

Table 3. Excess Forms at Two Nanjing Factories Nanjing Glassworks First Machine Factory % of % of Forms Proportion total Forms Proportion total Total 351 - 100% 242 - 100% Used within the Industry 192 100 55% 107 100 44% Used at the factory level 78 40.6 22% 47 43.9 19% For outside submission 159 100 45% 135 100 56% Planning 34 21.4 10% 32 23.7 13% Financial Accounts 50 31.4 14% 24 17.8 10% Statistics 66 41.6 19% 63 46.6 26% Technical Accounts 9 5.6 3% 16 11.9 7%

At both the Nanjing Glassworks and at the First Machine Factory, not only were too many reports generated, but they were also generated for several different destinations/clients—the statistics bureau was only one among many destinations. In each factory, a sizable number of forms, roughly half (192 out of 351, or 55%, for the

420 In arriving at the count, the same type of form was counted twice if it was used to make monthly and quarterly reports. The table is a faithful reproduction except for the two ‘% of total’ columns, which have been added and were not part of the original table produced by the Nanjing Statistics Bureau. See: BMA 133-001-00091: 57-58.

171

Glassworks; 107 out of 242, or 44%, for the Machine Factory) were produced for internal use. Among these, a further distinction was drawn between those used at the factory/shop level and those used at the industry/unit 单位 level. The remainder was produced for outside submission. Among them, those sent to the statistics’ offices accounted for 41.6% and 46.6% of the total respectively. The corresponding shares out of the total forms for the statistics bureau thus were 19% (66 out of 351) and 26% (63 out of 242). The share of forms actually issued by the SSB in Beijing accounted for an even smaller number: 12% and 14% respectively. 421 As a result, the survey noted, the bulk of the requests constituted a duplication of data.422 On the basis of such surveys, provincial bureaus requested that the center clarify and standardize only those reports it needed reporting up and not try and standardize all reports. The other tables could then be based on local requirements without having to conform to a national standard.

Attempts at simplification, touched upon in the previous section on collection, were also often ineffective. One 1957 survey of three bureaus—Education, Transport, and Public Health—in Jiangxi observed that two out of the three had failed to simplify and had on the contrary increased their output of forms. In another survey, it was found that twenty-two offices and organs spread across eight industrial enterprises produced

421 More numerical and textual examples in: Si Gengsheng 司更生. “Fanlan chengzai de diaocha tongji biao 泛滥成灾的调查统计表 (Swamped by Statistical Tables).” RMRB, September 12 1953, 3.

422 That the SSB was particularly mindful of this duplication of data is evident from the interview of a statistical worker from a county-level Industrial and Mining Bureau who Li Choh-Ming interviewed in August 1961. Sometimes such duplicate data was based on genuine need. At other times, Li reported Wang commenting “that some agencies liked to obtain all types of information regardless of whether the information would be of use to them or not.” Li also noted in his report that “Without prior permission from the bureau, other government agencies and offices were not allowed to send investigators to the factories and mines under the bureau’s control for the purpose of collecting statistical information.” See Li Choh-Ming. ““Statistics and Planning at the Hsien Level in Communist China.” The China Quarterly, No. 9 (Jan. – Mar., 1962b): 112-123. (quoted text is from p. 119). Li also described a system of introductory letters that the bureau issued to other agencies when they required data.

172

477 types of forms and 25,771 indicators for outside submission. Many of these forms were extremely complicated 繁杂 and some were also duplications of each other.423

Made repeatedly clear through such surveys was the point that what was of relevance to local agencies was not necessarily of relevance to the center.424

The picture on the ground thus looked quite different from that envisioned by the

SSB in Beijing. Every administrative level complained of a flood of papers and numbers, and little capacity to make any sense of them. Excess issuing 滥发 and chaotic issuing 乱

发 were the terms frequently used to describe such a state of affairs. The SSB tried to tackle this problem in September 1953 by going public, as it were. Prior to that, discussions about excess issuing and duplication had circulated internally via memos, telegrams, and reports. But on September 7 the People’s Daily published two directives that addressed the issue head-on.425 The first of these, “Instructions regarding the cleaning-up of existing statistical survey forms and [concerning] the cessation of their reckless issuing,” was issued by the State Administrative Council (SAC) and carried the authorial imprimatur of Zhou Enlai himself. The second was an SSB document that had been vetted by the SAC, titled “Temporary Methods to formulate and approve statistical forms.”426 In addition, the People’s Daily also published an editorial on September 7

423 BMA 133-001-00091: 27.

424 See for instance: BMA 133-001-00091: 9. Also see: Si Gengsheng 1953: 3. Si provides an exposition of the kind of detailed data collected by local bureaus.

425 See for instance, BMA 133-001-00091; BMA 008-011-00054; BMA 002-020-00670: 28, 30-32; BMA 002-005-00088; and BMA 133-001-00043. Also see: DSJ: 23 (Item 59) and Xue Muqiao 1957: 18.

426 RMRB 7 September 1953, 1, 2. Both documents were actually released 发布 on September 5. See: BMA 123-001-00416: 47.

173

1953 titled “Wipe out chaotic phenomena from statistical survey work.”427 These three documents together served as a rallying cry for reducing duplication and mitigating chaos.

As a Beijing Statistics Bureau notice 通知 from a couple of weeks later declared:

These three documents are extremely important for current statistical work, [and it is] hoped that statistical bureaus and offices at all levels will organize study sessions for all statistical cadre, [so that the instructions] may be earnestly put into practice.428

In this manner, solutions continued to be sought so that a gradual decrease in excess and chaotic issuing could effected.429 In some instances, such as instructions for statistical work for the year 1954, documents even contained claims that the “serious problem of excess reports had by and large been conquered.”430 And yet, by 1956, hardly any long- term improvement had occurred. On the contrary, as the SSB expanded data collection into new areas, the People’s Daily could only bemoan that their still excessive numbers, complicated content, constant emendations, and urgency, all meant that it was finally

“time to reduce statistical reports.”431 Indeed, things were so serious that internal memoranda from a year later deemed reports useless as soon as they were produced 随发

427 RMRB, 7 September 1953, 2. Another article, published five days later on September 12, provided numerous examples from across the provinces and from various sectors of the economy to bolster the claim of “too many, too complicated, [and] too chaotic.” Si Gengsheng 1953: 3.

428 BMA 123-001-00416: 47.

429 BMA 133-001-00091: 11.

430 BMA 004-010-00807: 29.

431 Jiang Ling 江陵. “Zhonglei duo, xiangmu fan, biangeng duo, yaoqiu ji shi jianshao tongji baobiao de shouhou le 种类多、项目繁、变更多、要求急是减少统计报表的售后了.” RMRB 25 October 1956 , 4.

174

随废.432 Unless action was taken, such an unchecked spread would, it was felt, lead to unmitigated disaster 泛滥成灾.433

The possibility of unmitigated disasters was very much on Xue Muqiao’s mind in

September 1957 when he observed that the problem of excess issuing remained unresolved. Given the demanding nature of exhaustive counting (investigations) 全面调

查化, he insisted that only the SSB should issue demands for periodical reports. If other levels of the statistical system needed their own data, they should resort to one-time surveys, typical surveys, or annual reports, and not monthly or quarterly reports.434

According to him, three principal reasons were behind the problem: micro-management, the twin tendencies of subjectivism 主观主义 and bureaucratism 官僚主义, and not enough oversight by the statistical organs. Looking ahead to the second five-year plan he exhorted the Planning Bureau, the Finance Bureau, and the Statistics Bureau to join together and combat the problem in one go.435

Reconciling Data 核对、整理

A second major problem was the presence of wildly disparate numbers for putatively the same observation or activity. The need to have reconciled data was particularly important in a fast expanding economy undergoing comprehensive nationalization, since terms and

432 BMA 133-001-00091: 10

433 BMA 133-001-00091: 30.

434 Xue Muqiao 1957: 9.

435 Xue Muqiao 1957: 18-19. Resorting to such sloganized reasons was of course hardly new and can be traced to the early 1950s and the campaigns against bureaucratism. Also, Mao’s “Wu duo 五多” essay raised similar concerns. See: Mao Zedong. “Jiejue wuduo wenti 解决‘五多’问题.” In Mao Tse-Tung. Selected Works of Mao Tse-Tung—Vol. V. Beijing: People’s Press, 1977, 77-79 (p. 89-81 in English edition).

175

categories were constantly in a state of flux. Wang Sihua had touted standardization as one of socialist statistics’ key characteristics, but that ideal too needed to be reconciled with the reality on the ground.

One reason for different numbers can be traced to confusion over what exactly was being measured. In 1953, for instance, a Beijing Statistics Bureau bulletin 通报 (sent to all industrial units in Beijing, and copied to various central government organs) required that all entries for the Annual Report as well as for monthly periodic reports be made according to the Catalog of Industrial Products 工业产品目录. This would help avoid the problem of leaky reporting 漏报, which was defined as consisting of not reporting goods produced for internal consumption or goods that factories considered unimportant. Similarly, sometimes there was confusion over units of measurement: should one report tons 吨 or bags 袋?436 Even four years later, ensuring that production was reported according to the Catalog of Industrial Products remained a challenge.437

But the challenges of standardization also reflected a more deep-rooted and systemic problem. In the earlier section on overproduction we have already seen that data was collected by numerous agencies. In order to ensure data was properly reconciled, two systems of reporting were used: the single track 单轨制 and the double track 双轨制.

The former system was used for data in fields where already a high level of professionalization existed, such as rail and air transportation, postal services, education, and public health. The business offices of these services were expected to directly report

436 BMA 008-011-00054: 41-42.

437 BMA 063-001-00088: 5. By this time, the catalog consisted of about 230 items: over 180 intermediate goods and over 40 consumer goods.

176

statistical data to statistical bureaus at the same level and also to make reports to their own higher levels.438 In contrast, the dual track system was used for the more important and much larger types of data, such as those relating to industrial and agricultural production, fixed capital investment, retail sales, and labor and wages. According to this system (see chart below) statistical and business data were collected independently; the former being reported through the statistical apparatus and the latter through the particular ministry concerned. The two sets of data thus produced were then compared to check the overall quality of data, the normative expectation being that the data would be the same.439

State Statistics Bureau Ministry 国家统计局主管部

Provincial (Self- Provincial (Self- administered region, city) administered region, city) Statistics Bureau Business Office 省(自治区, 直辖市) 统计局省(自治区, 直辖市) 业务厅

County Statistics Bureau County Business Office 先统计局县业务主管局

Grass-roots Unit 基层单位

438 Wu Hui 1985: 32.

439 Original schematic chart in: Wu Hui 1985: 33.

177

In practice, however, because of varying conditions of work and administration, the procedures of data collection were rarely standardized. The natural result was inconsistent data. In matter of fact, in most cases a third source of data also existed: the finance bureau also collected accounting data.440 But instead of solving the problem, this merely exacerbated inconsistencies across three sets of numbers. In other words, inconsistencies across statistics, accounting, and business reporting constantly brought into question the quality and reliability of all data being reported. The archives abound with bulletins, regulations, suggestions, and instructions that all attempt to grapple with this problem.441 Many of these discussions stressed the generally superior reliability of statistical data, observing that accounting or business reporting was relatively vague because it was prone to variations on account of local conditions and local practices.442

Yet others held the exact opposite view, asking not just for the abolishment of the dual track system but also for the removal of local level statistical units and a sole reliance on business/industry reports.443 The problem was severe enough that on June 20 1956 the

SSB and the Ministry of Finance jointly released “Common Responses to questions about

440 Dual-tracking to control for quality of data was likely used in other settings as well. For instance, a related dual-track system was used at the Beijing Pharmaceutical Company 北京医药公司 in 1956. The two tracks in this case were the statistical and the fiscal (accounting). Receipts 发票 was recorded and then passed on by the business clerk 业务员 to the statistics division 统计部门 and to the accounting division 会计部门. The statistics division then prepared an original voucher 原始凭证 while their accounting counterpart prepared a goods account 商品账. Each of these datum was then reconciled against the other on a daily, 10-day, and monthly cycle 统会双方按日旬月核对数字. See: BMA 35-2-203: 35. Another example (with chart) for Public-Private Commercial Enterprises, Catering Services 饮食业, etc., can be found in: BMA 133-001-00008: 67.

441 See, for instance: BMA 119-001-00118: 76-80; 133-001-00065; 004-010-00255; 004-016-00268; 133- 001-0091:32); and 004-010-00618.

442 See for instance, BMA 133-001-00065: 20.

443 For instance, BMA 0133-001-00091: 24.

178

inconsistencies in accounting and statistics,” which it was hoped would provide guidelines to begin addressing the problem.444

Delays

Excess issuing, duplication, and the need to reconcile data all also contributed to another chronic problem: delays. A case in point is the 1954 Agricultural Production Annual

Report. In order for the national version of this report to have been prepared in a timely manner, provincial reports, themselves compiled on the basis of periodical reports, should have been received by the end of January. But as late as 22February, only sixteen out of twenty-nine provinces and cities had submitted their reports.445 In similar fashion a 1955 notice from Xue Muqiao requested that all provincial and city level bureaus undertake reconciling work in a timely manner, observing that only a handful of provinces had actually begun such work, with most others lagging behind.446

Unclear procedures and confused systems of measurement also contributed to delays. For example, when reporting Commercial Statistics, the SSB regulations called for all forms to be filed thirty days after the end of the quarter. But the Commerce

Bureau’s regulations called for the reporting to be done within twenty-six days. And the

Finance Bureau, in turn, demanded them within forty-five days (though in reality it took sixty days).447 Such multiple deadlines did little to ensure the timeliness for any of the bureaus concerned.

444 DSJ: 46-47.

445 BMA 002-020-00746: 9.

446 BMA 002-020-00746: 24.

447 BMA 133-001-00091: 18-19.

179

Much as in collection, the agricultural sector again offered the greatest challenge for the reconciliation of data. We have already noted agriculture’s secondary position within the economy for much of the 1950s. In 1953 Mao had observed:

In the present transition period we cannot introduce unified and planned production in agriculture, except on the state farms, and interfere too much with the peasants. We can only guide agricultural production and coordinate it with industrial production by our pricing policy and necessary and practicable economic and political work so that it can be integrated into the national economic plan.448

It was as a result of such views, which acknowledged that agricultural sector could not be administered with the same degree of control as other parts of the economy, that some proposed making the county level planning and statistics office the main unit for reporting agricultural data. The reason behind such a suggestion lay in recognizing that it was the county level at which local conditions and requirements became particular enough to warrant local control.449 An even more drastic scheme suggested transferring the responsibility of agricultural statistical work to the Agriculture Bureau.450 But this was tempered by other less radical points of view, which recommended that the SSB and the Agriculture Bureau work together and reach a consensus on methods, requirements, and management.451

448 Mao Zedong 1977 (English edition): 91.

449 BMA 133-001-00091: 2.

450 BMA 133-001-00091: 45-46.

451 BMA 133-001-00091: 70-71, 75.

180

Solutions

By 1957, different solutions were offered for each of the different levels of government: provincial, city, county, and village. At the Province Level it was felt that neither a system that granted too much centralized control to the province nor one which let authority devolve entirely down to the various departments was appropriate.452 The problem at the provincial levels included limited manpower and an unfamiliarity with the work of the various departments, which resulted in much ado for very little result 忙乱尽

缓 and could not help mitigate the chaotic issuing of reports 乱发报表. If, on the other hand, authority to issue and manage reports was distributed to individual departments, then on account of their numbers and variety, a chaotic situation 混乱现象 was also unavoidable. The solution, as proposed by the Henan Statistics Bureau, was for the SSB and the various departments to issue a common set of periodical forms, which had to be used by everyone.453

At the City and Special 专 Administrative level, it was felt that forms should only be issued after all city-level statistical departments had come to an agreement. This meant granting significant autonomy to city level statistics bureaus. The logic to doing this lay in the fact that many city level governments were close to and familiar with local conditions and hence had a better sense of the ground realities. Therefore, if their expertise and experience was not brought into play, chaotic issuing 乱发报表 of reports could easily result. At the same time, if certain departments had a strong foundation in

452 This was, of course not the only opinion. In a federalist move, some called for greater power for the provinces, while others argued for increased centralization. See BMA 133-001-00091: 34, 41, and 26-27. For other solutions, see: BMA 133-001-00091: 55-56 (Nanjing); 58-61 (simplification).

453 BMA 133-001-00091: 9.

181

statistical work, then they could be permitted to issue reports themselves, as long as periodic checks were still performed.454

At the County Level, one-time reports/surveys usually outnumbered periodic reports. Since county offices were frequently far from the provincial office, they exercised greater freedom and could respond more freely to short-term requirements for data, issuing new forms per need 随用随发. This reality, and the fact that time was always in short supply, meant that most forms reported from the county level up to the province level could not be approved. One solution proposed was to give the authority to approve one-time survey results to the county office, and require that only periodic reports be submitted for approval to the provincial office. As for the village level, most provinces felt that statistical work at this level was poorly administered. Villages were too many and tables and reports were generated in an erratic fashion. Frequently, most reports were useless the moment they were produced 随发随废. The majority did not have a shelf life of more than one day and rarely made it to the County Level for approval.

Accordingly, even as late as 1957, the pressing need was to better oversee village level statistical work by introducing a simplified periodical reporting system and do away with the chaotic collection and issuing of reports. 455

By the mid-1950s, proposal such as these had led Xue Muqiao to promote the general policy of centralized leadership but dispersed management 统一领导，分级管理.

454 BMA 133-001-00091: 10.

455 BMA 133-001-00091: 10.

182

Research

In his report at the Sixth Annual Meeting of Statistical Work in Beijing in September

1957, Xue Muqiao had explained that the people’s economy was extremely complicated, and the result of the intertwining of a variety of intricate factors. It was the task of statistics and the statistician to explain and simplify this extremely complicated world by providing sound research and analysis of statistical data. Research, thus, was the third major service that statistics had to perform. The most important element of such research and analysis was the production of statistical indicators. As Xue noted:

When conducting investigations 调查, in order to make calculations easier we must simplify complicated phenomena by converting them into abstract [but easier to understand] statistical indicators.456

The calculation of indicators lay at the heart of most of kinds of analysis, ranging from studies of comprehensively balancing various sectors of the economy, to calculating the relations between different factors , to increasing savings and measuring the successful implementation of plans.457 In general, though, such work had been neglected because the primary emphasis had been on establishing the basic framework for collecting and collating data.458 As the Anhui Statistics Bureau observed in 1957:

Over the past few years, our work in this area [analysis] is extremely inadequate. Our statistics bureau has done a much better job than other bureaus of mastering [the collection of] abundant data, but we have not put this data to good use.459

456 Xue Muqiao 1957: 7.

457 See for instance, BMA 133-001-00091: 12, 35-36.

458 BMA 133-001-00091: 52.

459 BMA 133-001-00091: 35.

183

One reason for this was that given the overwhelmingly large numbers and types of forms and tables, most bureaus had no time for analysis.460

Until about 1955, most research and analysis was left to the SSB in Beijing, to whom provincial and lower level bureaus supplied Annual Reports. The SSB, in turn, collated and analyzed the data to produce annual bulletins 公报 that summarized economic growth and progress in plan implementation. These bulletins sought to achieve as comprehensive a picture of the economy as possible and included information in absolute and relative terms about changes in industry, agriculture, livestock, infrastructure, transportation, communication, investment, commerce and trade, prices, labor, education, publications, and public health.461

By the mid-1950s two additional developments had taken place that expanded the capacity to carry out research and analysis. The first of these was the establishment of a small scale computing facility at SSB headquarters. This foray into mechanized counting permitted the SSB to collectively summarize data from annual reports and censuses.462

By 1957, Xue Muqiao was able to encourage those provinces where statistical work was particularly onerous to also begin considering the gradual mechanization of their

460 As observed by the Henan Statistics Bureau. BMA 133-001-00091: 12

461 See for instance: 1953: 《关于一九五三年度国民经济发展和国家计划执行结果的公报》; 1954: 《关于一九五四年度国民经济发展和国家计划执行结果的公报》; 1956: 《关于 1956 年度国民经济计划执行结果的公报》; 1953-57: 《关于发展国民经济的第一个五年 (1953 年到 1957 年) 计划执行结果的公报》; and 1958: 《国家统计局关于 1958 年国民经济发展情况的公报》 All published by the Statistical Press.

462 In 1958, the SSB published a book on calculating machines: State Statistical Bureau. Jisuanji de susuanfa 计算机的速算法 (The Methods of Rapid Calculations using Calculating Machines). Beijing: Statistical Press, 1958.

184

calculation work.463 The mid-1950s also witnessed the creation of dedicated research sections within provincial and city level bureaus. I have already discussed the household livelihood survey, something which these new research sections became responsible for.

Now let us look at another example of the type of research that began to done by more local level units.

In 1956, the Beijing Statistics Bureau began producing and circulating statistical analysis briefings 統計分析簡報.464 Each issue was a few pages long and offered basic analysis of recent trends for a particular sector of the economy. In all, these briefings were organized under eight categories: summary, capital construction, commerce, industry, agriculture, communications (including transportation), [consumer] goods, and labor. Topics ranged from tracking agricultural production and urban incomes, forecasts of plan completion in various bureaus during the first quarter of 1956, construction work completed during the first half of 1956, the arrival of capital construction equipment for local industries during the first half of 1956, the result of attempts to reduce costs, the ahead-of-time attainment of targets for the sale of retail goods, etc. On the whole, the analysis drew upon actual figures and simple indicators of growth, along the lines of what was discussed in Part I Chapter III.

Each briefing was sent to two sets of recipients. The primary recipients included various city level leaders and related offices, such as the city committee and its general office, and the city planning commission. The reports were also sent to the director of the

463 Xue Muqiao 1957: 7. In interviews, Renmin University professor Ni Jiaxun 倪加勋 recalled using the abacus for all calculations during the early 1950s when he was part of a county level statistical unit.

464 See: BMA 004-010-00253. ( 北京市统计局统计分析简报) The first briefing in this folder is dated April 4 1956 and the last appears to be from October 1956. Altogether there appear to be about fifty bulletins spread across 145 pages.

185

SSB and to the various relevant departments at the SSB. So, for example, the industry and capital construction briefings were sent to the industrial statistic department , the agriculture briefing was sent to the agriculture statistics department, and the commerce briefing was sent to the department of trade statistics. The second set of recipients consisted of those that the briefings were copied to. Here we see a far wider set of agencies, including various ministry level bureaus. Four among them received copies of every kind of briefing, regardless of the category; they were the Bureau of Finance, the

Bureau of Taxation, the Beijing branch of the People’s Bank, and the Supervision Bureau.

By the mid-1950s, there was thus a fairly robust network of agencies sharing interim statistical research on the performance of the economy.

Supply

Supplying data was the last of the four principal duties that SSB and the statistical apparatus were tasked with. While the annual bulletins 公报 noted in the previous section were the most visible examples of the SSB’s work, they were not what comprised the bulk of the data that Xue Muqiao had in mind when he put forward his four-part categorization of SSB activities. Indeed, the bulk of the data and research that the SSB supplied was intended not for the public but only for designated governmental agencies and the leadership at various levels. Many of the reports and briefings cited in this chapter (the reports 报告, the bulletins 简报, analysis 分析 such as various indicators 指

186

标) often carried the stamp “internal issue; not for publication” 内部发行、不得对外发

表.465

It was the supply of this kind of data, which was spoken of as absolutely critical for the planning process that was the final and most important output of the statistical system. And yet, according to Xue Muqiao its fundamental importance was ignored in favor of more frivolous pursuits. In 1957 he noted that “in the past there were some comrades who underestimated the glorious task of supplying data. [Instead] they particularly emphasized the ‘supervisory role’ and ‘vertical integration’ of statistics.”

Xue explained that these cadres went to unreasonable lengths to focus only on the supervisory role of statistics. For Xue this was a gross error, especially when overall statistical capacity still remained rudimentary, when the SSB still struggled to produce reliable data in a timely fashion.466

In the previous three sections we have already seen some of the problems that plagued the accurate and timely supply of data. As we have seen, data was supplied at various levels: frequently between bureaus, to various central agencies, and eventually reported up 上报 to the SSB headquarters in Beijing.467 But the question of who used what kind of data remained: the center often only wanted annual data, but at provincial and local level, monthly and quarterly data were usually more important.468 The result

465 See, for instance: BMA 133-001-00090: 81. Other formulations included: “internal document, outside transmission not permitted” 内部文件，不得外传; see: BMA 004-010-00807:16 .

466 BMA 133-001-00090: 82.

467 See, for instance: BMA 002-020-00670: 12.

468 Xue Muqiao 1957: 9.

187

was to stimulate duplication and chaotic issuing and the competing deadlines also brought about severe delays.469

Conclusion

In the end, for all the valorization of an extensive, complete, and objective statistical system, Xue Muqiao admitted that in certain important situations, the timeliness of the data outranked their accuracy.

Even though statistical data should possess a high degree of accuracy, it is not necessary that all data must be highly accurate [all the time]. In order for the leading authorities 领导机关 to understand the situation, research questions, and decide on policies, they frequently needed reference data on a timelier basis. Such data need not possess a high degree of accuracy or be comprehensive, but it must be supplied in a timely fashion.470

Implicit in this statement was the admission that in most cases data produced was neither timely nor accurate. The archival materials that comprise the principal source base for this chapter suggest why such a bleak picture of statistical work during the 1950s is justified. While they demonstrate a tremendous degree of engagement and professionalization, they also clearly highlight how much a struggle it was to make manifest an idealized system centered on exhaustive enumeration via a periodical report system.

This was not, however, the way statistical work was presented in more public materials. To promote the work they were doing and to keep morale high, the SSB published a whole range of books and pamphlets that sought to share positive experiences.

469 See, for example: BMA 35-2-203: 21-27.

470 The statement was made during his main address at the sixth annual meeting of statistical work in Beijing in September 1957. See: Xue Muqiao 1957: 13; 10.

188

Usually composed of essays written by cadre at different local, regional, and provincial statistical offices, these works had titles such as “Compilation of Experiences of

Grassroots Agricultural Statistics Work,” “Our Country’s Experience in Industrial

Statistics Work,” and “How To Do Analysis of Industrial Statistics.”471 Even though such works extolled the successful experience of selected cadre and statistical offices, read against the grain they only serve to further sharpen the sense we already have of a system bogged down by the weight of its own expectations. At the same time, they also highlight the fact that statistical activity was as much about capacity building as it was about claiming legitimacy; that is, creating the appearance of capacity and ensuring that the

Communist revolution was seen to be a success. This is precisely what the Annual

Bulletins cited above sought to achieve. The 1959 publication of Ten Great Years 《伟大

的十年》 was perhaps the ultimate legitimizing weapon in this regard.

A critical element of both capacity building and claiming legitimacy was training adequate numbers of statistical cadre.472 Given the challenges to statistical work described here, this was a far from straightforward task. It is to that topic that we shall turn to next.

471 See, for instance: State Statistics Bureau. Jiceng nongye tongji gongzuo jingyan huibian 基层农业统计工作经验汇编 (Compilation of Experiences Conducing Grassroots Agricultural Statistics Work). Beijing: China Statistical Press, 1956; State Statistics Bureau. Woguo gongye tongji gongzuo de jingyan 我国工业统计工作的经验(The Experience of Industrial Statistical Work in our Country). Beijing: China Statistical Press, 1958; State Statistics Bureau. Zenyang zuo gongye tongji fenxi gongzuo 怎样做工业统计分析工作 (How to Carry out Analysis of Industrial Statistics). Beijing: China Statistical Press, 1956.

472 Xue Muqiao 1957: 5.

189

CHAPTER VI

To “Ardently Love Statistical Work”: State (In) capacity, Professionalization, and their Discontents

190

Comrade Li Ming Dismisses Statistics

In July 1955, the Tongji gongzuo tongxun (统计工作通讯 Statistical Work Bulletin), the principal journal of statistical work in the People’s Republic, published a short letter from a youthful statistical worker.473 Just twenty years old, Li Ming noted that he had already been engaged in statistical work for three years. Prior to beginning work he had received a year and a half of statistical training at a Statistics and Accounting Vocational School

专科学校. When he began work, Li told the editors of the journal, he was full of enthusiasm. He hoped to wholeheartedly devote all his abilities to statistical work and thereby serve the people. To that end, he spent an additional three months during the latter half of 1954 at the Central Fuel Industry Ministry’s Cadre School 中央燃料工业部

干部学校, supplementing his training. Such training and experience should have led him to achieve success at work. And yet, he lamented that the reality was not so. Quite to the contrary, Li Ming discovered that the more he worked the more he lost faith in statistics and the less he knew where the future career prospects of statistical workers lay. Having already worked for three years, he could not deny that he had a basic grasp of statistical work, but was this not merely the addition and subtraction of numbers or the calculation of percentages 加加减减，算算百分数? After all, he did not directly participate in production work and, as such, was reliant on the information provided by others. Even if he visited the factory floor, there was much about the production process he did not know or understand. Such ignorance about the production process meant that the blind application of statistical methods and tools was not in the least bit useful and rather a

473 Li Ming. “Comrade Li Ming’s Letter.” Tongji gongzuo tongxun, 1955 No. 7, pp. 42. Republished in: State Statistics Bureau. Re’ai women de tongji gongzuo 热爱我们的统计工作 (Ardently Love our Statistical Work). Beijing: China Statistical Press, 1956, 8-9. Herafter RWTG.

191

waste of time 一无所用, 白费时间. So were born his doubts regarding the usefulness of statistical work; doubts so acute that he had reached a point where he was unwilling to do any work.

Statistics’ perceived worthlessness made unambiguous and his own lack of desire powerfully expressed, Li Ming proceeded to note that his frustration was all the greater given his youth. The better part of his life lay ahead of him and he still yearned to make a practical and real contribution 实际的贡献 to the motherland. Statistical work, clearly, did not qualify. And so, he found himself in a dilemma. On the one hand he found himself increasingly unwilling to participate in statistical work; on the other, he could not but recognize the resources the nation had already devoted to his training in statistics.

How could he now let the nation down by seeking other work? Caught between the

Scylla of patriotic desire and the Charybdis of patriotic duty, Li Ming implored the editors of the journal to help him find a solution.

In publishing Li Ming’s letter, a solution is precisely what the editorial team at the journal hoped to unearth. In a short preface to the letter, they explained that the kinds of doubts and general attitudinal problems that Li Ming wrote about were present in a not insignificant number of statistical workers.474 Accordingly, they invited readers to respond with their reactions and suggestions. So was begun a national debate on statistical work. Responses were published over the following several months. The discussion was eventually brought to an end in the December issue of the journal. That issue included a brief overview of the letters that remained unpublished, as well as an

474 For instance, according to a survey carried out in 1954, of the thirty-seven statistical workers in a petroleum factory in the city of Fushun (抚顺), 40% did not pay attention to their work. In another example, statements delivered at the statistical meeting of the Kaifeng (Henan province) group noted that as many as 60% of the workers neglected statistical work. RWTG: 1.

192

essay summarizing the discussion since July and its lessons, written by Jia Qiyun 贾启云, one of the deputy directors of the State Statistics Bureau. The following year, the State

Statistics Bureau (SSB) compiled what it deemed were the best responses into a small booklet for wider dissemination. To those responses they added introductory and summary essays that provided additional didactic content regarding the value and revolutionary nature of statistical work. For the booklet’s title they appropriated the title of Deputy Director Jia’s December article: “Ardently Love our Statistical Work” 热爱我

们的统计工作. Published in August 1956, the booklet had a publication run of 5,500 odd copies.

It is more than likely Li Ming did not exist. He was probably a composite character created to embody the real concerns and problems confronting Chinese planners and statisticians.475 Indeed, the putative Li Ming and the campaign his letter instigated lay bare a crucial anxiety through much of the early PRC—the creation and maintenance of the ideal (statistical) worker. Several forces, each pulling in a different direction, made this an especially challenging task. While some of these forces were a part of the general problem of bureaucratization following revolution, and have been addressed in earlier scholarship476, many others were specific to the nature of statistical

475 Such practice is well known in PRC history. The most notorious example is, of course, Lei Feng. Lei Feng, a young People’s Liberation Army soldier, was at the center of a 1963 PLA orchestrated under the leadership of Lin Biao that sought to “emphasize the basic values of service to the party.” As explained by Jonathan Spence: “the posthumously discovered Diary of Lei Feng emphasized again and again the soldier’s undying love for the revolution, for his country, and for his comrades, as well as his unswerving devotion to Chairman Mao. The fact that the “diary” was fictitious, concocted by the PLA propaganda writers, should not conceal its basic significance, which was to launch an attack against the lack of revolutionary fervor displayed by many intellectuals and writers in the People’s Republic.” See: Jonathan D. Spence. The Search for Modern China. New York: W.W. Norton and Company, 1990, 597.

476 See, for instance: Hong Yung Lee. From Revolutionary Cadres to Party Technocrats in Socialist China. Berkeley: University of California Press, 1991; Leo A. Orleans. Professional Manpower and Education in Communist China. Washington, D.C.: U.S. Government Printing Office, 1961; Franz Schurmann. Ideology

193

work, and to statistics’ role in the economy and what it ‘produced.’ It is the notable lack of reverence for the power and positivism of numbers in Li Ming that sets up the basic contradiction. In Part I we saw that the early 50s were a time when a new theoretical and methodological paradigm for statistics and statistical work was established in the PRC.

But what was one to do when the common worker refused to even acknowledge the transformative power and potential of statistics, far less any putative resolutions to definitional and methodological distinctions? Indeed, on the ground, the concerns were far more basic: the creation of a professional cadre and attendant infrastructure, ensuring adequate number of such cadre, and maintaining their morale; all while keeping in mind the central and fundamental role that statistics and statistical workers were supposed to play in the economy. It is with almost metonymic, and perhaps entirely predictable, precision that Li Ming’s mini-biography and complaints mirror these concerns.

This chapter, then, is about the state’s attempts to build statistical capacity.

Alternatively, it is about how the state dealt with incapacity. My thinking about capacity draws upon Michael Mann’s formulation of infrastructural power. In his seminal 1984 essay he characterized infrastructural power, in contra-distinction to despotic power, as

“the capacity of the state actually to penetrate civil society, and to implement logistically political decisions throughout the realm.”477 Statistics is a powerful example of

and Organization in Communist China. Berkeley: University of California Press, 1968; and Ezra Vogel. “From Revolutionary to Semi-Bureaucrat: The “Regularisation” of Cadres.” The China Quarterly, no. 29 (Jan. – Mar. 1967): 36-60.

477 Mann focused his discussion primarily on capitalist countries, observing that such infrastructural power may have begun by bolstering despotism, but was eventually also used by actors outside the state, including civil society members and powerful capitalist corporations. My interest in this chapter is not to investigate the checks to power that these processes might generate. Regardless of the system of governance, the category of infrastructural capacity has great purchase for 1950s China (and indeed much of earlier Chinese history as well). See: Michael Mann. “The Autonomous Power of the State: Its Origins, Mechanisms and Results.” European Journal of Sociology 25, no. 2, (November 1984): 185-213.

194

infrastructural power. Such power, Mann argued, was a characteristic of modern industrial societies. Accordingly, in this chapter I investigate how the PRC state went about establishing infrastructural power in the realm of statistics, something that the

Republican state arguably lacked, and suffered on account of. In a much more recent working paper Francis Fukuyama, building on Mann’s ideas, has stressed the need to better understand such capacity as part of a larger definition of governance, regardless of the nature of the government in power.478 According to him, one key marker of this capacity is the level of education and professionalization among government officials.479

Thus, if the previous chapter focused on the nature of statistical work, this chapter shifts the spotlight to the creation and maintenance of a cadre capable of doing that work. Our opening story, the ensuing national debate in the TJGZTX, and the eventual publication of

RWTG, all hint at how massively challenging this task was.

A variety of stratagems—training, supplementary training, self-study, motivation, and reward—were employed to tackle these challenges. They began with training in preparation for employment and, in theory, were continually mobilized throughout one’s career. The chapter traces this training and motivation work, using Li Ming as a foil wherever possible. Beginning with his claims of youthfulness and education, the chapter first explores the kinds of statistical training offered by the state. For a select few, such training may have been available at the university level, but the vast majority was trained in technical schools through training courses 训练班. But even this was inadequate and

478 Francis Fukuyama. “What is Governance?” Center for Global Development, Working Paper 314, January 2013, 3. In his introduction, Fukuyama notes about the field of political science and policy studies: “everyone is interested in studying political institutions that limit or check power—democratic accountability and rule of law—but very few people pay attention to the institution that accumulates and uses power, the state.”

479 Fukuyama 2013: 7.

195

by the middle of the decade the SSB observed that it was unable to train enough statisticians through such means. A solution was suggested via self-study while on-the- job 在职自学. Self-study, in turn, casts a firm and unwavering spotlight on morale and its maintenance. The final third of the chapter is devoted to discussing this issue. Starting with an exploration of the role of the TJGZTX, which straddled the line between training and motivation, serving as a forum for both activities, I proceed to a sustained discussion of the RWTG, dwelling on its production and its content. I conclude by observing that if

Li Ming had been at the heart of the campaign, serving as a model subject in need of reformation, then by 1956 other more locally organized campaigns began to promote local activist model statistical workers as subjects worthy of emulation.

Red and Expert: The Need for Revolutionary Professionals

The Communists had won the Civil War on the back of a large-scale peasant revolution.

Their experience in governance prior to 1949 had been restricted to small pockets of territory and tiny populations in Jiangxi and the Shan-Gan-Ning border regions.480 The transition to greater political control, starting in 1948 in the Northeast and spreading elsewhere over time, was hamstrung by a severe shortage of trained and educated personnel.481 Such personnel were needed to effectively man the various and expanding

480 In some of these areas statistical work was undertaken as early as 1941. For more, see: Renmin University. Tongjixue jiangyi—diliu juan 《统计学讲义—第六卷》 (Lectures in Statistics—Volume 6). Beijing: Renmin University Press, 1959.

481 As Ezra Vogel observed in a 1967 essay: “In China, with a population larger than Europe and Africa combined, a vast majority of the population in 1949 had no school training, did not speak the major dialect, and was not part of a national communication network.” See: Ezra Vogel 1967: 40. In anticipation of this problem, which would only grow in magnitude, the Politburo had articulated the need to train cadre at a meeting at Xibaipo in September 1948. See: James Gao. The Communist Takeover of Hangzhou: the Transformation of City and Cadre, 1949-1954. Hawai’i: University of Hawai’I Press, 2004, 16. Also see: Lee 1991: 14, 45-46.

196

organs of the state and help govern the newly established republic. Given that the revolution and civil war era “conception of the cadre as a combat leader was not ideally suited to the tasks of civil administration that faced the Chinese Communists after 1949,” new cadres were needed.482 Recruits were, for the most part, drawn from three strata of society.

First, the “positivists” who gushed up from the movements; most of these were working class-people with more “virtue” than “ability”; we know that many workers were rapidly promoted to supervisory positions in factories, but probably only a few of them were given complex administrative positions. Second, the students, undoubtedly a much more promising source of recruitment; most of them had been sympathetic to the Communists, even though they were usually not involved in the illegal Party organizations of the cities; these were the “new intellectuals.” Third, the “old intelligentsia”: the new regime preferred young to old intellectuals, but, as An Tzu-wen indicated, the older ones were needed as well.483

Accordingly, the number of civilian cadres increased fourfold by 1953: from 720,000 in

1949, to 2,750,000 in September 1952, to 3,310,000 the following year.484

While discussions linking “virtue” 德 and “ability” 才 have a long history in

Chinese political and moral philosophy, they appear to have been first deployed in a communist revolutionary context by Mao in speech he made in 1938.485 In its later expression as Red and Expert, the dichotomy and relative dominance of one over the other at any given point in time has been an influential heuristic device for understanding

482 Schurmann 1968: 164.

483 Schurmann 1968: 168. Also see Vogel 1967: 54-59, for a discussion of the major groups of cadres based on their origins. On the general question of how the PRC state dealt with intellectuals in the early 1950s, see: Eddy U. “The Making of Zhishifenzi: The Critical Impact of Registration of Unemployed Intellectuals in the Early PRC.” The China Quarterly, no. 173 (2003): 100-121.

484 Schurmann 1968: 168; and Vogel 1967: 40; with both citing different works by An Tzu-wen.

485 Schurmann 1968: 165. Schurmann observes that Mao himself have been inspired by a speech by Stalin.

197

events and policies in China since the 1950s.486 But we would do well to remember that the ideal cadre was to be Red and Expert in equal measure.487 He was to have the technical and organizational skills to promote the work done in his unit, while at the same time also be brimming with ideological purity and revolutionary fervor so that any counter-revolutionary elements could not regain a foothold. Mao himself would note twenty years after his 1938 speech: “there is no doubt that politics and economy, and politics and technology should be united. It has been so in the past and will be so forever.

This is what red and expert mean.”488 The Red Expert thus drew his authority both from his expertise and from his ideological correctness. To the extent that experience was important, it was the cadre’s personal experience garnered in the present. The only significant past experience of value in this formulation was the advanced experience of the Soviet Union. In this way, the Republican era dichotomy between experience and expertise as two clashing sources of authority was neatly side-stepped.489 The PRC aimed to produce a new kind of cadre (professional).490

486 For a more recent instance, see: Joel Andreas. Rise of the Red Engineers: The Cultural Revolution and the Origins of China’s New Class. Stanford: Stanford University Press, 2009.

487 Meisner 1999.

488 Mao continued: “political workers must have some knowledge of business. It may be difficult for them to have a lot, but it may not do for them to have only a little. They must have some. To have no practical knowledge is to be pseudo-red, empty-headedly political. Politics and technology must be combined together. In agriculture, this means carrying out experiments; in industries, understanding advanced models, trying out new techniques, and producing new goods.” See: Mao Tse-Tung 1969. Red And Expert— January 31, 1958. Source: Long Live Mao Zedong Thought, a Red Guard Publication, 1969; included in Mao’s Collected Works, Vol. 8. Source: http://www.marxists.org/reference/archive/mao/selected- works/volume-8/mswv8_04.htm

489 This distinction has been illustrated particularly well in the field of medicine, where the experience of traditional Chinese medicine was pitted against the expertise of modern medical science, setting up two contrasting sources of authority. See: Daniel Asen 2012. “Dead Bodies and Forensic Science: Cultures of Expertise in China, 1800-1949.” Columbia University, PhD dissertation, 2012.; Sean Hsiang-Lin Lei 2002. “How Did Chinese Medicine Become Experiential? The Political Epistemology of Jingyan.” Positions: East Asia Cultures Critique 10.2 (2002): 333-364.

198

Students and young people more generally were quickly identified as the target group for achieving the ideal balance of Red and Expert. The “positivists,” because they were often too old to train, and the “old intelligentsia,” who were needed in the immediate short-term but would always be politically and ideologically suspect, were both deemed less suitable. Accordingly, the training and ideological indoctrination of the young was carried out at all possible levels, keeping both expediency and long term goals in mind.491 On a national scale, “of the 2,750,000 cadres in China in mid-1952, 66,000

[2.4%] had graduated from regular universities, 100,000 [3.6%] had been processed at

“people’s revolutionary universities,” but over 1,100,000 [40%] had been trained in rotating training courses.”492 Furthermore, this ideal of Red and Expert was applied not just to cadres who were, after all, defined by their leadership roles and by specific minimum ranks, but to all young skilled and semi-skilled workers needed by the state.493

University Education and Training: The Case of Renmin University

Discussions concerning the setting up educational institutions to train cadre picked up pace after the official declaration of the People’s Republic in October of 1949.494 Surveys

490 On an early discussion of the professional as possessor of authority, see: Talcott Parsons “The Professions and Social Structure.” Social Forces 17, no. 4 (May, 1939): 457-467. Recent works on question of professionalization have called for the need to move beyond Talcott Parson’s now “traditional” definition and focus instead on precisely the issues of expertise, skill, technique, and experience. See: Gordin 2004: 805; and Mackenzie 2006: 11 (on the role of professional judgment).

491 On the challenges faced by Wang Sihua in training adequate numbers of statistical cadre in the northeast in 1950, see: Wang Sihua 1986: 9-10.

492 Vogel 1967: 45; citing An Tzu-wen in turn.

493 It bears pointing out that the words 德 and 才 are never mobilized, though they are very much the implicit categories at work, in most of the materials I consult in this chapter.

494 For a summary of communist education and cadre training pre-1949 see Douglas Stiffle. “Building Socialism at Chinese People’s University: Chinese Cadres and Soviet Experts in the People’s Republic of

199

by visiting Soviet advisors had already investigated the nature and content of education at various institutions across the country. Eventually, it was decided that most major pre- existing universities would not disturbed.495 Instead, a new university would be established that would serve as a national model. Dubbed the People’s University 人民大

学 (Renmin Daxue), its principal focus was to train students in “economic-managerial specialties that were the regime’s highest priorities in 1949-1950.”496 The University was inaugurated on 3 October 1950.

In plans drawn up in November and December of 1949 it had been proposed that the University would offer two types of courses. The Regular-Course 本科 would involve two to four years of study in one of eight departments: Economics, Economic Planning,

Finance-Credit, Trade, Cooperatives, Factory Management, Law, and Diplomacy. Short- term training classes 训练班, envisaged as six-month courses offered by the same departments, emphasized economic management and state administration with the idea of producing trained cadre who could be put to work straightaway. In the December plan, the name of these short courses was changed from “training class” 训练班 to “specialized

China, 1949-1957.” University of California, San Diego, PhD Dissertation, 2002, 89-102. Also see: Michael Lindsay. Notes on Educational Problems in Communist China, 1941-1947. New York: Institute of Pacific Relations, 1950.

495 In Beijing, for instance, both Beijing and Tsinghua universities were allowed to continue operations without interference. See: Stiffler 2002: 57.

496 Stiffler 2002: 22 Later, Stiffler elaborates: ““In directing that a substantially revised plan be drafted, the CCP leadership reaffirmed Liu Shaoqi’s clearly-stated priority for the university: Renda would train Chinese cadres in economic management and state administration. It would not emphasize the training of cadres in Marxist-Leninist ideology, as in the Yan’an and Civil War-era cadre-training schools, and as the Soviet advisors recommended.” (p. 64). “By 1952, however, Renda’s priorities would shift back in the direction of the Marxist-Leninist ideological training that the Soviet experts had recommended.” On a possible precursor northeastern precursor (东北人民大学) see: Wang Sihua 1986: 9-10.

200

advanced classes” 专修班.497 As for who would attend these classes, it was decided that

“students of the “young intellectual” category should be eighteen to thirty years old and graduates of senior middle schools or “young intellectual” CCP cadres who had graduated from junior middle school or the equivalent and had at least three years of experience in CCP revolutionary work or specialized organs. These students would be enrolled in the Regular-Course Division.” In contrast, the Short-Courses would enroll cadre with education equivalent to lower middle-school, who had more than three years’ experience working in the “liberated area” and were between twenty and forty-five years in age, as well as retrained, old intellectuals.498

While not part of the November list of disciplines, Statistics was added by the time of the revised December plan.499 At this early stage, Statistics was conceived as being offered only in a short-term capacity and the November 1949 plan called for 300 students out of a total planned enrollment of 3,800 short term students. Of the planned

2,700 regular course seats, not a single one was in statistics. In the December plan, which involved a general downsizing, the corresponding numbers were 200 out of 3,000 short term students; and zero out of 1,400 regular-course students.500 Dedicated teachers too were downsized, from five out of 185 in the November plan to two out of fifty in the

497 Stiffler 2002: 66. Soviet advisors number decided at 50 (down from 185). “The CCP leadership decided to put in first place the six-month Short-Course Division, which emphasized economic management and state administration, while cutting the two-to four-year Regular-Course Division. This represented a clear victory for Liu Shaoqi’s pragmatic, short term, technocratic priorities, and for the regime’s goal of rapid economic recovery.”

498 Stiffler 2002: 145.

499 Pedagogy was also added at this time.

500 Stiffler 2002: 70.

201

December one.501 It was from such humble origins that Renmin University became the center of research and training in statistics over the course of ensuing the decade.

By June 1952, when a formal department of Statistics was established, the

University had become the central node for theoretical and pedagogical research in statistical training.502 As Table 1 shows, over the rest of the decade, at least 2,200 students received some form of training in statistics at the University.503

Table 1. Renmin University Statistics Graduates/Enrollees504 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 Total

专修生 Short-term Students 71 145 493 172 136 0 135 99 80 51 1382

研究生 Research Students 0 13 5 0 25 31 0 2 0 0 76

本科生 Regular Students 0 0 23 109 0 192 99 104 82 135 744

Total 71 158 521 281 161 223 234 205 162 186 2202 Source:

It is unclear whether the targets set in November and December 1949 for the first batch of students were actually met. Data from the Alumni Record indicates that only students attending ‘advanced specialized classes’ were graduated in the first two years.505 The first

501 Stiffler 2002: 72. Stiffler also notes that early curricular plans had extreme scarcity of specialized courses; curricula for several different courses actually similar. See pp. 79.

502 Its role in curricular organization and design, as well as in directing the general direction of statistical education, is discussed in Part I Chapter III.

503 This estimate and the individual breakdowns in the table are based on annual rosters of graduates produced by Renmin University. I do not know if these lists are exhaustive, so the likelihood that additional students were trained remains.

504 This table was compiled based on annual rosters from the statistics departments’ alumni publication. For details, please consult: Xiaoyou jiniance 校友纪念册 (Alumni Record). Beijing: People’s University Press, 2007, 1-7.

505 The data from the Record may be incomplete and as such the assertions made and conclusions drawn can at best be tentative.

202

class of graduates with the equivalent of an undergraduate degree was produced in

1953.506 This makes sense if, per the plans, regular courses were to last two to four years and students had started instruction in 1950. Even so, only twenty-three undergraduates were produced that year, less than five percent of the total (521) students enrolled in the statistics program. 1953 was also the year with the highest number of graduating students, coinciding with the start of the first five year plan (1953-57) and a push towards greater planning and statistical work.507 After that peak, the number of people trained in any given year was in the neighborhood of 200, with the share of undergraduates gradually rising to constitute the majority by 1960. The total number of graduate students or researchers 研究生 was just seventy-six for the entire period, with the majority concentrated around 1955 and 1956. The most significant and suggestive result for our purposes here is that even at a place like Renmin University, the focus was on preparing statistical workers via short term courses: over the entire decade 63% (1382 out of 2202) of all students were trained in advanced specialized classes.

In addition to Renmin University, other departments and schools were set up at a variety of other places during these early years. Table 2 provides a basic chronology of some of the major statistical schools or departments that were set up during the decade of the 1950s.508

506 Throughout the entire decade, statisticians could not attain the academic degree of Dr. 博士 or Assoc. Dr. (副博士), nor gain professional designations like engineers. See: The Self-Study of Statistical Cadre 1957: 1

507 After 1955 pressure to replace cadres (per Vogel) had reduced (find page citation), and this seems to fit with the numbers above, where the number of 专修生 declined after 1955.

508 The Renmin University School of Statistics’ Alumni Record for 2007 observes that the Renmin Department of Statistics was the first to be set up under the aegis of the new economics of the PRC. Right now, I am not sure how to evaluate this claim against information in Table 2 above.

203

Table 2. Establishment of Dedicated Statistical Schools or Departments Source: 中华人民共和国统计大事记，1949-2009 (DSJ)

Date Name of Bureau/Activity Province/ Region

Sometime in 1950 Xiamen University Statistics Department Fujian August of 1951 Northeast Planning Statistics Institute Northeast June of 1952 Renmin University Statistics Department Beijing November 29 1952 Northeast Finance and Economics Institute Northeast (Changchun?) July of 1954 Chongqing School of Statistics Chongqing August of 1954 Xi'an School of Statistics Xi'an June of 1955 Shanghai School of Statistics Shanghai Control of the Shanghai, Chongqing, and Changchun Statistical Shanghai, Chongqing, March 27 1958 Schools transferred to respective provinces Changchun Control of the Nanchang and Xi'an Statistical Schools transferred to March 27 1958 respective provinces Nanchang; Xi'an

Some of these schools and departments offered Regular Courses, but most of them also provided shorter term training. In many instances these were institutions that had existed prior to 1949 and had been reorganized and re-designated special schools. Even if we assume that these schools trained a similar number of students as Renmin University, we would arrive at a total of roughly 20,000 students, at most. Yet, by 1956, claims that there existed 200,000 statistical workers had been put forth.509 Clearly most of the training that occurred was not at major universities but elsewhere. One major training site was the vocational school: the 专科学校. As Li Fuchun had explained in 1951:

[We] cannot train advanced statisticians only at the university level; rather, under specific requirements and within the overall plan for education, every province, municipality, large region, should train grassroots statisticians.510

Vocational/Technical Schools

Li Ming did not have the good fortune to be trained at Renmin University or at any of the dedicated schools of statistics established in the early 1950s that are identified in Table 2.

In his letter, he denies us the name of his training center, noting only that it was a

509 RWTG: 3; elsewhere in the RWTG, a figure of 190,000 is also cited. See RWTG: 10.

510 DSJ: 5.

204

vocational/technical school. In this experience he was not alone. Expediency and overwhelming demand meant that most of the tens of thousands of statistical workers, a significant portion of the 1955 claim of 200,000, were trained in vocational or technical schools. Such schools, often established anew or created by the takeover of pre-existing institutions, dotted the entire country and came to supply an entire generation of state employees, including, of course, statistical workers. This section uses the example of the

Beijing Finance and Economics School to provide a sense of what such schooling may have been like.

The Beijing Finance and Economics School 北京财经学校 (BFES) was established on 31st March 1951. In its pre-liberation avatar it had been the Beijing

Business and Vocational High School 高级商业职业学校, which had been founded in

1929.511 Originally located in Wuying Hutong, by the time of its conversion to the BFES, the school had moved to a much more spacious location and could accommodate 900 students in eighteen classrooms and 180 dorm rooms. Facilities also included a library and rooms for various other activities. By the end of that year, when a short report on the school’s activities was drafted, total enrollment at the school stood at 575 students divided across Regular and Intensive classes. During those initial months, the school also graduated 179 students in the Regular Class stream, as well as ninety-eight students through short-term courses in accounting and statistics.512

Four courses of study were offered by the BFES during these initial years:

Regular Classes, Intensive classes, and two types of short-term classes. Each course of

511 BMA 004-010-00008: 7-9.

512 BMA 004-010-00008: 3-4.

205

study was designed with a different target audience in mind, and was set up to meet different goals. Table 3 summarizes these differences.

Table 3. BFES-Early characteristics (1951-1952)513 Short-term: Short-term: Regular Classes Intensive Classes Accounting & Tax Statistics

Educational Middle-school Middle-school Middle-school Qualifications equivalent High-school equivalent equivalent equivalent correct thought, healthy, willingness to Qualities work Correct thought, healthy

20-40; mostly 18-25; mostly 18-28 unemployed unemployed unemployed Age 15-20 intellectuals intellectuals intellectuals 6 months; per 3 years; receive 3 months; per pressing proper education, 3 months; per pressing pressing requirements, master modern needs, receive basic requirements, specialized finance and scientific and econ./fin. specialized training in training in relevant Course of Study economics skills relevant skills skills No. of classes 16 1 1 1 No. of students 543 32 49 49

Participate in city become middle government work (likely After graduation level cadre low level?) Discontinued after Discontinued after Remarks July 1951 July 1951

The Regular Course was designed as a three-year program for young teenagers with roughly the equivalent of a middle-school education. The School offered more than 16 regular courses (more than the stipulated 12) but could not meet the stipulated student teacher ratio of 2.5 to three teachers per course.514 Graduates were subsequently assigned to work in various bureaus of the city and central government as middle-level cadre. A report written in November of that year observed that while they were enthusiastic about work and willing to take on responsibility, their skill levels were inferior to the average university trained economics graduate .515 The content of the Regular Courses was often

513 Table prepared based on information in BMA 004-010-00008: 2, 1-9

514 BMA 004-010-00008: 3-4.

515 BMA 004-010-00008: 7-9

206

modified to meet the requirements of relevant city and central government bureaus but, in general, three types of majors were offered: Accounting, Statistics, and Factory

Management. Accounting majors would focus on cost accounting, government accounting, banking, and other specialized areas. Instruction in Statistics included advanced statistics, industrial statistics, and other related specialized fields. Factory management included such courses as Industrial economics and management and labor organization. In addition, basic disciplinary classes were part of each major, such as math, introductory statistics, and economic geography.516

In contrast to the Regular Classes, the intensive and short term training classes were designed to meet immediate and urgent staff shortages, usually in the city government. Such training normally lasted only three months and dealt with the highly specific skills required to carry out specific tasks. Students were generally older, and in many cases had complicated backgrounds because of their identification as unemployed intellectuals.517 For these students, political study was deemed especially vital in ensuring their loyalty to the revolution. The report lamented, however, that this was not done in an organized manner and their subsequent performance at work clearly indicated this. The two short-term training classes in Accountancy and Statistics appear not to have been permanent course offerings since they were discontinued after July 1951. A total of ninety-eight students (forty-nine in each course) were trained this way.

While Li Ming’s self-description of his education in statistics does not exactly fit any of the forms of training offered by the BFES in its first few years, it was probably closer to a truncated version of the three-year program described above. This may have

516 BMA 004-010-00008: 5

517 On the categorization and treatment of intellectuals in the early 1950s, see: U 2003.

207

been as much an outcome of differing courses of study drawn up in different parts of the country as it was an outcome of expediency, which may have led to a premature work assignment. It is likely, however, that the nature of his study itself was a close approximation of the situation in Beijing, where, by the end of 1951, sixteen regular courses were being offered by the BFES. The standard teaching method employed involved the use of the Five-Stage Teaching Method 五段教学方法.518 The five stages were: planning the class; checking previously assigned homework; teaching the lesson; consolidating (revising?) the lesson; and, finally, assigning homework for the next class.

Study materials involved a variety of textbooks designed for specific courses. In many instances, some of the Renmin University texts discussed in Part I were the principal study materials.519 Just as significant was study outside of the classroom. This mainly involved three things: i) monthly current affairs broadcasts; ii) the regular reading of papers; and iii) additional subject matter reading relating to class-work. In addition, various forms of teaching, support and guidance were provided in class and outside.520

We shall return to these extra-class forms of training in a later section.

On occasion, the BFES’ facilities were also borrowed by other agencies and companies to carry out additional training. This was especially sensible in the early years when the emphasis on practicality and expediency meant using all available resources to solve pressing concerns and because the school had space to spare. The November report

518 Based on Germany Educational Theory of the late nineteenth century; in a small minority of instances the three-stage method (san duan jiaoxue fangfa 三段教学方法) was also used. See BMA 004-010-00008: 6, 7-9.

519 At this early stage, books on mathematical statistics 统计数学 were also included.

520 BMA 004-010-00008: 6.

208

on the state of affairs at the BFES had noted with some satisfaction that after the move to its new location, the school’s infrastructure was more than adequate. In matter of fact, in

November 1951, only 575 students were registered when there was space for 900; similarly, only 521 students lived in dorms that could accommodate 900.521 As various agencies and companies sought to adapt to the changing economic goals put forward by the state, they were able, on their own initiative, to take advantage of this extra space to meet their short term requirements. One such instance is that of the Beijing City Public

Enterprise Company 北京市公营企业公司 and their attempts to deal with personnel shortages. In June 1951, the Manger and Deputy Manager of the company approached the City Government and the City Finance and Economics Committee with a request that they arrange short-term statistics and accounting courses over the summer and avail the facilities of the BFES to do so. They explained that the company was going through a major expansion of existing factories and the establishment of new ones but lacked cadre for such work. In particular, the lack of statistical workers and accountants was gravely impeding their work. They estimated that the University graduates already assigned to them that summer would not be enough to meet their ever-mounting needs. To solve this widening incongruity, they planned a three-month short-term course to develop statistical and accounting talent 会计、统计人才. To that end, they requested they be permitted to borrow 借 empty classrooms at the BFES to carry out this training.522 The request was

521 BMA 004-010-00008: 3-4.

522 BMA 004-007-00065: 1-2.

209

subsequently approved and in August the company reported that they had gotten in touch with the BFES and classes had commenced the previous month.523

By 1954, such ad hoc arrangements appear to have become less common. This is most obvious in the changed nature of reports and planning documents from that year and the year before.524 Differences with the 1951 report exist at the levels of form and content both. The nine-page 1951 report was hand-written, concise, and focused primarily on an empirical discussion of classrooms, types of classes, and style of teaching. Everything about it suggested immediacy, urgency, and improvisation. In contrast, the planning documents from 1953 and 1954 ran to several tens of pages, were typed or printed, and adopted formulaic language and standard outlines. Planned action, via regulation and the setting of specific objectives with specific deadlines, was the stated ideal. By 1953 already, certain basic themes were evident. The first and foremost amongst these was the elevation of the Soviet Union to the status of role-model. As a planning document noted,

“the teachers, after having studied and analyzed Soviet syllabi and teaching materials, have realized their superiority and the concordant importance of studying them intensely.”525 A 1954 planning document was even more explicit in discussing past achievements and future plans.526 According to it, an educational system was being built on a foundation of Marxism-Leninism. For guidance and advice it relied on the advanced experience of the Soviet Union and on the devotion of students and teachers alike, all of whom wanted to contribute toward the construction of a socialist society. Accordingly,

523 BMA 004-007-00065: 6-7.

524 BMA 004-001-00096 and BMA 004-001-00103.

525 BMA 004-001-00096: 3 Quoting of teachers becomes more and more common a strategy—can be understood perhaps as the application of the mass-line to education?

526 BMA 004-001-00103.

210

the report noted that over the past few years the level of planning had been raised.

Teachers had also raised their political and professional standards. New testing methods had been put in place and specialized syllabi and course materials had been prepared.

Teacher training was also actively being carried out at Renmin University and Beijing

Normal University.527

The School’s brief had also become narrower by this time. From a general vocational school for the training of cadre, it had transformed into a dedicated school focused on Industrial Management 工业管理, preparing middle level vocational/technical cadres 中等专业干部.528 Statistical training remained a core component of the school’s activities, though the scope of the courses was reduced to reflect the School’s new character. In 1953, statistics course offerings had included:

General Principles of Statistics 统计学原理; Industrial Statistics 工业统计; Capital

Construction Statistics 基本建设统计; and Social and Economic Statistics 社会经济统

计.529 By 1954, the latter two were jettisoned, leaving two qualitative forms of classes, one focusing on general principles and the second on specialized vocational statistical skills.530

527 BMA 004-001-00103:16.

528 BMA 004-001-00103.

529 BMA 004-001-00096: 2-3.

530 BMA 004-001-00103: 12.

211

Supplemental Training

A further form of training was supplemental, often carried out by taking a break from one’s job. Li Ming claimed to have received such supplementary training, which lasted three months. Such training likely became more and more common as the nature and quantum of work increased during the first five-year plan period. For instance, in

February of 1954, a joint notification from the Beijing Statistics Bureau and the Bureau of Personnel Administration was delivered to eight city bureaus concerning the transfer of already employed cadre for the purposes of further study.531 The notification stated that the BFES, in order to co-operate with the needs of the large scale economic construction going on in the country, would train on-the-job 在职 statistical cadre in a planned way and thereby raise the level of their professional skill. The notification went on to lament that in recent times the city’s statistical cadres were lacking in backbone 缺

乏骨干 and this problem needed to be addressed. After an investigation, it was resolved that additional training was especially needed for statistical cadre involved in capital construction work. Of the sixty-one cadre to be selected for training, thirty-five were to be selected from that group and, if feasible, the training was slated to begin the same month and last a total of six months. In addition to the thirty-five seats reserved for the

Construction Bureau, the remaining twenty-six seats were divided across seven bureaus:

Construction and Engineering Bureau (建筑工程局): 35 Local Industry Bureau (地方工业局): 8 Public Works Bureau (建设局): 5 Sanitation Engineering Bureau (卫生工程局): 5 Public Health Bureau (公共卫生局): 1 Public Services Bureau (公用局): 3 Education Bureau (教育局): 1

531 BMA 004-010-00812.

212

Housing Administration (房管局): 3 Total: 61

By 1957, a two-tier structure had emerged for such supplemental training classes.

In the first tier were classes organized by schools that were directly overseen by the State

Statistics Bureau. These were normally attached to universities or to dedicated statistics schools (see table 2). Every year, the SSB approached Provincial Statistics Bureaus with a request to send students to these schools. Candidates had to have the rank of section chief or higher and had to be vetted by the SSB before they could begin their studies. In addition, some schools insisted on administering admission tests. The training usually lasted one year. In the second tier were training classes set up by Provincial and City- level Statistics Bureaus, on occasion working in concert with the local Planning

Committees. The BFES was likely one such school. The student body consisted of cadre already employed in statistical workers in the local district/area. The training period was usually in the three to six month range. In 1956, approximately 5,000 students were attending such classes in schools across fourteen provinces (including Shandong,

Zhejiang, Heilongjiang, Henan, and Hunan).532 In addition, a similar two-tier system of cadre schools was also used to provide training. Employees of the central government were sent to cadre schools attached to various central ministries whereas provincial employees attended provincial cadre schools. Statistical training in these schools was organized in an ad-hoc manner and most likely substantial variation existed with regards to students, curriculum, and duration of study.533

532 The Self-Study of Statistical Cadre : 1957: 76.

533 The Self-Study of Statistical Cadre 1957: 76-77.

213

Two other methods were also employed to provide (supplemental) training: night schools 夜校 and correspondence courses 函授班. In 1957, night-classes were offered by the Shanghai, South-Central, Northeast, and Sichuan Academies of Finance and

Economics. Correspondence courses were offered by the Correspondence Division of

Renmin University directly from Beijing. The Division had also established coaching stations 辅导站 in nearby cities such as Tianjin 天津, Taiyuan 太原, and Baoding 保定.

Each of these schools targeted already employed statistical workers who possessed the equivalent of a senior-middle school education. Even so, they were usually required to pass an admission test and the nature and content of training was likely modeled on the regular classes.

The formal and supplemental training described here, in addition to university level education described in the earlier section, thus comprised the central bulwark of the state as it sought to create basic human resource infrastructural capacity in statistical work. Li Ming was a recipient of a mixture of such training. His initial training was at a vocational school, which was later supplemented by a short-term course in 1954. While I have found no aggregate data that indicate how many cadres were trained through these methods, we are now in a position to make some very general estimates.

Confronting Incapacity—Personnel Shortages and the Advocacy of Self-Study

Let us begin with some simple arithmetic. By 1955, the SSB claimed a workforce of

200,000 cadres. In an earlier section I estimated that the number of graduates trained in statistics at the university level was unlikely to have exceeded 20,000 for the decade of the ‘50s. This gives us an estimate of about 10,000 graduates by mid-decade. It is

214

somewhat harder to estimate the numbers trained at vocational schools. By the mid-50s there were schools in at least fourteen provinces. The BFES had just below 600 students enrolled in three-year courses in 1951. By 1955 this would mean three full cycles of students. However, we do not know how many of these were trained for statistical work.

For the purposes of estimation, let us assume other schools across the fourteen provinces were operating in similar fashion. In addition, we know that in 1956 about 5,000 students were attending three to six month long supplemental courses in these schools. That gives us an estimate for the year that ranges from 10,000 to 20,000. Assuming the same rate for

1951-55 would give us a range of 50,000 to 100,000 students trained via this method. We should remember though that supplemental education implies that these cadres had already received basic training in statistics, either at vocational schools, on the job, or they were a hold-over from the pre-1949 regime. Like Li Ming, many of them shouldn’t be counted twice in our attempt to estimate total numbers trained. Statisticians were also trained at other venues, notably cadre schools and schools run by various ministries.

Accepting that this method offers us an extremely imprecise estimate, we can still reasonably claim that it would be a real stretch to make the numbers add up to 200,000.534

So what can we conclude? Training of cadre, where it existed, consisted primarily of short-term courses, after which cadres were assigned to work in the hope that they would learn on-the-job. This mirrors the story at the national level as far as training of cadres went. The state was unable to train enough personnel through traditional schooling.

534 A regional example of such incapacity is the case of statistical work in Beijing in 1953. Of the 3,829 full-time and part-time statistical workers spread across the statistics bureau and various other industries, government departments, and related organs, only 618 were at the rank of scientific staff 科学员 or above. Furthermore, only 1,291 of the 3,829 had studied any statistics at all. See: BMA 004-008-00291:16.

215

The response, as far as statistics was concerned, was the advocacy of ‘self-study.’535 In fact, in 1957 the Education Department of the SSB published a volume entitled The Self- study of Statistical Cadre 《统计干部的自学》 where it went so far as to substantially abnegate responsibility for training sufficient numbers of personnel. As they explained in the volume’s first chapter:

But, we must realize one thing: no matter whether in recent times or in the future, the country cannot nor must it establish statistical schools and training courses on a large scale and get students to attend them. Even night school and correspondence classes can only realistically be organized in large cities and in limited numbers. Therefore, the education of statistical cadre must rely on one method—self study while on the job [在职自学].536

In addition to addressing the key limitation that not enough personnel were being trained, self-study also offered a solution to another common complaint. Li Ming had expressed dissatisfaction in his letter that in spite of his training there was much about the production process he did not understand. He worried that the blind application of statistical method was not in the least bit productive. One of the published responses to his letter began by confirming this precise problem: Li Yuren, a university graduate, acknowledged in his October 1955 letter his dismay at being unable to apply to statistical work the theory he had learnt in school. Only by learning on-the-job was he was able to bridge this divide.537 The publication of The Self-study of Statistical Cadre was thus aimed at addressing both these issues.

535 Another way to understand the need for self-study can be appreciated from a general sense of the statisticians’ brief: Bureaus and their employees, Prefecture and County Offices—2,200 + , Village cooperatives overseeing 750,000 villages. Then you had the industrial units and ministerial employees. Suddenly, 200,000 appear insufficient!

536 The Self-Study of Statistical Cadre 1957: 3.

537 Li Yuren. “Wo dui tongji gongzuo de renshi 我对统计工作的认识 (My Understanding of Statistical Work).” Tongji gongzuo 1955.10: 53. Also published in RWTG: 33-35.

216

In justifying the need for self-study, the volume offered some preliminary clarifications. The first of these had to do with the object of their study. China, it explained, was passing through a great historical epoch, transforming itself from a backward agricultural country to an advanced industrialized one. Such a goal could only be realized if the nation possessed the most advanced science and technology. But owing to the recent history of semi-colonization, and the even longer history of semi-feudalism,

China’s culture, science, and technology had all been utterly backward 十分落后. Since

1949, leaping progress 飞跃的发展 has been made, but much remained to be done. The previous year the party had prepared a twelve-year plan for science and technology, so that the nation might match or surpass the most advanced countries in the world. And now, the youth must march towards science and accomplish this great and formidable task. Statistical cadres were no exception to this expectation. The goal of their study was to unceasingly raise the standard of statistical work, to enrich the content of statistical science, and bring together statistical theory and practice, thereby contributing to nation- building. In other words, they were to transform themselves into statistical experts 统计

专门家. This kind of expertise was particularly important—empty headed theoreticians

空头理论家 whose study was not functional and divorced from actual requirements were labeled expressly undesirable.

The second clarification addressed the question of whether dedicated study was necessary. The volume explained that while statistical work naturally involved addition and subtraction 加加减减, it definitely was not limited only to those activities. Nor was statistics merely a numbers game 数字游戏. Rather, statistics possessed abundant political and economic content because they indicated the results of the labors of China’s

217

hundreds of millions of people. Study would ensure the avoidance of errors and accurate analysis, and thereby help in the deduction of the intrinsic laws of all objective things. In particular, knowledge of economics and business, political theory, and statistical theory were all critical to becoming statistical experts. But, the volume went on to lament, in recent times the level of statistical workers had been lacking. The only way of solving this contradiction was for workers at all levels to self-study. Admittedly, this was no easy task—time was limited and guidance often absent. And yet, the volume exhorted workers to persevere and carry out long-term self-study.538

Having already deployed the rhetoric of expertise, the volume in equal measure, also asked workers to be warriors 战士 contributing to the construction of a socialist society. Guidelines for self-study were divided into two sections: those dealing with elementary or foundational knowledge common to all statistical workers; and specialized materials and guidelines for sub-discipline specific statistics (e.g. industrial statistics, agricultural statistics, etc.).539

Elementary knowledge began with ensuring that all statistical workers had attained at least junior-middle-school literacy. A basic knowledge of mathematics, natural science and social science was also required. Next on the agenda were subjects such as

Political Economy, dialectical and historical materialism 辩证唯物主义; 历史唯物主义, and China’s revolutionary history. These subjects were viewed as the foundation of all social science and hence essential. This was followed by knowledge of the Party’s

General and Specific Policies and familiarity with practical professional work 业务知识.

538 The Self-Study of Statistical Cadre 1957: 2-3.

539 The Self-Study of Statistical Cadre 1957: 4.

218

Familiarity with the Party’s policies was deemed necessary since it would enable statistical cadre to provide data and analyses to the leadership so that policy might be formulated and implementation scrutinized. Practical professional work consisted of familiarizing cadre with the two most important methods of statistical data collection: (1)

The Periodical Report System 报表制度 and (2) Survey Plans 调查方案. And in mastering them, workers were also required to master the various relevant indicators 指

标 used within a bureau and their calculation methods 计算方法. Thus, in order to carry out good work one had to have familiarity with every indicator in the periodical reports and its overall place in organization 编制 and planning, in deciding policy, and in management. 540

General statistics related classes were divided into three categories.541 The first of these was a study of the basic problems in statistical theory, where the goals and methods of statistics, the collection of data, and their ordering and analysis were introduced.

Concepts included were: mass surveys 大量观察法, summarization 汇总, stratification

分组, averages 平均数, trends 动态数, and index numbers 指标数.542 The next category

540 The Self-Study of Statistical Cadre 1957: 5-6.

541 The Self-Study of Statistical Cadre 1957: pp. 7.

542 Recommended study materials included: i. Tongji lilun yiban wenti jianghua 统计理论一般问题讲话 (Common Problems in Statistical Theory). Beijing: China Statistical Press, 1956. ii. Tongji jiangzuo jiangyi 《统计讲座讲义》第 1-11 讲 (Statistical Lectures—No.1 to 11). Beijing: Statistical Press, 1957. iii. Tongjixue yuanli jiangyi 《统计学原理讲义》(Lectures on Statistical Principles). Beijing: Peoples University Press, 1956. Of these, the first is popular and easy to understand and meant for those with a low educational level, who have just begun the study of statistics. The contents of the second and third are relatively more detailed (abundant) and are suited for cadre with a higher level of education. [it is not specified what high and low here might be] For those statistical workers who want to study even further, they could study: Tongji lilun 《统计理论》第 1-10 章 (Statistical Theory). Edited by the Soviet Statistics Bureau and translated by the

219

consisted of branch specific statistics, for example, capital construction statistics focuses on those phenomena relating to construction work and their analysis.543 Finally, there was the study of statistical economics, which involved the collection of data relating to economic activity (indices), their content, calculation, and analysis. The objects of such analysis were central economic concerns: national income, balance sheets, the organization of economic work, etc.544 In addition, two related subjects that workers were encouraged to also study were accounting and the analysis of Economic Activity. And finally, there were materials on Marxist-Leninism related treatises on economics and statistics. Only by studying (and mastering) all these materials could one aspire to become a true statistical expert.545

The basic organization of self-study was calibrated based on the education and experience of individual cadre.546 For those cadres who did not possess a Junior-Middle

Statistical Press, 1957. Additionally, they could consult《统计学》by .Г. 斯特鲁米林 and 《统计一般理论问题》by Т.И. 科兹洛夫 et al. For details, please see: The Self-Study of Statistical Cadre : 9-10.

543 Much of the rest of the volume was dedicated to the discussion of such branch specific study—more on this below.

544 Recommended study materials included: i. 《统计讲座讲义》Every lecture following lecture no. 12 (国家统计局教育处编) ii. 《经济统计学讲义》(Xu Qian 徐前 et al. Statistical Press, 1957) iii. 《经济统计学教程》(by Prof. А.И. 彼得罗夫, Statistical Press, 1957) iv. 《统计理论》第 11-26 章(published by the Soviet Statistical Bureau) Of these the first two were meant for cadre whose education levels were relatively low; they were relatively easy to understand and their contents relatively comprehensive. The last of these was described as a good combination of statistical economics and basic problems in statistical theory; its portions on statistical theory were relatively comprehensive and it could serve as a principle reference book. For details, please see: The Self-Study of Statistical Cadre : 9-10

545 Study materials included: Documents relating to 8th People’s Congress as well as other important meetings of the government and party; articles from RMRB, other journals, especially relating to the state and party’s policies; Individual department’s internal newsletters (内部刊物) and Tongji gongzuo, Jihua jingji 计划经济; and the three volumes of Tongji gongzuo zhongyao wenjian huibian published by the Statistical Press in 1955, 1957, and 1959; and Jisuan jishu jiangyi 《计算技术讲义》. Beijing: People’s University Press, 1955.

546 The Self-Study of Statistical Cadre : 11-13.

220

School Education Level, the first task was to attain a level of education equivalent to

Junior-middle school and to familiarize oneself with the basic professional knowledge required for one’s job. Spare time outside of work was to be utilized to become literate.

Additional time, whether during or after work, had to also be dedicated to learning about party policies, and their implementation. At the next level were cadres who possessed education equal to or above the junior-middle school level but who had not systematically studied political and statistical theory. Such cadre had to exploit their familiarity with work and use it as a foundation for the study of theory. Their goal was to convert themselves into statistical cadre with the ability to work independently. Such study had three components: Political Theory, the Party’s Policies and Knowledge of

Professional Duties, and Statistics and related classes. During their study, the primary focus was to be on understanding the nature of work, especially the method of organizing periodic reports.

Cadre with several-years’ experience of statistical work comprised the third group.

In order to further raise their existing professional work experience, so that they could perform even better, their primary goal was to systematize their knowledge and experience and meld it with theory. These cadres were to dedicate the bulk of their spare time to study “statistics and related classes.” In order to do this they could begin by focusing on the specialized statistics of their branch and proceed to expand the scope of study based on specific needs that arose out of work.547 In the final category were cadres who had attained a university degree or had some university-level education. Their main tasks were to further research Political Economy and branch-specific economics, and to

547 In so doing, they were not, however, to neglect the “Party’s Policies and Knowledge of Professional Duties (党的方针政策和实际业务知识), since there was often new material to keep up with; see: The Self-Study of Statistical Cadre : 12.

221

master economic statistics and branch-specific statistics. In so doing, they would become

Statistical Experts, capable of producing reports with statistical analysis, and able to scientifically and practically address complicated questions surrounding statistical methods. Once the general foundational content had been clarified, statistical workers were directed to the main content of The Self-study of Statistical Cadre. Details of pedagogical materials and study methods were laid out in seven branch-specific chapters, their sequence indicating their relative importance: Industrial Statistics; Agricultural

Statistics; Capital Construction Statistics; Commercial Statistics; Material Supply;

Culture, Education, Health, and Hygiene; and Surveys of Family Income and Expenditure.

Finally, instructions were also provided on how workers might access the various study materials listed in The Self-study of Statistical Cadre. All books that had been made public 公开 could be purchased at the local Xinhua Bookstore. If a particular book was not available locally, then it could be ordered from the Beijing Mail-Order Bookstore 邮

购书店. Since the volume only listed a small proportion of the actual numbers of books available on economics and statistics, cadre were encouraged to browse catalogs and purchase other relevant books. As for those materials published or compiled and printed by the Interior 内部, somewhat different procedures would have to be followed.

Materials published by the Renmin University Press or by the Statistical Press, such as different volumes of the compendia Compilation of Important Documents on Statistical

Work 《统计工作重要文件汇编》, could be purchased from the relevant presses with a letter of introduction from the cadre’s home bureau. Compiled and printed materials, such as Statistics and Planning forms and their rules, survey plans, compilations of answered problems (问题回答汇编, as well as materials such as documents from last year’s

222

activist statistical workers meeting, and so on, could usually be obtained from the workers’ own units. Failing that, they could be sourced from an administratively higher

Statistical Agency.548

That self-study was largely a fiction, a utopia, is not hard to imagine. The program of study outlined in The Self-study of Statistical Cadre is sufficient to intimidate a full-time student, let alone a worker with a full-time job. Thus far, I have yet to find conclusive evidence that its adoption and use as a manual achieved any degree of popularity among statistical workers. And yet, such an assessment of The Self-study of

Statistical Cadre is also important for it points to another basic incapacity. Self-study was a road to self-improvement, a pathway to greater contribution to the nation. But such self- improvement could only be actualized if morale was high. For, only when morale was high would one expect to see cadres willing to dedicate their energies to advancing their own training and education. Let us, therefore, return once more to Li Ming and his frustrations.

Responding to Li Ming

Since its first issue in 1953, the Tongji gongzuo had served the dual role of a professional journal and didactic mouthpiece.549 By 1955, each monthly issue contained approximately 25 articles that spanned such topics as methodology, statistical data on specific questions or economic activities, reports of annual and local meetings concerning statistical work, translations of works by Soviet statisticians, reports from specific work

548 The Self-Study of Statistical Cadre : 75.

549 Related journals fulfilled similar roles as well. See for instance: Jianshe 《建设》; Jingji jihua《经济计划》, among others.

223

units on recent achievements, announcements declaring the start of a new campaign, and descriptions or transcripts of of major speeches that touched upon statistical work. By disseminating experience and information about statistical activities to various parts of the country many of these essays served a pedagogical purpose. Every issue also carried in it reports that touted achievements in statistical work or essays that wished to share experience 经验 about a particular kind of recently completed activity. Included also were essays by enthusiasts/activists 积极分子, meant to serve as encouragement to their fellow workers. On occasion, one could also find articles publicly shaming the poor work of a particular statistical worker or unit.550

By publishing Li Ming’s letter, however, the editors of Tongji gongzuo were adopting a new strategy. In what appears to be the first instance of its kind, the journal was used as a platform for a more open-ended debate on the very nature and purpose of statistical work. Instead of specific discussions on method or exhortations to do well a specific kind of statistical work, the editors (and the SSB as a whole) hoped to demonstrate and disseminate a mass ‘correct’ response to several key questions.551 The first of these centered on the claim that statistical work was mere arithmetic jugglery— adding, subtracting, calculating percentages, and so on—and, therefore, utterly boring 枯

燥无味.552 Was such an understanding correct, the editors asked? How should one arrive at an accurate understanding (appreciation) of statistical work? The second question dealt

550 See, for instance, the shaming of a unit leader for not filing statistical reports in: Tongji gongzuo 1955.11: 48.

551 Li Ming was, of course, not the only one to cast doubts on statistics. An article in the RWTG named a handful of other workers who had registered similar complaints. Other names included: Liu Guang 刘光, Zhi Ge 止戈, etc. See: RWTG: 23, 63.

552 Kuzao wuwei; lit. dry and tasteless.

224

with the issue of usefulness. The editors invited readers to ascertain whether Li Ming’s verdict on the uselessness of statistics was correct. They encouraged readers to consider the role and status of statistics in the ongoing national construction (nation-building).

How could statistics’ full potential be brought into play? The final question addressed Li

Ming’s fears of the lack of advancement and the poor future prospects of statistical workers in general. Li Ming had lost faith 失去信心 in statistics, they acknowledged.

But was this justified? What, after all, were the prospects of statistical workers; could they rise through the ranks?553 In October 1955, two months into the debate, the editors of the journal decided to collapse the three questions down to one central question: Why

[are we] unwilling to do statistical work?554

Over 600 responses were received by the journal in the fourth months from

August to November.555 A representative handful was published each month: five in

August, four in September, five in October, and three in November. The December issue brought the debate to a close and carried two articles. The first was a general overview of the debate, summarizing content from the numerous letters that had been received but not published. The second, as noted at the start of this chapter, was the essay by SSB Deputy

Director Jia Qiyun. The eighteen responses as well as the summary essays were subsequently published as RWTG. Before an analysis of the debate can be offered it is worth probing the logic of switching to a single question prompt in October—the

553 Li Ming 历明. “Comrade Li Ming’s Letter.” Tongji gongzuo 1955.07: 42. Republished in RWTG: 7.

554 RWTG: 64.

555 RWTG:55.

225

unwillingness to do statistical work. For that, we need to consider the peculiar status of statistical work itself.

The (Peculiar) Status of Statistics

For a whole generation that came of age in the late 1940s and early 1950s, the gun-toting, landlord-slaying peasant revolutionary was a heroic ideal. Even if many did not directly participate in the revolution, sympathy for the communists was widespread. Accordingly, one of the principle challenges for the CCP once they came to power in 1949 was to successfully channel this popular fervor towards the task of national construction—the building of a new China. The broad economic strategy adopted to achieve this goal was that of industrialization. Any surplus produced in the economy was to be invested in industrial production. Agriculture, the dominant sector of the economy, was marginalized to the extent that its primary purpose was to support the ongoing industrialization. This bias in favor of production, a matter of economic policy, in turn dictated certain social policies. The emphasis on productivity and rapid industrialization at the expense of agriculture went hand-in-hand with a shift in focus from the country-side to the city. For a rural, peasant based, and largely peasant-led revolution, urban China had become the new field of action and achievement.556 It is not surprising that concerns with the un- regulated movement of people, especially from the rural country-side into cities, began to grow, and were often justified via a need for stability and the desire to ensure the best conditions for production.

556 Of course there was land reform, which was rural, but the major thrust now was towards the city. Old and new works have focused on this transition from a variety of angles. See, for instance: Vogel 1969 (Canton), Pickowicz and Brown 2007 (various regions in China), and Gao 2004 (Hangzhou).

226

One of the results of this shift was the birth of a new type of hero, who engaged in production for the country. This applied to agricultural and industrial work in equal measure, but the emphasis on industry meant that the especially heroic work was done by those working in factories, producing the raw materials and finished goods that would build a new China. This had implications for the valuation of administrative and service- based work. The “new hero was not the administrative cadre but the engineer.”557 For skill-based technical professions, such as statistics, the repercussions were dire.558 As

Wang Xiangyuan explained in his essay in Tongji gongzuo:

Right now there are many comrades who are unwilling to do non- productive technical work [非生产技术性], but are willing to do productive technical work. This kind of thinking is present not just among statistical workers, but also those who work in accounting, note-taking, official correspondence, typing, etc.559

Statistics, along with the other professions that Wang listed, did not enjoy the common man’s respect because they did not produce tangible goods.560 This judgment is reflected in a table in Ten Great Years, which lists the relative labor distribution between productive work 生产部门 and non-productive work 非生产部门 annually from 1949 to

1958. Titled “Rapid Increase in Percentage of Workers and Other Employees in

Productive Work,” the table shows a clear upward trend in the share of productive work:

557 Vogel 1967: 48.

558 An interesting contrast to the PRC was India, where statistics in the early post-independence years was highly valorized and its practitioners very powerful; more on that in Part III.

559 Wang Xiangyuan. “我一定把工作做得更好.” In: RWTG: 38.

560 On the distinction between productive and non-productive work and its roots in Smithian and later Marxian analysis, see: Susan Peter-Smith. “Defining the Russian People: Konstantin Arsen'ev and Russian Statistics Before 1861.” History of Science 45, part 1 (2007): 47-64, 50.

227

climbing from 65% in 1949 to 85.1% by 1958; the share of non-productive work declining correspondingly from 35% to 14.9%.561

Even amongst ‘non-productive technical’ professions, statistics stood out. A science that straddled the administrative and the productive, it required skilled workers capable of collecting and analyzing vast amounts of data so that tangible production could be better understood, organized, and improved. Its close cousin, accounting, had similar characteristics but was different in significant respects.562 Unlike statistics, it was not part of a national organization linking the center to village and work units.

Accountants were usually restricted to a work-unit, village, cooperative, factory, and so on, and rarely operated at higher levels of aggregation. Statisticians, in contrast, were part of a national organization that numbered approximately 200,000 workers by the mid-

1950s. And statistics was fundamental to the rhetoric (and reality) of a socialist state, whose logic dictated that the centrality of planning was not merely economic theory, but also a philosophical and practical necessity, predicating all other activity. The collection and tracking of all information—productive and non-productive—so that informed and appropriate decisions regarding planning and allocation of men and materiel could take place was the very cornerstone on which everything rested. But making the argument theoretically and in the halls of power was one thing, getting the average statistical worker to do the work, quite another.

To address this problem two related strategies, one a broad appeal to logic and the other an emotional appeal to the romance of production, can be located. Let us look at the

561 Ten Great Years 1959: 164. The last year in the table, 1957-58, accounts for a significant portion of the change: a swing of about 12.2%.

562 See: The Self-Study of Statistical Cadre 1957: 7-8, where this distinction is dwelt upon; and Ezhov 1957: 40-41 (Soviet Statistics) .

228

broad appeal first. The logical solution was to claim that all kinds of activity were important and revolutionary—this, after all, was the view put forward in all textbooks and self-study books, explicitly or implicitly. As one statistical worker explained in his response to Li Ming:

…I have come to realize [through my work] that all kinds of revolutionary work [一切革命工作] are important. There are no distinctions such as ‘important and unimportant,’ ‘small and large,’ ‘valuable and worthless;’ rather, every profession produces its own leading authority [行行出状元]. All one needs to do is to work hard, and no matter the nature of the work, success will follow…563

The valorization of productive work over non-productive work was also attacked.

Another interlocutor observed that “although statistical workers did not directly participate in production and, therefore, did not create value, their effect within the larger process of socialist development could not easily be calculated using such a metric.”564

The most cogent and sustained exposition of this line of reasoning was provided by deputy director Jia Qiyun 贾启云. In his December essay he wrote that he wished to explain why every type of work was an important part of national development and that if one worked hard and with a focused mind, success was guaranteed. The logic to this was straightforward. Jia proceeded to explain that even though heavy industries were the basis of building a socialist society, and working in that sector of the economy was no doubt an honor, other sectors of the economy were just as important. Indeed, without development in other sectors of the economy, such as light industries, agriculture, transportation, commerce, and so on, developing heavy industries alone was impossible. Therefore, each type of work was equally honorable, equally glorious. Furthermore, it was imperative to

563 RWTG: 37. For another instance, see: RWTG: 42.

564 RWTG: 57.

229

recognize that all these important activities could not proceed unless they were supported by administration, a justice system, diplomacy, national defense, education, and health.

Regardless of whether it was productive 生产性 or non-productive 非生产性, each type of work was of equal importance and as honorable and full of glory as any other. And underlying all these activities was planning, for which statistics and numbers were absolutely critical. Statistics was thus the basis 依据 for everything: whatever the job, if statistical work was set aside, then it was tantamount to setting aside any prospects for progress.565 In such a way was statistics placed at the very heart of revolutionary work and the socialist enterprise.

Cold, calculated logic, however, also found a worthy partner in powerful imagery and rhetorical flourish. If one was unmoved by the seemingly irrefutable logic of Deputy

Director Jia’s argument, as indeed many may have been, then Feng Jixi’s formulation may have held greater appeal. Feng, another interlocutor in the Li Ming debate, began his essay with these words:

Every time I complete a statistical table/chart, my happiness is like that of a peasant on his field catching sight of a golden ear of wheat, my excitement like that of a steelworker observing molten steel emerging from a Martin furnace, (and) my elation like that of an artist completing a beautiful painting.566

The sleight of hand was obvious. In one sweeping sentence, Feng had equated statistical work with work in agriculture, industry, and the arts. The key lay in equating a statistical table (and the report it would invariably be a part of) with the tangible goods produced in other fields. The table, thus, was the tangible, material product that a statistician

565 RWTG: 67-68. In a subsequent passage he similarly dismissed the notion that statistical work was subordinate to other kinds of work.

566 RWTG: 10.

230

contributed to the economy. In the previous chapter we have already encountered the kinds of problems this valorization of the table-as-product generated, in particular the oft- repeated and much-lamented phenomenon of excess issuing of reports 滥发报表.

Feng’s was not the only case where Li Ming’s interlocutors resorted to rhetorical flights of fancy, but it was perhaps the most flowery and emphatic of them all. The romanticization of numbers and statistics had been legitimated just a few months earlier by none other a figure than poet and author Ba Jin 巴金 (1904-2005). In an essay published in the People’s Daily he relayed with approval the description of the ongoing five-year plan as a “song of numbers.”567 In the essay, Ba Jin accepted that in the past people liked to say that numbers were absolutely boring 枯燥无味, but felt such a formulation was no longer tenable.568 Numbers, he explained, contained in themselves abundant meaning. Far from being inert, they could speak; they could relay countless touching stories and paint beautiful paintings:569

I have always believed: "Poetry should give people the joy of creation, poetry should spread life [?]." These numbers not only spread life and happiness, they not only depict to us a happy future, but they also guarantee that happiness. Not only do these numbers gather the sentiments of 600 million people, they also embody their common aspirations and are their signpost [pointing to the future]. With them, correctly, step-by-step, we shall arrive on the road to [building a] socialist society. They are like a bright lamp, illuminating the hearts of 600 million.570

567 Ba Jin 巴金. “‘Shuzi de shi, xinfu de baozheng 数字的诗’，幸福的保证 (‘The Song of Numbers,’ the Guarantee of Happiness).” Renmin ribao 16 July 1955. Also in: Ba Jin. Ba Jin quanji: di shiwu juan 巴金全集—第十五卷 (The Collected Works of Ba Jin: Vol 15). Beijing: Renmin wenxue chubanshe, 1986, 90- 94.

568 As the interlocutors in RWTG would remind us, the value of statistics had indeed changed in the new China. See: RWTG: 34.

569 Feng had also acknowledged the revolutionary strength of numbers; see: RWTG: 10.

570 The Collected Works of Ba Jin: Vol 15: 92. A variation of this was also quoted in the TJGZTX and RWTG: 60.

231

The implication was clear. If Li Ming grasped the power of numbers and statistics in the

New China, he would not find them boring at all. And the evocation of collective spirit, of course, was not incidental either; nor was the focus on the people’s common aspirations. These themes provided grounds for what was the most concerted and popular attack on Li Ming.

Attacking Individualism, Honoring Model Workers

Eleven of the eighteen responses published in the TJGZTX and the RWTG highlighted Li

Ming’s problem as one essentially stemming from individualism. The broad basis of the critique was provided by the Qingdao Statistics Bureau’s Newsletter Group. In their letter, they observed:

It is our opinion that the basic problem can be traced to the fact that Comrade Li Ming has not considered his statistical work in light of the needs of the nation and the people. He has not recognized that every kind of work carries with it great revolutionary significance…Instead Li Ming persists in separating himself and thinking only of his personal prospects and advancement...571

In similar fashion, another interlocutor explained that Li Ming was deceived, even blinded, by personal profit. As a result, he was unable to see the uses and effects of statistics.572 Indeed, for a revolutionary worker, individualism was very dangerous. It was like a noose tying the worker down, making him loose direction, much like a man lost in the darkness of the night. Such a worker would only go round and round in his small circle, paying attention only to himself, ignoring the community. He would focus on the

571 RWTG: 17.

572 RWTG: 11.

232

small, not the big; on the near, not the distant; and even after the passage of one, two, or three years, he would remain unchanged.573 Numerous other instances, declaiming Li

Ming’s individualism, were published.574 On occasion, capitalist was added to the charge, making it an error not just of individualism but of capitalist individualism.575

Tied to individualism was the implicit tendency towards careerism 野心, even though it was never articulated as such. The term used instead to discuss the matter was

‘future prospects’ 前途. Li Ming had raised the issue in his letter as well. One key problem stemmed from the fact that, unlike various other technical professions, one could not earn advanced degrees or certifications in statistics. The Self-study of Statistical

Cadre had acknowledged this problem on its first page, observing that the SSB had received numerous letters decrying the absence of advanced degrees for statisticians.

Neither advanced academic degrees of doctor or associate doctor 副博 nor any professional qualifications such as those granted to engineers were available to statistical workers.576 Such circumstances naturally raised concerns about career advancement, professional development, and pay scales among a statistical work force that was extremely young.577

The response to problems of careerism began with the assertion that the pay- scales were appropriate, but also sought to reframe the object of study and work. Cadre

573 RWTG: 12.

574 See, for instance, RWTG: 16, 26, 36, 38, 43, 48, 61, 63, 67.

575 RWTG: 35, 65; the latter in Jia Qiyun’s summary essay.

576 The Self-Study of Statistical Cadre 1957: 1.

577 One interlocutor noted, for instance, that even accountants and other service based professions were doing better on these fronts.; see RWTG: 24. On the pay and hiring structures among statisticians in 1950s PRC, see: RWTG: 74-76.

233

were told that rather than the attainment of a doctorate, certification as an engineer, or some similar qualification, the goal was to develop and make prosperous the nation’s culture, science, and technology, and to bring about rapid industrialization. Statistical cadres were no exception to this call to arms. In other words, once people were clear about this, they would no longer obsess about degrees and status, and their achievements would be recognized appropriately. In the final analysis, accomplishments at work and not degrees or pay scales were the true measure of study.578

It is at this point that Li Ming, the negative model, had served his purpose. In order to stress achievement over qualification, positive models, worthy of emulation needed to be promoted.579 As Jia Qiyun wrote in his December essay, over the past few years there had been countless heroic model workers who had made stellar contributions across the various fields of national development work. And, as Jia clarified, statistical workers were very much among such heroes:

At this year's "National Meeting of Youth Activists for the Construction of Socialism" were there not just model engineers and (industrial) workers, but also model shop assistants, model cooks, model pig-farmers? Among our statistical workers was there also not an advanced youth group led by comrade Chen Jinbao? They all have the same [glorious] future prospects.580

578 The Self-Study of Statistical Cadre 1957: 2.

579 The general promotion of model workers had been happening at the national level through conferences where activist/enthusiastic workers exchanged experiences, their reports subsequently being compiled and disseminated for wider didactic purposes. A discussion on the Mass Line noted: “leaders must therefore be skilled in uniting the small number of active elements to form a leading group and must rely on this leading group to raise the level of the intermediate elements and to win over the backward elements.” See: De Bary 2000: 417

580 RWTG: 69.

234

Indeed, much like other cadre, statistical workers were also recipients of various model worker awards.581 Such awards were usually made after year-end appraisals had been carried out.582 One of Li Ming’s interlocutors had reported that because of his stellar work he had been recognized as a Third Degree Model Worker in 1953 三等工作模范.

Continued excellence at work had resulted in his elevation to a Second Degree Model

Worker the following year. This honor was not just limited to within the company but also across the city of Taiyuan, and he was handed the label of Shanxi’s Advanced

Worker.583 Similarly, another statistical worker recalled how she was made a First and

Second Degree Advanced Worker.584 The point was thus reiterated: statistical workers had prospects 前途; they could become statistical experts and model workers 模范工作

者.585

By 1956, the time was ripe for a model worker conference dedicated to statistical work. The Beijing Statistics Bureau decided to take the lead at the local level. On April

18 1956, it sent out a notification to all bureaus/departments, communes, and associations within the city to begin making preparations for the city’s (and nation’s?) first Activist

Statistical Workers Meeting 统计工作积极分子大会.586 Held the following month, the two-day meeting brought together nearly 300 people and recognized and honored

581 As early as 1950 Wang Sihua had spoken of the need to master and popularize the experienced of advanced workers (“掌握与普及先进分子的工作经验”). See: Wang 1986: 7.

582 RWTG: 37.

583 RWTG: 52-53.

584 RWTG: 59. Another instance can be found on pp. 37.

585 RWTG: 51.

586 BMA 035-002-00203: 2. A similar meeting was organized in Sichuan in March 1957, which was graced with the presence of Zhu De. See: DSJ: 52.

235

advanced workers who were on the battle lines 战线 of statistical work.587 Their testimonies were compiled for dissemination in the hope that they would, in turn, inspire other statistical cadre to give free rein to their enthusiasm and creativity.

Conclusion

Given the PRC’s limited resources, the story of capacity building in the field of statistics very quickly turned into one of how best to handle incapacity. Such incapacity manifested itself in three successive areas. The first incapacity was reflected in the need to build training infrastructure, both in terms of schools and training centers and in terms of a statistical curriculum that could accommodate a very wide range of students. Indeed, barring a small minority of cadre who had college or high-school level education, most cadre were in their late teens with extremely limited education. A second incapacity was characterized by an inability to train cadre in sufficient numbers and in a rapid enough fashion. Finally, morale was another area in which the Statistics Bureau encountered tremendous challenges. By mid-1950s, as work was stabilizing and the PRC had entered the high point of its bureaucratic phase, the SSB could take stock of its achievements and also start on activities like self-study and model-worker meetings. This chapter has provided a sense of what the state aspired to do, how and where it fell short, where it abnegated responsibility, and what kinds of actions it took to address that shortfall.

The consequences of some of the rhetorical flourishes one sees with regard to statistics and numbers have already been traced in the previous chapter. In particular, the fetishization of periodical reports—their being touted as the ‘product’ that statistical

587 BMA 035-002-00203: 7-9.

236

workers were laboring to produce—exacerbated what was already a pretty common problem, the excess issuing of reports. As discussed in the previous chapter, the problem became increasingly insurmountable as leaders in the SSB recognized that the very system of periodical reports as much as their fetishization was responsible for masses of paper and no way to actually deal with them.

237

CHAPTER VII

Statistical State, Quantified Society

238

By 1956, an integrated statistical system modeled on the Soviet Union had allowed for unprecedented levels of penetration stretching down to the village level. With a staff of about 650 personnel, the SSB, headquartered in Beijing, oversaw 200,000 statistical cadre spread across provincial bureaus, roughly 2,200 county offices, and about 750,000 village cooperatives. The scope and nature of statistical data collected, putatively, not only increases a state’s capacity to regulate society, but through the creation of new categories also reflects its vision of social change.588 1950s China was marked by several massive nationwide social engineering projects, such as cooperativization and collectivization of agriculture, state take-over of the private sector, minority classification, household registration and restriction of rural mobility, and perhaps most ambitiously, economic planning.

Normally understood as top-down outcomes primarily of ideology, in this chapter

I offer a preliminary investigation of the role of statistics and statisticians in such processes, taking as case studies economic planning and the institution of the work points system in the countryside. In each case, the object is to demonstrate the centrality of statistics to the conceptualization and execution of large scale social and economic transformation. In the case of the first five-year plan, the consistent challenge was to better coordinate the planning and statistical apparatuses of the state. Drawing upon

Beijing Planning and Statistical Bureau documents from 1954, I first show the

588 There is a large literature on this subject. See, for instance: Bernard Cohn. An Anthropologist among Historians. Oxford: Oxford University Press, 1990 (in particular pp. 224-254, on the Indian census); Ian Hacking. “Biopower and the Avalanche of Printed Numbers.” Humanities in Society 5 (1982): 279–95; Ian Hacking. “Making up People,” (Essay based on “Tenth Annual British Academy Lecture,” April 2006; the full text of the lecture appeared in the Proceedings of the British Academy) London Review of Books, August 17 (2006): 7-11; Martha Lampland and Susan Leigh Starr, eds. Standards and their Stories: How Quantifying, Classifying, and Formalizing Practices Shapes Every Life. Ithaca, NY: Cornell University Press, 2009; James C. Scott. Seeking Like a State: How Certain Schemes to Improve the Human Condition Have Failed. New Haven: Yale University Press, 1998; and many others.

239

relationship between statistics and planning and between the statistics bureau and the planning bureau, as it was understood during such an early an early stage of the plan. I then move forward in time to 1958 and look at the discussion surrounding the impending second five-year plan (1958-1962).

Even if, in the end, the plan was haphazard and not very plan-like in comparison to other contemporary plans, the drive to plan and the idealization of total statistical information that it was based upon wrought several changes in the country side. While resistance to the state’s attempts to collect information and tax, the universal historical pattern one would expect to find, was indeed one response, the creation of new forms of statistical data and categories of accounting also resulted in numbers informing new conceptions of everyday life and work. In the second half of the chapter I explore one such instance: the work points system and the concomitant rise of what I term enumerative consciousness.

The First Five-Year Plan and Statistics

In most standard (economic) histories of the 1950s, the decade is divided into three periods: 1949 to 1952 is the period of recovery 恢复时期, 1953 to 1957 is the period of the first five-year plan 第一个五年计划, and 1958-1962 is the period of the Great Leap

Forward 大跃进.589 The first five year plan thus occupies a central position in the decade, temporally as well as in terms of its impact on China’s development. Even though it

589 Needless to say, variations on this theme exist. The original prospective formulation in 1949 had called for “three years of recovery, then ten years of development.” (Riskin 1987: 53). 1949-1956 is also frequently referred to as the period of transition 过渡时期, implying that by its end the three major socialist transformations—of agriculture, handicrafts, and the capitalist economy—had largely been successfully achieved. 1958-1962 would originally have been the period of the second five-year plan; the Great Leap Forward put paid to those plans.

240

started in 1953, it was only after July 1955 that the plan itself was released as a document.590 Over the years, general scholarly consensus has been that it was less a plan in the strictest sense of the word and more a set of guidelines.591 As the economist (verify)

Audrey Donnithorne summarized in 1967:

China’s economic planning has been restricted mainly to the setting of targets, to drawing up lists of resolutions. It does not attempt to effect close integration of different economic sectors, nor is it much concerned with the optimum allocation of resources. Throughout, and this can scarcely be stressed too much, economic planning in China is constrained by the deficiencies of the information on which it has to work, as well as by the weaknesses in the administration and supervisory organs charged with implementation of plans and with checking this implementation.”592

A significant component of the deficiencies of information that Donnithorne spoke of had to do with the quality of statistical data. Writing twenty years later, Carl Riskin would echo this assessment, noting that national economic planning could not proceed without the “establishment of an elementary infrastructure, including planning and statistical capacities.”593

590 This was done after Li Fuchun 李富春, Vice-Premier of the State Council and Chairman of the State Planning Commission, delivered his address on July 5 and 6 1955 at the Second Session of the First National People’s Congress in Beijing. The second part of Li’s speech is a full summary of the Plan. The Foreign Language Press produced an official English translation of the speech by October of that year and a full translation of the Five-Year Plan was issued in 1956 (First Five-Year Plan for Development of the National Economy of the People’s Republic of China 1953-1957). For more, including the text of the 1955 Foreign Language Press publication, see: MacFarquhar, Roderick and John K. Fairbank, eds. The Cambridge History of China, Vol. 14: The People's Republic, Part 1: The Emergence of Revolutionary China, 1949-1965. Cambridge: Cambridge University Press, 1987, 42-91.

591 The first draft of the plan was rejected by Soviet advisors in the Soviet Union in 1952. The plan that was subsequently approved focused primarily on the state-owned economy and favored industrial production and capital construction; investment in agricultural production constituted an almost negligible proportion of overall investment. I do not offer here a general discussion of the plan and its contents. Trenchant discussions can be found in Riskin 1987: 53-60; Li Choh-Ming. Economic Development of Communist China. Berkeley: University of California Press, 1959; Meisner1999.

592 Audrey Donnithorne. China’s Economic System. New York: Praeger, 1967, 457; cited in Orleans 1974: 51.

593 Riskin 1987: 53.

241

The importance of statistics to planning was stressed from the very earliest discussions about building statistical and planning infrastructure but the relationship was usually articulated in general terms. Recall Wang Sihua’s 1950 speech (from Chapter II), wherein he stated that by establishing a system of socialist statistics, China was already on the path toward a planned economy. A somewhat more precisely stated link between the planning and statistics can be found in a document entitled “Written Answers to

Questions Concerning State Planning in China.” This document was prepared a few months prior to Li Fuchun’s speech in July 1955 and delivered to the Indian Prime

Minister, Jawaharlal Nehru (1889-1964) through the offices of Vice-Premier Chen Yun

陈云 (1905-1995). 594 According to the document, the plan was drawn up by working out control figures, making a draft plan, and approving it for delivery to lower organs. Two systems were critical to this process: the ‘administration under the central authority’ and the ‘system of local administration.’ In the first step, the State Planning Commission

(SPC) worked out control figures for the whole nation. Upon the approval of the State

Administrative Council (SAC), these figures were sent down to the lower organs.595 At this point, various ministries and local administrations would use these figures to draw up their own draft plans and submit them for approval to the SAC, with copies also delivered to the SPC. On the basis of these lower level draft plans, the SPC then revised the

594 The questionnaire had been prepared by P.C. Mahalanobis (more on him in Part III), who had requested Nehru to present it to the Chinese on his official visit to the People’s Republic in 1954. Nehru had handed the questionnaire to Vice-Premier Guo Moruo 郭沫若 (1892-1978) on October 22 1954. The Chinese original of the document was translated into English by the Chinese Foreign Office and both documents arrived in New Delhi (through the Indian Embassy in Beijing) by the end of April 1955. For more, see: NAI _EA_FEA_10(48)-FEA/55: 1, 3.

595 Set up in 1952, the SPC came under the direct control of the State Administrative Council after the 1954 reorganization.

242

national draft plan.596 The process was recursive and relied on coordination across the various administrative levels and with statistical organs:

The State Planning Commission maintains connections with local administrators through the planning commissions of the provinces, municipalities and autonomous regions. In addition to the planning commissions of the provinces, municipalities and autonomous regions, local planning organs, to be known as “planning and statistical sections” are set up in the counties.597 (Original English translation)

At the lowest levels, planning and statistical functions were thus combined within a single office. At the national level, the SPC’s principal partner was the State Statistics

Bureau (SSB). Both the SSB and the SPC came under the direct control of the SAC. To quote the document: “the state takes the figures of the Statistical Bureau as base for drawing up its plan, and reviews the execution of the plan on the strength of the figures of the Statistical Bureau.”598 (Original English translation)

Much like Mao’s exhortation to “serve the people” 为人民服务, statistical work thus had to “serve national planning” 统计工作要为国家计划服务. As a report on

Beijing’s Third Meeting on Statistical Work explained, such service involved three tasks.

The first task was the supply of correct data 首先应提供正确的资料. Second, analysis and research was conducted on the correct data to discover problems 根据正确的资料进

行分析研究, 发现问题. Finally, based on the analysis, solutions to those problems were suggested 提出解决问题的办法. The statistical numbers so produced were, in turn, the basis for preparing plans 编制计划, inspecting plans 检查计划, deciding on policies 决

596 See: NAI _EA_FEA_10(48)-FEA/55: 4 (see p. 17 for Chinese version).

597 “计划统计科”; see: NAI _EA_FEA_10(48)-FEA/55: 4 (see p. 17 for Chinese version).

598 See: NAI _EA_FEA_10(48)-FEA/55: 4 (see p. 17 for Chinese version).

243

定政策, and for checking the implementation of those policies 检查政策执行情况. So, for example, the statistics bureau could check the rate of industrialization 观察工业发展

速度, the rate of nationalization 公私比重变化, or formulate a variety of supply plans 制

定各种供应计划. The report argued that all these activities were important because they were closely linked to national construction 国家建设, to production 生产事业, and to the people’s lives 人民生活.599

Given the need for statistics to work closely with planning what was the reality on the ground? At the level of the central headquarters of each organization, coordination appears to have been the de facto norm. Such coordination was no doubt facilitated by the fact that both the State Statistics Bureau and the State Planning Commission were located on different floors of the same building in Beijing. But if we take the example of statistical and planning work at the city level, things are less clear. According to materials connected with Beijing’s Third Meeting on Statistical Work in 1954, of the 114 sanctioned attendees to the meeting, only five were from the planning bureau.600 The rest were statistical cadres spread across forty-five units 单位 and key factories and mines 重

点厂矿. On first glance, these numbers indicate that the Beijing Planning Bureau’s input in statistical work was limited. But when the total number of cadre at the Beijing

Planning Bureau is taken into account (forty-one), we find that a significant portion of the

Planning Bureau was actually in attendance (roughly twelve percent; or five out of forty- one). In order to investigate further the relationship between the two bureaus further, let

599 BMA004-010-00807: 81.

600 BMA 004-010-00807: 9. The post-meeting report stated that in reality there were as many as 148 attendees and over sixty observers. See: BMA 004-010-00807: 80.

244

us therefore consider the annual work summaries 工作总结 for 1953 of the Beijing

Planning and Statistics Bureaus.

Reflections on Planning and Statistics, 1952-1958

A summary report of statistical work carried out in 1953 in Beijing was prepared by the

Beijing Statistics Bureau in April 1954.601 The report stated that there were 3,829 full- time and part-time cadre involved in statistical work in Beijing.602 Over the course of

1953, some major improvements were made in the timeliness of reports and in ensuring that numerical data were accompanied by textual explanations 文字说明.603 At the same time, most of the issues discussed in Part II Chapter II above, such as duplication, chaotic reporting, etc., were also pointed out in the report. One of the reports five sections

(section three) was devoted to a discussion of the bureau’s contribution to planning work.

The section on planning began by observing that the bureau had provided statistical data to the leadership to help them draw up plans and decide on policies. For instance, in July of 1953, under the leadership of the Finance Committee, the bureau had compiled and consolidated annual reports for every year from 1949 to 1952 to aid in long-term planning. In the few cases where data was not available, they had provided estimates. These materials were then cross-checked by various relevant ministries and organs and printed in booklet form 印成册. In order to supply data needed for drawing up a plan for 1954, the bureau had also provided forecast figures for 1953. Data was also

601 See: BMA 004-008-00291: 9-22. For earlier drafts of this report, see: BMA 004-011-00034: 13-22, 24- 42.

602 See: BMA 004-008-00291: 16.

603 See: BMA 004-008-00291: 11.

245

provided to aid policy decisions and carry out key tasks 中心工作. For example, in order to research the relative magnitudes of the private and public economy, the bureau provided quarterly data on the share of the commercial economy. In similar fashion,

Beijing’s population was verified so that the state monopoly for the purchase and marketing of grain could be effectively implemented. Data on average wages and labor productivity were also provided in order to facilitate research into the problem of wages .604

Even though the bureau felt that its various offerings had substantially met contemporary planning requirements, it acknowledged that shortcomings nonetheless remained. For starters, the bureau admitted that it was not good at utilizing statistical numbers from all possible sources, ignoring in particular statistical data from the administrative and professional departments 业务部门. Even in those cases where a mastery of all the numbers had been achieved, the data was often not fully utilized.605 A second concern was that specific needs were not addressed directly. Problems were discovered and the leadership informed only once data had been consolidated. As a result, even though the data provided to the leadership was plentiful, it played a limited role in decision-makin.606

What did the Planning Bureau’s annual summary of work for 1953 say about statistics? Drafted in January of 1954, the annual summary noted that the Beijing

Planning Bureau had officially been set up on October 21 1953 and had more than forty-

604 See: BMA 004-008-00291: 13.

605 “See: BMA 004-008-00291: 14.

606 See: BMA 004-008-00291: 14.

246

one cadre distributed across departments responsible for planning in industry, agriculture, infrastructure, trade, health and education, as well as five others.607 The Bureau received its instructions from the State Planning Commission, the City Committee, the City

Finance Committee, and the city government. Its tasks were to examine, consolidate, and balance long-term, annual, and seasonal plans for local industry, agriculture, trade cooperatives, capital construction, health and education, and wages using data obtained from the statistics bureau.608

The report went on to note that planning work had suffered on account of three reasons. Leaders had been unable to focus on the organization of planning because they were busy with the state purchasing monopoly on food grains. The planners also encountered problems with the quality of index numbers and often did not get corrected numbers delivered to them on time. Finally, they complained about inadequate numbers of cadre and their poor overall training.609 Neither statistics nor statistical work was directly mentioned, and yet two of the three problems they listed were essentially statistical in nature. It was only later in the year, that the SPC directly addressed the issue of planning and statistics on a national level.

On August 24 1954, the State Planning Commission passed a set of regulations/provisions 规定, which directly addressed the relationship between planning and statistics. “Regarding Uniform Provisions for a number of Problems [Encountered] in

Planning and Statistical Work” was a six-page document that was sent to all central committees, departments, institutes, and to the statistics and planning bureaus of every

607 BMA 004-011-00034: 1-6.

608 BMA 004-011-00034: 2.

609 BMA 004-011-00034: 3.

247

province and municipality.610 Additionally, copies were delivered to all units of the planning commission and to the State Statistics Bureau offices in Beijing.611 The document noted that problems existed both at the systemic level and at the level of coordination. Coordination between Planning Units and Statistics Units, in particular, was frequently inconsistent. Such inconsistency, in turn, affected the calculation and meaning of specific indicators. Furthermore, subordinate planning and statistics departments often did not strictly enforce the SPC and SSB’s jointly issued regulations, or they habitually supplemented these regulations with their own contradictory regulations. Such ad hoc localized decision-making included changing a planning unit’s affiliations and economic structures without consultation.612 Finally, a single unit frequently possessed multiple sets of planning numbers and did not clearly specify which set was to form the basis for checking plan implementation. The consequence was that statistical work and planning work at any given level were disjointed. Duplication of data and the chaotic nature of statistical work adversely affected the quality of planning work.

In order to resolve these problems of coordination, the SPC proposed five provisions. First, they insisted upon greater cooperation and understanding between the planning and statistics offices. At every level of administration, the statistics office must be in compliance with the corresponding planning committee. It was the statistics office’s responsibility to provide, examine, and approve all basic data needed for planning. This structure was to be replicated at all levels from the province down to the county. In order

610 BMA 004-011-00255: 1-7.

611 It notes that copies were also delivered to (SSB?) Committee members 张、按、韩 and to 王副秘书长.

612 BMA 004-011-00255: 2.

248

for this to work in practice, planning departments were required to make their requests for statistical data well in advance of their use date, so that statistical activity could be properly planned and the data delivered in a timely fashion.613

The second provision focused on problems with indicators, stressing that their meaning and the methods used to calculate and classify them should all be in agreement across statistics and planning work. Without such commensurability, it would be impossible to produce any meaningful comparisons. Where planning guidelines already existed, they were to be used to rationalize the indicators. Similarly, if in places the only guidelines were statistical ones, then they were to serve as the basis for the rationalization.

And if neither statistical nor planning guidelines were available, then representatives from both statistical and planning offices were required to work together and devise mutually agreeable ones. In case of any special conditions or problems, the opinion of both the SSB and SPC had to be sought. Under no circumstances were local officials supposed to act before obtaining the SPC and the SSB’s approval. In general, indicators were to be calculated based on the following SSB catalogs and classification schemes:

i. Catalog of Industrial Products

ii. Catalog of Agricultural Products

iii. Fixed Prices for Agricultural and Industrial Products

iv. Industrial Sector Categories

v. Classification of Workers and Wages

If planning or statistics offices at any subordinate level considered a particular calculation method or categorization of indicator required improvement, they had to seek the joint

613 BMA 004-011-00255: 3.

249

approval of the SSB and the SPC. And as noted above, they were not to take any action prior to obtaining such approval.614

Among the key tasks for statistical organs was checking the rate of plan implementation. Accordingly, the third provision stressed the need for statistics reporting units and summary (comprehensive) units to be in compliance with planning units.

In particular, the planning units that fell within the National Economic Plan (NEP) were required to use the SSB’s periodical reports and survey estimates. Any changes in the relationship between planning units or in economic type again could only be carried out in consultation with both the SPC and the SSB and all catalogs used by planning units had to be copied to statistics offices.615

For statistics to effectively aid drafting of plans and to track their implementation, the fourth provision focused on the unifying the methods of plan implementation. In all cases, nationally approved numbers were to be used. So, for instance, when checking the implementation of a particular annual plan, the National Annual Plan was to serve as the basis for comparison. Similarly, when checking quarterly plans, quarterly numbers were to be drawn from the National Annual Plan and used as the basis for comparison. If such numbers did not exist, corresponding quarterly figures from the next higher level were to be used. In the case of monthly plans for grassroots industries, the quarterly plan of the superior administrative unit was to be broken into monthly estimates and used. As with earlier provisions, any major changes or corrections could only be made after obtaining

614 BMA 004-011-00255: 3-4.

615 BMA 004-011-00255: 4-5.

250

approval. In the absence of any approval, the units were to continue using the older

[incorrect] numbers and copy all results to the statistics division.616

The final provision reiterated the relative administrative ranking of planning and statistical units. At every level planning had to provide leadership to statistical work.

Cadres from both units were required to periodically meet and discuss statistical work, so that day-to-day work was better and more closely coordinated. In addition, cadres from each unit were required to attend each other’s work meetings and when necessary, to jointlyreport and discuss their work with the central government.617

Reading against the grain, these five provisions as well as the annual reports for

1953 from the Beijing planning and statistics bureaus suggest that in the first year of the five-year plan there was limited and poor communication between planning and statistics activities. Planning and statistics offices frequently operated on an ad hoc basis, calculating their own indicators without appropriate approval, thereby adding to the cacophony of data and analyses. The relatively skewed numbers of personnel for Beijing, where a little over forty-one planning cadre had to coordinate their work with nearly

4,000 statistical cadres, is also suggestive of how difficult it must have been for the planning organs to provide leadership and oversight of statistical organs. Oversight by the planning organs also meant that statistical organs were under two different authorities— people’s committees (as discussed in Chapter V) and planning committees. In many instances, and especially at the local level, there may have existed significant overlap in personnel between the two. But even so, administratively they were different and therefore could easily be the source of different and possibly contradictory requests as far

616 BMA 004-011-00255: 5-6.

617 BMA 004-011-00255: 6.

251

as the statistical units were concerned. In spite of these difficulties, however (and perhaps remarkably), the SSB did produce annual reports on economic development and plan implementation for every year during the first five-year plan.

Even though the problems plaguing statistics and planning were accurately assessed by 1954, their resolution remained elusive over the following years. A notice issued jointly by the SPC and the SSB in April 1958 repeated many items from the 1954 provisions just discussed and indicates that many problems remained unresolved even at the end of the first five-year plan.618 At the same time, developments during the four years since 1954 had also generated new challenges. The 1958 notice remarked that in keeping pace with expansions in the scope 口径 [lit. aperture] of annual plans, the range of statistical work had also expanded every year. As a result, statistical work had surpassed the scope 口径 of the first five-year plan. For example, in the case of investment in infrastructure, the five-year plan did not include investments in the military commission, in the bureau for the management of government offices, nor in administrative departments. However, such investments were all part of individual annual plans by the end of the first five-year period. In similar fashion, the five-year plan for geological surveying did not include several items labeled infrastructure investment that were included within annual plans or in the plans of individual departments.

Continuing differences in the classification methods used by statistical and planning organs also resulted in inconsistencies. For example, the sown area for tobacco

土烟 and for different kinds of hemp and flex (Cannabis Hemp 大麻; Velvet Leaf 青麻; and Kenaf 羊麻) was included in the five-year plan. However, in regular statistical work,

618 BMA 005-001-000298: 1-6.

252

which was more attuned to the requirements of annual plans, they were listed within the sown area for technical crops 技术作物. In another example, the five-year plan did not include tubers 薯类, the finances of food supply,619 and catering 饮食业用 in its figures for retail sales of food commodities 粮食商品零售量. These were included, however, in the annual statistical figures. Finally, there were quite a few indicators whose calculations methods were not consistent. For instance, the indicator for food did not include data for soybeans and indicators of infrastructure investment in agriculture, forestry, or water management did not include investment data from the forest industry. 620

Much like in 1954, the SSB and SPC again recommended a slew of measures to try and combat these problems. In those instances where the scope of statistics exceeded that of planning, they recommended that the results be publicly announced, without correcting for the differences. This was particularly acceptable when the expansion in scope was more-or-less offset by the reduction in prices or other factors, such as in the case of infrastructure investment.621 In those cases where the scope of planning exceeded that of statistics, the statistical data needed to be adjusted by making appropriate estimations, with the caveat that the adjustments could be avoided if the differences were small and did not affect claims about the overall completion of the plan. On the question of inconsistencies in classification, the SPC and the SSB now called for planning categories to take precedence over statistical ones.

619 Need to verify this; possible error in transcription or in original.

620 BMA 005-001-000298: 3.

621 BMA 005-001-000298: 3.

253

The notice recognized that changes in the scope of the plan were sometimes made during the implementation process itself through a variety of actions, such as increasing or decreasing the number of construction units. Under such circumstances, instead of constantly revising plans, the notice asked that such changes be carefully recorded so that once the plan was completed the figures could be compared to the original plan figures.

Furthermore, because the nature of the economy underwent substantial change during the plan period, when analyzing plan completion rates, it was best to use annual statistical data, since the planning data may not be commensurable for different years.622 It was with such considerations in mind that discussions about the second five year plan were undertaken. In late 1957 most provincial statistics bureaus drew up their own research plans. The Fujian Bureau, for instance, listed thirty-two questions ranging from analysis of various aspects of the first five-year plan to discussions about changes necessary for the next plan.623

Work Points

Even though agriculture was not the central thrust of planning, over the course of the decade the party-state progressively nationalized land, reorganized social relations in the countryside, and instituted a universal system of wages that fundamentally reshaped rural

China. Statistics and quantification were at the heart of these changes. Measuring agricultural labor, in particular, developed into an increasingly important task as the state increasingly became not just the sole purveyor but also the sole provider of people’s

622 BMA 005-001-000298: 3-4.

623 BMA 133-001-00091: 43-45. Similar lists were produced by the Xi’an Statistics Bureau, the Inner Mongolia Statistics Bureau, and many others. See: BMA 133-001-00091: 49, 63-34.

254

incomes. The principal method of keeping track of income earned through the performance of labor was the work-points system. The work-points system quantified every aspect of rural life by classifying every activity within the rural economy in terms of the number of points needed to perform it. Putatively, every activity was thus reduced to single number, accounted for on a daily basis. As a result, people’s relationship to numbers changed in new and fundamental ways.

The Work-Point System 工分制, also known as the Labor-Day System 劳动日制, was a system of quantifying daily labor and allocating consumer goods in the rural countryside throughout most of the Mao era.624 According to Huang Yingwei 黄英伟, the concept of evaluating (someone’s) work and recording that evaluation in terms of points

评工记分 was presumably in use before 1949.625 The first documentary evidence of the use of the phrase “evaluate work and record points” can be found in a mid-1950 report in

China Agricultural Report 《中国农报》. By July of 1952, three methods for estimating a labor day had been proposed. In the first case, work points were preset and work was assigned based on those ratings. In the second case, work points for a task were predetermined 预定 but the actual assignment of points was carried out after the work had been completed by a democratic reckoning. Finally, the third method called for the

624 See: Huang Yingwei 黄英伟. “Gongfenzhi de zhidu jingjixue fenxi 公分制的制度经济学分析 (Analysis of the Institional Economics of the Work-Point System).” Ershiyi shiji shuangyue kan 二十一世纪双月刊, no. 131 (June 2012): 87-96.; and Huang Yingwei 黄英伟 Gongfenzhi xia de nonghu laodong 工分制下的农户劳动 (Rural Labour under the Work-Point System. Beijing: China Agriculture Press, 2012, 44-70. Much the discussion that follows, draws upon Huang’s work. A useful contemporary source is: Andrew Nathan. “China’s Work-Point System: A Study in Agricultural “Splittism.” Current Scene— Developments in Mainland China 2, no. 31, (April 15, 1964): 1-13.. Also see: Peter Schran. The Development of Chinese Agriculture, 1950-1959. Urbana: University of Illinois Press, 1969, pp. 30-32. For a description of work points during the 1960s and later, see: Mobo C.F. Gao. Gao Village—A Portrait of Rural Life in Modern China. London: Hurst and Company, 1999, 61-63.

625 Huang 2012a: 88; Huang 2012b: 45.

255

implementation of a system where quality, quantity, and the nature of work were all contractually fixed. These three systems would in later years be dubbed as ‘fixed points, fixed valuation’ 死分死评; ‘fixed points, negotiable valuation’ 死分活评; and ‘fixed labor’ 定额工.626 Sporadic discussions about these methods and their employment in tracking daily labor continued over the following few years.627 At this early stage, when mutual aid teams were the norm, “work points were used to keep track of the work put in by Farmer A on the plot of Farmer B, so that the latter could reciprocate fairly accurately, in accordance with the principle of “exchanging work for work”.628 Points were thus based solely on time, and did not distinguish between different types of work and their relative physical and mental requirements. Furthermore, as an exchange purely of labor, work points were not income in the strict sense.629

It was not until November 9 1955, however, that a formal definition and interpretation were authorized by the Standing Committee of the National People's

Congress at its twenty-fourth meeting. From that point on, the work points system became a system of wages, operating on the principle: “calculate payment based on labor; the greater the labor the greater the gain” 按劳计酬、多劳多得.630 A work day 工作日 was calculated as the quantity and quality of work that could be performed by a medium

(average) amount of labor 中等的劳动力, taking into account a variety of factors such as

626 Huang 2012b: 46. The first two have been understood as ‘time-work’ and ‘piece-work’ in English language scholarship. Cf. Nathan 1964: 2.

627 See: Huang 2012b: 46-47.

628 Nathan 1964: 3.

629 Nathan 1964: 3.

630 Huang 2012a: 88; Huang 2012b: 47-48.

256

the nature of work, the quality of land, farm animals, agricultural tools, daylight hours 天

时, etc.631 One such work day was equal to ten work points 工分.632

The following year, at the end of June, the transition from lower-state cooperatives to higher-stage cooperatives was initiated and all land was moved into collective ownership.633 Unlike in the previous period of lower-stage cooperatives, when land had remained privately owned even as it was collectively worked, this stage witnessed the quantification of every possible activity through what is known as the

‘piece-work’ system.634 Accordingly, all forms of jobs were considered and rates were established for each in terms of work points. So, for instance, ploughing wheat could be worth 10 points per mu 亩, threshing it 9 points per mu, collecting manure 8 points per ton, and so on.635 Income from personal land was thus no longer possible and one’s entire earnings were based on wages calculated as work points.

Enumerative Consciousness

From the state’s perspective, a work-point system allowed for a degree of control over the harvest that payment in direct monetary terms would have made much more difficult.

As Andrew Nathan explained:

If wages are expressed in money terms, the State would be committed to certain expenditure in farmer’s wages; but with work

631 Huang 2012b: 47.

632 For a more detailed discussion of the various labor grades that were applied, based on gender, skill, and other criteria, see: Nathan 1964: 4; Huang 2012b: 50-57.

633 Huang 2012b: 48.

634 Nathan 1964: 2..

635 For a table of various jobs connected with farming wheat 小麦 see: Huang 2012b: 55-56.

257

points, the money value is only known after State procurement. Then the total number of work points earned on the team is divided into the total value to be distributed to arrive at the value of each work point. Meanwhile, the farmer has competed with others to get more points only in order to get a higher proportion of the distribution ultimately allotted by the State.636

Monetary reckoning was usually done at the time of the Lunar Festival. Therefore, for much of the year work-points retained their abstract, purely numerical value. Even so, the rhetoric behind work points was very much one of emancipation from earlier feudal structures of state extraction. If in the pre-modern and early-modern past, statistics was a tool whereby the state and gentry taxed and oppressed peasants, then in the new China it was the tool of a munificent state that ensured the well-being of peasants and the development of the nation as a whole. In a section titled “The Abacus Clicks” in Lu Qi’s

1963 poem “Willow Village Revisited,” this transformation is captured through the changing role of the abacus.

…This abacus started clicking early: That year we staged the struggle, It computed exploitation, added up lost lives, Click, clack, click, clack… It’s the thousands of rifles firing salvos, And the battle drums thundering together; It added and added until Heaven and earth turned over and man changed his fate!

It kept clicking till today, and today It no longer computes the rent, but work points. Yesterday it figured out whose hate was the strongest Today it tells us whose love is the deepest. Listen now, listen! Throb the thousands of communers’ hearts.637

636 Nathan 1964: 2

637 Lu Ch’i 陸棨 (Lu Qi). “Willow Village Revisited.” In Literature of the People’s Republic of China, edited by : Hsu Kai-yu, 729. Bloomington and London: Indian University Press, 1980. Originally published in Poetry Journal, no. 3 (1963): 4-7.

258

The abacus was a tool that originally calculated what peasants owed an extractive state and gentry. During Land Reform it became a tool in overturning the feudal order. It helped assess the value of that same extraction by quantifying the scale of exploitation so that peasants could be reimbursed through accurately calculated land redistribution. And shortly thereafter, even such harsh duties were set aside: “it no longer computes the rent, but work points.” Work points thus became indicators of peace and prosperity for all and were crucial to the social and economic transformation of the country-side.

Attitudes toward work points also pointed to whether one whole-heartedly supported the ongoing cooperativization. In Ma Feng’s “I Knew Three Years Ago,” a story written in the late 1950s but which was set earlier in the decade, the chief protagonist Uncle Chao (Man-tun) is portrayed as a “smart cookie” who had a difficult time adjusting to the cessation of his private commercial activities and the forfeiture of his assets, both of which were done to further the communal interests of the cooperative.

Fretting over work points was one way in which his disgruntlement manifested itself:

“Huh! Me grabbing for work points! Who doesn’t want to earn a few more work-day credits? You keep saying the co-op is your family. Why don’t you labor just for the co-op then? Or, “Even if you grew gold from the ground, if you don’t earn any work-day credit, you could stare till your eyes pop out and still have no share in it!”638

A little later in the story, Uncle Chao’s wife confides in the author, noting that his attitude has improved:

She paused and then said, “It’s much better now. For the past few years, our share at the end of the year has never been less than what we would

638 Ma Feng 马烽. “San nian qian zao zhidao 三年前早知道 (I Knew Three Years Ago).” In Literature of the People’s Republic of China, 565. Bloomington and London: Indian University Press, 1980. Translation based upon version published in: A Bumper Crop of Short Stories, Beijing: 1959.

259

get if we worked on our own [land]. This has put his mind at ease. Nowadays, he only adds up his work points every few days…639

That Uncle Chao was only adding up his work points “every few days” is suggestive of how people may have fixated on their daily point tally as they sought to keep track of not just their own work points but also of their fellow co-op members. Both these literary examples thus point to the importance of work points as an increasingly pervasive and foundational element of life in the countryside. This impression is also reflected in reports sharing positive work experience that were filed by village level statistical workers.

On April 18 1956, the Beijing Statistics Bureau sent out a notification to all bureaus/departments, communes, and associations within the city to begin making preparations for the city’s [and nation’s?] first Activist Statistical Workers Meeting 统计

工作经济发展大会.640 Held the following month from May 24 to 26, the meeting summarized and disseminated model work experience in statistics.641 In all, about 300 statistical cadres from fifty-three departments, bureaus, and administrative units participated in the meeting.642 Among the at least twenty statements delivered at the meeting were two that focused on work in agricultural statistics.643 In each case, work points were central to the statistical cadre’s testimony.

639 Ma Feng 1980: 566.

640 BMA 035-002-00203: 2-11.

641 BMA 035-002-00203: 15.

642 Of these, a significant number—a little over a fourth—were involved in industrial work. Three hundred was the sanctioned number. The actual numbers may have been different. See: BMA 035-002-00203: 7-9.

643 BMA 035-002-00203: 41-46, 73-75.

260

The first statement was made by Yang Shi’en 杨世恩, who was the statistician of the Number Nine Team of the Czech Agricultural Cooperative 中捷友好农业社 in the suburbs of Beijing.644 Yang was a graduate at the higher-primary level and during school would often listen to talks introducing Soviet collective farming and the happy lives of its members. Deeply influenced by them, he decided to volunteer in agricultural production.

After graduating in 1953, he decided to put down his book bag and joined the village’s

Tile Production Group 本乡的农业生产瓦助组.645 That same winter, Zhanglin

Township 张邻庄乡 started to build a cooperative and he was sent (by the cooperative) to study agricultural technology. Upon returning, he was assigned the dual tasks of co-op statistician and technician. Even though he was initially concerned he didn’t have the ability to take care of statistical work, encouragement from his Team Leader Guo Baojin

郭保金 assuaged those fears.646

Yang encountered many difficulties during the first few months of statistical work at the cooperative. He discovered that the data being collected were completely unable to meet both the needs of the periodical reports that were reported up as well as the more local needs of the team and cooperative leaders. In many instances data was not freely provided; in those instances when it was, much of it appeared to be based upon estimation 估计. Yang proceeded to offer an example: while calculating the area under cotton he asked the Team Leader, “How many mu 亩 have you planted?” The Team

644 Yang’s was the fifth statement in the compilation. The following paragraphs are drawn directly from his statement, the text of which can be found in: BMA 035-002-00203: 41-46.

645 BMA 035-002-00203: 42.

646 “At that time I was quite afraid. I would think: one, I cannot calculate; two, I cannot remember (things), how will I be able to do statistical work?” See: BMA 035-002-00203: 42.

261

Leader replied, “Roughly 17 or 18 mu.” “Exactly how many,” Yang persisted. He replied:

“You can write 17.5!”647 In similar fashion, no exact estimates of the amount of wheat sown or land under wheat cultivation existed. As a result, numbers couldn’t be successfully recorded and even when they were, the data recorded and the reality were frequently at odds with each other.

Yang discovered that work points were recorded in an especially chaotic manner.

And unlike other kinds of data, their improper recording generated a tremendous amount of tension among members of the cooperative. To demonstrate this phenomenon, Yang recalled the case of an old cooperative member 老大爸 whose points had been under- recorded. The old man had furiously accosted Yang, saying: "Right now the land does not ___, [only] through these [points] can we eat and drink. [If] you forget it [recording work-points] it's okay [for you], [but] for me each point is my very lifeblood."648 Yang found himself deeply troubled by such accusations, which he had to admit were true. But how, then, was one to apportion work points properly? The problem came to occupy him completely: “even food [I] was unable to eat; if [I] wanted to sleep [I] was unable to sleep.”649 In the end, it was Old Man Guo’s 郭大爸 statement that each point was his very lifeblood 纷纷就是命根 that made Yang resolve to improve the quality of statistical work. Work points, he recognized, lay at the root of everything: “if work points are

647 BMA 035-002-00203: 42.

648 See: BMA 035-002-00203: 43.

649 BMA 035-002-00203: 43.

262

recorded well, then not only will that influence the life and production of the masses, but it will also influence the correctness of statistical materials.”650

And so, he decided to work on improving his understanding of the relationship between work points and business and accounting calculations at the cooperative. During the spring planting season in 1955, Yang sought out the production team leaders to discuss and devise an improved system of recording work points and related statistical data. Based on these discussions, he devised a system whereby two types of data were simultaneously recorded. On the one hand, he accurately recorded each worker’s work points. On the other hand, he recorded in numeric terms the daily value of each worker’s completed farm work, such as plowing the land, sowing, removing pests, spreading manure, etc. This data was compiled into a daily interim summary. Every ten days, he would produce a full summary, which he would send to the county office. Yang explained that such a system allowed him to judge whether the allocation of work points was reasonable. In addition, he was now also able to assess in a more timely fashion the progress being made in production. His daily and ten-day reports had also improved the timeliness of statistical data—once members of the cooperative had reported to their teams, it only took him a day to produce summary reports on such things as amount of cultivated land or the number of livestock.

Yang also compiled a couple of production handbooks 生产手册 for the team based on the production plans of the work team.651 These handbooks covered such things as how much area should be allocated for different crops, instructions for calculating how

650 BMA 035-002-00203: 43.

651 BMA 035-002-00203: 44.

263

much human and animal labor were required per mu of land and how much fertilizer to use per mu. In addition, he also delineated every job’s work point assignment. He gave copies of each handbook to leaders of small teams. In this way, team leaders were able to effectively and efficiently calculate each workers daily work points. Yang asserted that such changes in practice not only helped reduce production losses, they also helped avoid the nightly wastage of time when everyone gathered to discuss the allocation of work points. The small-team-leader was so pleased with the outcome that he went to the extent of giving Yang’s handbook the epithet ‘teacher’ 老师.652

The second statement was filed by Liu Changming 刘长明, the statistician of the

Number Seven Work Team of the Heping Agricultural Production Cooperative 和平农业

生产合作社, located in Heping Village 和平乡 in the Eastern Suburbs of Beijing 东郊

区.653 In 1954, Liu had left the Army and returned to the village and begun participation in agricultural work. The following year he took on the joint responsibilities of cashier 出

纳 and work points recorder 记工 for the South Lake Drainage Cooperative 南湖渠社.654

At that time, Liu noted that the entire team’s work point distribution had not been checked for errors in a very long time. In order to correct this state of affairs, he focused on professional learning and worked with his subordinates so that all work-point recorders became comfortable with how to record work points. He taught all work point recorders in the cooperative and firmly established in their minds the importance of their work. In addition, whenever a meeting was organized, he would ask leaders of the

652 BMA 035-002-00203: 44.

653 Li’s statement was the eleventh in the compilation. The following paragraphs are drawn directly from his statement, the text of which can be found in: BMA 035-002-00203: 73-75.

654 BMA 035-002-00203: 74.

264

cooperative to invite other leaders to participate. This raised the meeting’s profile and made all attendees participate more seriously. Such a strategy had the added benefit of solving on-the-spot any difficulties that arose.

In keeping with the Mass Line, Liu stressed that recording work points must also rely on the leadership of the masses. He would frequently go down to the grass-roots level to understand the actual problems that work-points recorders faced and helped them solve their problems. Liu also set up a team-wide system of mutual checks and comparisons. Good practices would receive praise, and the bad ones were criticized.

Every ten days the work-point recorders and the members of the cooperative got together to reconcile labour manuals so that there was no incongruity between the logging of work points and the actual work that had been carried out. Every month, after reconciling with the entire team, Liu and his fellow points recorders also put together a work points account book .The recorders also took advantage of days and times when inclement weather made farm work impossible and organized fora 座谈会 where work-point recorders could jointly study methods for recording work points, share experience, and thereby improve the quality of their work. It was as a result of such measures that the recording of work points was no longer error prone.

The institution of work points thus served to invert the relationship to numeracy that we may have come to expect. As the power of the state expanded enormously over the past few centuries, scholars such as James Scott have located modes and sites of resistance to state intrusion.655 In a recent workshop paper, Jacob Eyferth sought to add to

Scott’s strategies of ‘a-literacy’ and ‘a-historicity’ by proposing ‘a-numeracy’: not

655 See in particular, James C. Scott. The Art of Not Being Governed—An Anarchist History of Upland Southeast Asia. New Haven: Yale University Press, 2009.

265

knowing how to count was a way of keeping the state at bay during the tumultuous transformations taking of 1950s China.656 At the same time, the discussion of work-points in this chapter indicates that as higher-stage cooperatives became the de facto reality in the rural countryside, a-numeracy may have transformed into a new form of enumerative consciousness. As every day was reduced to a single nominal digit which represented one’s entire income, it became increasingly impossible to claim ignorance or disinterest vis-à-vis numbers and quantification.

Conclusion

Socialist statistics was central to both the way in which the state approached planning and to the changes that were wrought in the countryside. Complete statistical information demanded unmatched levels of cooperation and data sharing across all levels of the planning and statistical apparatus, even as at the grass-roots level the impetus to quantify each and every basic activity in the rural country-side generated enormous social tensions as numbers in the most basic form of work points came to occupy an increasingly central position in the lives of peasants.

656 Jacob Eyferth. “Learning to Count, Learning Not to Be Counted: Numbers, Governance, and Resistance in Rural China.” Unpublished Paper Presented at the Workshop on “Nation and Enumeration in Modern China,” Columbia University, New York, 26 October 2013.

266

PART III

ALTERNATIVES

267

CHAPTER VIII

The Turn to India: Sino-Indian Statistical Exchanges, 1951-1959

268

Sunday, 9 December 1956. Chinese Premier Zhou Enlai’s visit to the Indian Statistical

Institute (ISI) in Calcutta had not gone entirely according to schedule. The original itinerary called for a visit of half an hour. Strenuous protestations by P.C. Mahalanobis, the ISI’s director, had resulted in a further 45 minutes being granted. But the Premier’s party arrived almost a half hour late, throwing the carefully planned schedule into disarray. Various items on the agenda had to be jettisoned.657 Zhou himself did not help matters by getting deeply involved in several displays, asking questions, seeking clarifications. In the National Sample Survey Unit, he sat down on a table and refused to move until his questions had been satisfactorily answered. By the time he finally emerged out in the open, several minutes behind schedule, the West Bengal officials including the

Chief Minister Dr. B.C. Roy, who had accompanied Zhou through his tour of the Institute, were all in a hurry to depart. Suddenly, Zhou caught hold of Mahalanobis’ arm, “side- stepped the waiting car, waved his hand to shake off everyone,” and took him aside.

“What are the countries most advanced in statistics? Are you in touch with them?”

Zhou inquired. Mahalanobis listed the United Kingdom, the United States and the Soviet

Union, noting further that while they had been in touch with all three, it was not desirable to copy from any country and that “in India we are trying to adopt and develop the methods to suit our own needs.” Zhou nodded his head in vigorous approval. He observed that a group of Chinese statisticians would be visiting the ISI very soon. “I want them to see everything in detail. We want to learn from you. How long should they stay?..I shall tell them to stay as long as you think it advisable...Also, we want to send research scholars to work here. What is the kind of men who should come?” Mahalanobis

657 The program for his visit included short speeches on the grounds of the Institute, where Zhou also planted a Mango tree, before proceeding to various departments. For details see: Chinese Foreign Ministry Archives (WJB) 203-00084-01: 118.

269

told Zhou that he wanted a mix of types, some advanced and some with good experience of practical work. He also wanted an economist or two who could share China’s planning experiences. Zhou again approved. “Yes, we must have exchange of men.” On the question of how long the research scholars should stay at the ISI, Mahalanobis suggested a period of at least six months and preferably one year, with some staying longer as necessary. Zhou was determined: “We shall arrange this. I shall tell them that they should consult you, and you should make the final decision.” He concluded the conversation by taking Mahalanobis’ two hands in his and stating very firmly: “And next year in

Peking.”658

The exchange may have had a redemptive quality for Mahalanobis. For long, he had tried to engage with the Chinese on matters of planning and statistics, only to be rebuffed. But now, things seemed different. And Zhou’s urgency was real, the results tangible. On December 11, 1956, just two days after Zhou Enlai’s visit, a team of

Chinese statisticians arrived at the ISI. Ostensibly to participate in the silver jubilee celebrations of the Institute, they stayed for a month. Led by one of deputy directors of the State Statistics Bureau (SSB) whom we have already encountered in Part I, Wang

Sihua, the four-person delegation also included the noted demographer, statistician, and

People’s University professor, Dai Shiguang 戴世光 (1908-1999). Within months of returning to Beijing, deliberations over inviting Mahalanobis to China began. “And next year in Peking,” did indeed become a reality. Mahalanobis, accompanied by his wife and by the statistician D.B. Lahiri, spent three weeks in China during the months of June and

July 1957. The two statisticians lectured to large audiences at the Planning Commission,

658 P.C. Mahalanobis Memorial Museum & Archives (PCMMMA), Correspondence with Shri Pitambar Pant File No. 915, 1953-56: 11-12.

270

Statistics Bureau and various universities, held numerous small group discussions with statisticians and planners, and planned future cooperation and exchange. Later that year, two Chinese statisticians, Wu Hui 吴辉 from the Industrial Statistics Division and Gong

Jianyao 龚鉴尧 from the Agricultural Statistics Division, arrived at the ISI and spent over an year in India studying statistical theory and its applications.

* * *

Ever since the establishment of the People’s Republic of China on October 1 1949, the

Chinese had hewn close to the Soviet Union as a role model, regarding it as an older brother with advanced experience [先进经验 xianjin jingyan], who needed to be learned from and emulated. Made famous by Mao’s phrase of “leaning more to one side,” (一边

倒 yibian dao), the practical implications of this policy had been the disregard for those kinds of relationships and gatherings that fell outside of this special mentoring program.

That the People’s Republic was in the Soviet camp fits a well understood Cold-War paradigm. The world had two centers, with their own zones of influence, and they each vied for control and influence over the vast regions they did not influence. In the immediate years following the end of World War II, the rapid unraveling of empires and the new emergence of nation states provided them with a playground for this exercise in influence. This center-periphery framework has come to dominate recent studies of the period, and does not always account for periphery-periphery links and what they can tell

271

us.659 As a result, studies of scientific exchange in the post-1945 years have frequently focused on West-centered or Soviet-centered networks at the neglect of regional- or self- avowedly “Third World” links such as the Sino-Indian exchanges of the 1950s.

As we have seen in Parts I and II, the story of statistics in the People’s Republic also appears to follow the general pattern: if Western involvement and influence in the spread of (social) sciences in China is one of the characteristics of the Republican period

(1911-1949), then Soviet presence is often cited as a key factor in the evolution of

(statistical) science in the immediate post-1949 years. But, as this chapter demonstrates, this fits a neat Cold War framework that elides even as it explicates. Starting as early as

1951, and with increasing urgency after 1956, Chinese and Indian statisticians and planners traded visits as they sought to learn from each other’s experiences. The later exchanges, in particular, were qualitatively and quantitatively different from those in

1951, and help shed light not only on a particular instance of scientific cooperation and knowledge production outside of the Cold War framework but also allow us to better appreciate the history of statistics in the early People’s Republic. In so doing, they also permit a more grounded understanding of the possible causes for the break-down in Sino-

659 There is a growing literature on the history of development and aid in the post-WWII era, much of it through the prism of US or Soviet centered networks of aid and influence. See for instance: Nick Cullather 2010. The Hungry World: America’s Cold War Battle Against Poverty in Asia. Cambridge: Harvard University Pres, 2010; David Ekbladh. The Great American Mission: Modernization and the Construction of an American World Order, 1914 to the Present. Princeton: Princeton University Press, 2010; the materials at the Cold War International History Project at the Wilson Center; Thomas P. Bernstein and Hua-yu Li, eds. China Learns from the Soviet Union, 1949-Present. New York: Lexington Books, 2010; Daniel Immerwahr. The Village Ideal: The United States, Community Development, and the World. Cambridge, Mass.: Harvard University Press, 2014; and Odd Arne Westad, ed. Brothers in arms: the rise and fall of the Sino-Soviet alliance, 1945-1963. Stanford: Stanford University Press, 1988. For an insightful critique of the scholarship on the history of north-south scientific exchanges in the post- World War II years, see: Alexis De Grief A and Mauricio Neito Olarte. “What We Still Do Not Know About South-North Technoscientific Exchange—North-centrism, Scientific Difussion, and Social Studies of Science.” In The Historiography of Contemprary Science, Technology, an Medicine—Writing Recent Science, edited by Ronald E. Doel and Thomas Söderqvist, 239-259. London: Routledge, 2006.

272

Soviet relations, moving us beyond causes/arguments that can be located at the highest levels of politics and in the egos of leaders.

At the heart of the Sino-Indian exchanges was the desire to learn more about mathematical statistics and its applications, particularly the cutting-edge statistical method of random sampling, which, while technically complex, held great practical salience for large and diverse countries such as China and India. This chapter traces these exchanges, explaining their timing and the motivation behind them. The story of these exchanges provides a specific and highly illuminating case of the growing disaffection that the Chinese experienced with a strategy of overt reliance on the Soviet Union for technical and scientific advice. At the same time, these exchanges also highlight the possibilities of south-south cooperation and exchange that became particularly appealing after the April 1955 Afro-Asian Conference in Bandung. Each set of actors in these exchanges had their own agendas. The Indians were particularly keen to learn more about

China’s planning methods. The Chinese, in turn, were much more focused on learning about statistical theory and methods, and more broadly, about an approach to statistics, which for political and ideological reasons had been rendered anathema.

P.C. Mahalanobis and the Promise of Post-War Statistics

As we have seen in Part I of this dissertation, the first half of the 20th century was a time of tremendous statistical innovation, benefiting in particular from genuine attempts at melding theory with practice in order to solve the pressing concerns of the day. To review briefly, until the end of the nineteenth century, the evolution of statistics had progressed

273

on two parallel tracks, one descriptive and the other analytic.660 The descriptive traced its origins to the “political arithmetic” of the seventeenth century, and through it to a much more fundamental and age-old phenomenon—the desire of the state to collect information about its constituents, its land, its produce, and so on. The analytic can be traced at least as far back as the 16th and 17th centuries and was principally concerned with concepts of chance, probability, and estimating and controlling error, all in the abstract , mathematical, sense. Starting in the 19th century, these tracks began to intersect.

Statistical methods were brought to bear on the most pressing social concerns of the day.

This cross-pollination led to tremendous innovation, resulting in many of the concepts that form the bedrock of modern day statistics—correlation, variance, regression, random sampling, and so on.

By the 1940s, statistics, operations research, and other numerically driven methods of understanding and organizing different aspects of day-to-day life were increasingly gaining legitimacy. A somewhat well known example of this phenomenon is found in the career of the Kennedy and Johnson era Secretary of Defense Robert

McNamara, who during WWII worked in the Air Force’s Office of Statistical Control, applying statistical and management theory (the ‘Harvard Method’) to enhance the efficiency of B-29 bombers.661 The end of the Second World War witnessed new

660 P.C. Mahalanobis. “Why Statistics?” Sankhya: The Indian Journal of Statistics 10, no. 3 (September 1950): 195-228. Quote from: 195-96.

661 As Phil Rosensweig noted in a 2010 essay, “perhaps more than anyone else, Robert McNamara personified management in the 20th century. In his legacy we see the triumphs of modern management as well as its most troubling limitations.” McNamara’s obsession with statistics is well documented. He put his skills to good use during World War II and later in the 1950s by helping resuscitate the ailing Ford Motor Company. During the Vietnam war this obsession with ‘statistical control’ led to the overt reliance and favoring of numbers and models over information coming in from other, in retrospect, more reliable sources, thereby contributing to the prolonging of the war. From 1968 to 1981 McNamara also served as the President of the World Bank. See Phil Rosensweig 2010. “Robert S. McNamara and the Evolution of Modern Management.” Harvard Business Review, December 1, 2010. On McNamara’s leadership during

274

attempts at global organization, and coincided with the post-colonial wave of newly independent countries. Statistics came to be viewed as a key tool for national and international development and stability. Newly formed nation states sought to establish statistical bureaus, collect data, and address issues of governance and growth. At the global level, newly formed international organizations like the United Nations and the

World Bank sought to render such data commensurable by establishing norms for its collection, collation, and analysis.662

Prasanta Chandra Mahalanobis was in many ways among the most articulate spokesmen for the promotion and use of statistics. “A physicist by training, a statistician by instinct and a planner by conviction,”663 Mahalonobis melded scientific and intellectual heft with a tremendous drive to engage in practical matters, and to use statistics to understand and address the problems of every-day life. As a result, he became intrinsically involved not just with independent India’s first forays into economic planning and development, but perhaps even more significantly, with the development and establishment of statistics, most notably sample survey techniques, as a principal tool for knowing the world.664 Mahalanobis’ original training had been in physics, and it was a subject he continued to teach into the 1950s. But a chance encounter with statistical journals in a Cambridge University library led to his first exposure to statistics. By the

the Vietnam war years, also see David Halberstam. The Best and the Brightest. New York: Random House, 1972.

662 For a history of the UN and its role this process, see: Michael Ward. Quantifying the World: UN Ideas and Statistics. Bloomington: Indiana University Press, 2004.

663 C.R. Rao. “Prasantha Chandra Mahalanobis, 1893-1972.” Biographical Memoirs of Fellows of the Royal Society 19 (December 1973): 454-492, 455.

664 A useful overview of Mahlanobis’ polymath activities and achievements is the above cited C.R. Rao biographical memoir. For a longer treatment also see: Ashok Rudra. Prasanta Chandra Mahalanobis: A Biography. New Delhi: Oxford University Press, 1996.

275

mid-1920s he had abandoned plans to return to Cambridge to continue a PhD in physics, choosing instead to stay in Calcutta and pursue various projects whereby he could apply statistical methods to practical problems. In 1932, along with a few likeminded individuals, he established the Indian Statistical Institute (ISI) in Calcutta. Initially operating out of a tiny room at Presidency College, by the 1950s the ISI had moved to a sprawling campus in the north of the city and evolved into one of the premier centers for statistical research and application across the globe.665

In his Presidential Address to the thirty-seventh session of the Indian Science

Congress in January 1950 Mahalanobis had astutely observed that, “at each upsurge of social and political development, or during war, there is a rapid growth and expansion of statistical practice.”666 Later in the same speech, he added:

Statistics is not only an applied science but is also a public science. It is because of the close connexion with public activities that big developments in statistics have always occurred only when there has been need of unified policy and co-ordinated action in times of war or peace.667

665 In an interview, long time ISI professor C.R. Rao recalled getting an offer in 1946 “from Cambridge University, UK, to do statistical analysis on some skeletal measurements, based on methods developed in India.” He added he “was glad to accept the offer considering it as transfer of technology from an underdeveloped country to an advanced country.” See: “ISI Honorary Member Interviews: Professor C.R. Rao,” International Statistical Institute Newsletter, Volume 29, No. 2 (86) 2005. Accessed online on March 9, 2012 at: http://isi.cbs.nl/Nlet/NLet052.htm#09HonoraryMemberInterviews. Rao got his PhD in statistics under Ronald A. Fisher at Cambridge before returning to the ISI and carving out a distinguished career.

666 Mahalanobis 1950: 196.

667 Mahalanobis 1950: 211. “In statistical research the greatest stimulus has always come from the need of solving practical problems. R.A. Fisher’s work on the design of experiments was due to the urgent need of solving the deadlock in agricultural field trials. W.A. Shewhart’s work on quality control arose from the need of improving the efficiency of inspection in large scale production.” Mahalanobis went on to provide three examples to prove his point: Post-New Deal United States where definite statutory responsibilities were assigned to the Division of Statistical Standards; the exigencies of war time planning in the UK, which led to the formation of the Central Statistical Organization; and the USSR, where the Central Statistical Bureau was an integral part of the national planning organ, GOSPLAN.

276

And, to his mind, the 1940s and 1950s were precisely such a time. War time exigencies had spurred the application of statistical theory to practical problems, and now in the post-war world, statistics had a foundational role to play. In 1947 he brought his considerable expertise, acumen, and drive to the UN’s newly formed Statistical

Commission and served as the chairman of its Sub-Commission on Statistical

Sampling.668 Mahalanobis was the driving force behind the Sub-Commission. “With the vision of a pioneer he perceived how sampling methods could overcome the otherwise insoluble problems of obtaining accurate information about a vast and still largely illiterate country like India”669

Sampling, of course, has a long history. In his 1950 Presidential Address

Mahalanobis had explained that its twentieth century formulation was a combination of the long history of purposive sampling with that of the recent developments in probability and mathematics. Purposive sampling is a form of sampling in which the samples chosen for analysis are selected because of some specific characteristic they possess. Today, we would identify it more closely with qualitative research, not quantitative.670 For a long time it was regarded as the only proper method of sampling and as late as 1926 it was given considerable prominence by a committee of experts of the International Statistical

668 A part of the UNESCO, the Commission is the “apex entity of the global statistical system. It brings together the Chief Statisticians from member states from around the world. It is the highest decision making body for international statistical activities especially the setting of statistical standards, the development of concepts and methods and their implementation at the national and international level.” http://unstats.un.org/unsd/statcom/commission.htm accessed February 29 2012.

669 Jelke Bethlehem. “The Rise of Survey Sampling.” Discussion Paper (09015). The Hague: Statistics Netherland, 2009.

670 To put it crudely, imagine, for instance, that we were to select Morningside Heights for intensive statistical analysis, and then proceed to claim that the results reflected the larger reality of Manhattan.

277

Institute.671 But since then evidence had been consistently mounting that it produced results that were neither representative nor accurate. The method of complete enumeration was also brought under increased scrutiny. In a 1950 paper in the Bulletin of the International Statistical Institute Mahalanobis had demonstrated “that with an acceptable level of precision the costs of sample surveys are only about ten percent of that of a complete enumeration.”672 In the same issue, Mahalanobis’ friend, collaborator, and colleague on the sub-commission, the statistician Ronald A. Fisher, went even further and stated that not only were sampling procedures ideal because of their adaptability, speed, and economy, but that they were also more scientific than complete enumeration.673 Mathematical theory undergirded sampling, and this permitted the

“careful design of sample surveys with a preset level of precision.”674

“Large-scale sample survey techniques as practiced today,” declared the statistician C.R. Rao in his 1973 memoir of Mahalanobis for the Fellows of the Royal

Society, “owe much to the pioneering work of Mahalanobis in the forties and fifties.”675

According to Rao, Mahalanobis made three notable contributions to sample survey

671 Rao 1973: 472.

672 Bethlehem 2009: 16.

673 Ronald Aylmer Fisher (1890-1962) was statistician, geneticist, evolutionary biologist, whose contributions in statistics include the analysis of variance (ANOVA), the method of maximum likelihood, the derivation of various sampling distributions, and experimental design. As a result of these contributions, many regard him as one of the principal creators of modern statistics. For more, see: F. Yates and K. Mather. “Ronald Aylmer Fisher 1890-1962.” Biographical Memoirs of Fellows of the Royal Society, 9 (1963): 91–129.

674 Bethlehem 2009: 16.

675 The mathematical statistician Harold Hotelling observed: “No technique of random samples has, so far as I can find, been developed in the United States or elsewhere, which can compare in accuracy with that described by Professor Mahalanobis.” Fellow statistical luminary R.A. Fisher noted: “The I.S.I has taken the lead in the original development of the technique of sample surveys, the most potent fact finding process available to the administration.” See: Rao 1973: 472.

278

techniques: pilot surveys, the concept of optimum survey design, and interpenetrating network of samples.676 He developed these methods during the 1930s and 1940s, starting initially by using sampling procedures to estimate jute yield in Bengal in 1937. Over the next two decades, along with his colleagues at the Indian Statistical Institute, he carried out numerous sample surveys, developing new procedures and fine-tuning old ones.

Mahalanobis’ contributions were not limited to the development of theory and method and he also assisted in establishing the framework and standards through which they could be used across the world. As Chairman of the UN Sub-commission on Statistical

Sampling, he oversaw the publication of “The Preparation of Sample Survey Reports,” which continues to provide the common guidelines for modern sampling procedures in use by statistical bureaus across the world today.677

Mahalanobis’ interest in statistics was almost entirely instrumental, a point he made most forcefully in his essay “Statistics must have a purpose.”678 This desire to use statistics to solve real world problems also got him involved in economic planning by the

676 Rao 1973: 473. For a discussion of Mahalanobis’ critics and alternative viewpoints about best sampling procedures during this time in India, see: A.P. Gore 1997. “P. V. Sukhatme: A 'Social' Statistician.” Economic and Political Weekly 32, no. 6 (Feb. 8-14, 1997): 257-258.

677 United Nations Statistical Commission. “The Preparation of Sampling Survey Reports.” Statistical Papers, Series C, No. 1. Lake Success, New York, 1950. Members of the first Sub-Commission included statistical luminaries such as G. Darmois, W.E. Deming, F. Yates, and R.A. Fisher. A follow-up report was published in 1964, with Mahalanobis serving as the Sub-commision’s Vice-Chairman. See: United Nations Statistical Commission. “Recommendations for the Preparation of Sampling Survey Reports (Provisional Issue).” Statistical Papers, Series C, No. 1, Rev. 2. New York, 1964. Members included F. Yates (Chairman), T. Dalenius, N.Keyfitz, V. Monakhov, and P. Thionet. S. Zarkovic participated as a representative of the Food and Agriculture Organisation, and D. B. Lahiri of the Indian Statistical Institute was present as an observer. P. J. Loftus, Director of the Statistical Office, represented the Secretary- General and M. D. Palekar served as secretary of the Group, p. v.

678 “Statistical data may be required to solve problems of a practical nature or for purely scientific or theoretical investigations. Some of the data may be required for current use; some may have to be collected for future utilization. The point to be stressed, however, is that statistics must have some purpose, practical or theoretical, and for immediate or future use.” P.C. Mahalanobis “Statistics Must Have a Purpose.” Samvadadhvam 1, no. 1 (July 1956): 3-10, 4.

279

late 1940s. His closeness to India’s Prime Minister, Jawaharlal Nehru, led to an appointment on the Planning Commission, and before long he was charged with drawing up India’s second Five Year Plan (1955-59). In preparation for this work Mahalanobis spent the summer of 1954 traveling through the United States, the United Kingdom,

Eastern Europe, and the Soviet Union, conferring with economists, statisticians, and planners.679 A two-sector growth model that he had created in 1953, similar to G.A.

Feldman’s 1928 model, became the basis for India’s second Five Year Plan.680 Earlier in the decade, he had succeeded in bringing the biennial Conference of the International

Statistical Institute to India. The Institute’s twenty-seventh session was held in Delhi in

December 1951.681 With the exception of Japan, which had played host in 1930, this marked the first time that an Asian or African nation had hosted the prestigious conference. By the 1950s Mahalanobis had also built the ISI up to be one of the leading centers for statistical innovation in the world. It was a magnet for all kinds of scientists, from either side of the iron curtain and beyond the curtain.682 And with the help of the

United Nations (UNESCO), in 1950 he had also set up at the ISI in Calcutta an

679 Mahalanobis was also seeking technical and material aid in statistical and planning matters, in particular computers and related technology. See: Papers of Pitambar Pant (190 LII), Nehru Memorial Museum and Library, New Delhi.

680 On alternatives to the Mahalanobis model and on contemporary debates surrounding planning strategies, see: Meghnad Desai Desai, Meghnad. “Development Perspectives: Was there an Alternative to Mahalanobis?” In India's Economic Reforms and Development: Essays for Manmohan Singh edited by Isher Judge Ahluwalia, 43-52. New Delhi: Oxford University Press, 2012.

681 Mahalanobis had tried to bring the Conference to India in 1949, but Switzerland’s invitation was favored instead. India was assured that its invitation would be given serious consideration for 1951. For more see File “EA/UN-I/9(58)-UNI, 1949,” at the National Archives of India (NAI), New Delhi.

682 The ISI Annual reports carry long lists of visiting scientists of varying disciplinary and ideological persuasions. Mahalanobis himself was a great traveler and the ISI Annual Reports usually also devoted a few pages to his global travels. By the end of the decade, some exasperated friends and colleagues began to worry that the administration of the ISI was suffering as a result. In the early 1960s, this issue even drove J.B.S. Haldane, then a full-time professor at the institute, to quit in protest. See: Ramachandra Guha. An Anthropologist Among Marxists and Other Essays. New Delhi: Permanent Black, 2001, 134-135.

280

International Statistical Education Centre (ISEC), the principal purpose of which was to train statistical workers and students from Asia and Africa. By 1956, students from fourteen to fifteen countries were receiving training at the Center.683

Initial Overtures and Estrangement

The People’s Republic of China had largely stayed outside this newly globalizing world of statistical exchange. That this was so was not for want of trying on the part of

Mahalanobis. He was acutely interested in what the Chinese were doing. In the summer of 1954 he had confided to Pitambar Pant:

I had two talks with Blackett [P.M.S. Blackett, left-leaning English physicist] and on one occasion he mentioned that he could foresee that in 10 years time India would be holding her own or even somewhat better her conditions of life but he thought China during the same time would go very far ahead; and he asked me whether this was not likely to have a profound influence on India. This, of course, is always on the back of my mind.684

And he was keen to visit China as well. Later that summer, while in the Soviet Union, he wrote again: “I must also go to China – but after our ideas have become more clear

[sic].”685 Mahalanobis’ efforts at engaging with the People’s Republic had begun when

India hosted the twenty-seventh biennial session of the International Statistical Institute in 1951. As the General Secretary of the National Committee charged with organizing the

683 “Professor Mahalanobis Welcomes Chou En-Lai.” Samvadadhvam 1, no. 3 (March 1957): 30-31.

684 Written on June 24th 1954. Papers of Pitambar Pant (190 LII), Nehru Memorial Museum and Library (NMML), New Delhi, p. 14. In the same letter, Mahalanobis also mentioned a meeting with the economist Joan Robinson, who too was deeply impressed by China’s progress, having visited there the previous year. This letter is part of a long series of letters that PCM wrote to Pant during his global travels in the summer of 1954.

685 Papers of Pitambar Pant (190 LII), p. 91, NMML, New Delhi.

281

conference, he oversaw the delivery of official letters of invitation to seventy-one countries, including the People’s Republic of China. Individual invitations were also sent to four Chinese members of the International Statistical Institute: Chen Da 陈达, J.P. Chu

[Zhu Junyi 朱君毅], K.P. King (Jin Guobao 金国宝), and T.L. Woo. These are familiar names; prominent Republican era statisticians, they had attended the 1947 meeting in

Washington, D.C. and we have already encountered them in Part I of this dissertation.

The invitees were also informed that a special session of the International Statistical

Association for Asia and the Far East (ISAAFE) would be held shortly after the ISI

Conference with the aim of:

(a) Converting the ISAAFE into a wider Asian Statistical Institute with both individual and governmental and institutional membership and working partly at an individual and partly at a governmental level and (b) considering statistical and economic problems of interest to Asian Countries.686

Chinese participation was “earnestly” sought. It was hoped that a delegation from China would be sent, failing which, perhaps someone from the Chinese Cultural Delegation, then already visiting India, could be requested to attend.

There appears to have been no response from the Chinese government or from the individual invitees. Mahalanobis was forced to resort to other means. Shortly afterwards, at a party for ambassadors hosted by Nehru, he engineered a chance encounter with the

Chinese Ambassador and urged the Chinese to attend.687 Within a few days a new set of invitations arrived at the Chinese Embassy in New Delhi. The ambassador deputed Shen

686 WJB 105-00229-01: 7-8. Additional material on the ISAAFE, including the Calcutta sessions, can also be found in “File 393-International Association for Asia and the Far East,” PCMMMA, Kolkata.

687 Evidence of Mahalanobis’ doggedness exists only in a short undated report, which provides a quick overview of Chinese participation in the 1951 ISI Conference. See: WJB 105-00229-01: 2-3.

282

Jian 申建, a counselor in the Embassy, to meet with Mahalanobis. By 21November, So- wan Pitt, the First Secretary in the Embassy was able to inform Mahalanobis that the

Chinese government had accepted the invitation, and that “one or two members of the

Chinese Cultural Delegation, which is now visiting India, will be instructed to participate in it.”688 A little over a week later, So-Wan Pitt confirmed the participation of Ti Chao- pei (Di Chaobai 狄超白), a member of the Chinese delegation already in India.689

As it happened, Di was only able to attend the first day of the Conference. The

Chinese delegation was on its way to Burma and he was forced to leave Delhi the very next day, on the 6th. To compensate, he drafted a short statement which he wished to have either read out or distributed during the Conference. In keeping with our discussion in Part I, this statement laid out the general view of statistics as it was taking shape in the

PRC at that time:

To us, statistics is one of the important means in the work of social construction and scientific researches. At the same time, we are of the opinion that the theories, methods and systems in connection with statistics adopted in a country cannot but be closely linked with the social system of the country in question; that only in a country where planned economy is practicable that statistics could be accessible to the work and daily life of the people and made for the people, its scope could be broad, correct, timely, and its utilization the maximum.690 (Original in English)

With the jibe at the notion of a universal statistics successfully delivered, he lauded the attempt at mutual understanding and exchange of statistical theories and methods that the

688 WJB 105-00229-01: 15.

689 Di Chaobai (1910-1978) was an economist who had taught at various universities prior to 1949. In July of 1948 he was appointed the director of the department of statistics 统计处处长 of The Central Financial and Economic Committee. In 1953 he was the head of the department of comprehensive statistics at the SSB. See: DSJ: 1.

690 WJB 105-00229-01: 22.

283

Conference represented. Yet, he felt that “all these tasks could be satisfactorily fulfilled only after the realization of a genuine international peace and the extinction of the aggressive designs of imperialism.”691 Di made no mention of the Calcutta sessions but an undeterred Mahalanobis responded the very same day, thanking Di for his participation and contribution, and reiterated his desire that China, in spite of Di’s unavailability, be represented at the Calcutta sessions:

“We are particularly eager that the People’s Republic of China should be represented at the speical [sic] meetings which will be held in Calcutta on 18 and 19 December to consider the organization and programme of work of a regional international statistical association or institute for Asian countries…I hope this will be possible.”692

Within a few days Mahalanobis was informed that Mr. Tai Ping (戴平 Dai Ping), the

Consul at the Calcutta Consulate-General, would attend the Calcutta sessions.693

The Chinese appear to have come away unimpressed from the Calcutta sessions.

They had likely been wary of the whole Conference right from the time the invitations first arrived. And their fears may have found justification as early as the inaugural session of the Conference on 5 December itself, when India’s Finance Minister, Chintaman

Deshmukh, made the following observation toward the end of his (keynote) address:

The International Statistical Association for India and the Far East, (which was established in 1949-50 through the efforts of the trainees from 10 or 12 countries who assembled in Delhi in [sic] training centre in population and agricultural censuses under the auspices of the United Nations and the Food and Agricultural Organization), is arranging to have a special session on the 18th and 19th December in Calcutta to considers its future

691 WJB 105-00229-01: 23. For good measure he also added: “To fellow statisticians from Asian countries who have shown us friendliness and co-operation, I wish to extend our respect and regards.”

692 WJB 105-00229-01: 24.

693 WJB 105-00229-01: 26.

284

organization and programme of work. The Government of India appreciate the benefits which would be secured through the building up of an active International Statistical Centre for the Asian countries, and would give their support to any suitable programme of work. Such an organization should prove a valuable adjunct to the Colombo Plan.694

UN auspices for the Association could not have been viewed favorably by the Chinese who were themselves excluded from the UN in favor of the government of Chiang Kai- shek (Jiang Jieshi 蒋介石) on Taiwan. But references to the Colombo Plan would also have raised flags. Its full name was the Colombo Plan for Cooperative Economic and

Social Development in Asia and the Pacific and it had been conceived at the

Commonwealth Conference on Foreign Affairs held in Colombo, Ceylon (now Sri Lanka) in January 1950. It was formally launched in mid-1951 and by the end of the year the

United States had also become a member country.695 As a result, it would not have been difficult for the Chinese to construe the ISI Conference and the Association, which was renamed the International Statistical Association for Asia and the Far East during the

Calcutta sessions, as ideologically compromised.696

In a series of telegrams exchanged between the Calcutta Consulate General, the

Embassy in Delhi, and the Foreign Ministry in Beijing that date from December 1951 through March 1953, just such an understanding was arrived at. By the end of April 1952 all documents and reports pertaining to the Conference sessions in Delhi and Calcutta had

694 WJB 105-00229-01: 36. This, Di no doubt would have heard, given that he was present only on this one day of the Conference.

695 For more visit: http://www.colombo-plan.org/ .

696 It is worth noting that the Korean War was raging during these months and that the US had a prominent role to play in the ISI. The 1947 session of the ISI, its twenth-fifth, had actually been held in Washington, D.C.

285

been delivered to the Foreign Ministry in Beijing.697 It took another six months before an initial opinion was forthcoming. In the considered judgment of the Ministry, it was felt that the Chinese should “in principle not plan to maintain links with the [said] organization.”698 Early the following year, further explication was put down in writing:

The [stated] Association has extremely close ties with the international statistics society, operates under the influence and control of imperialism, and there is a great probability that in the name of learning and science it will be used to conduct intelligence [related] activities.699

Two months later, in March 1953, clear instructions were issued that it was no longer necessary to treat with this Association or respond to its overtures.700 The People’s

Republic was much better off without involving itself in imperialist misadventures, especially when it had radically different views on the definition and purpose of statistics.

As discussed in Part I, a close and profitable relationship had already been established with the Soviet Union, and little was to be gained by dealing with what appeared to be suspect organizations like the ISAAFE or the two ISIs.701

697 The materials were sent out on 9 March 1952 by Mahalanobis, who was the President of the Association. See “File 393-International Association for Asia and the Far East,” PCMMMA, Kolkata.

698 “原子上不擬與該組織聯繫.” See: WJB 105-00229-01: 44.

699 WJB 105-00229-01: 45.

700 WJB 105-00229-01: 47.

701 In 1953 Mahalanobis had also tried to invite the Beijing University mathematical statistician P.L. Hsu (Xu Baolu 许宝騄; mentioned briefly at the end of Chapter IV) to spend several months at the ISI. The archival trail runs cold, but from what I can determine, P.L. Hsu never visited the ISI. For more, see: “File titled P.L. Hsu,” PCMMMA, Kolkata.

286

Statistics as a Case Study in Disaffection

By 1956, however, confidence in the Soviet Union and its advanced experience was no longer taken for granted. Even though recent scholarship has highlighted the tensions in the Sino-Soviet story, including specific cases where initial idolization of Soviet models was confronted by the reality of their execution or their sheer inapplicability to Chinese conditions, the causes of the disenchantment that began in 1956 are still largely attributed to two more over-arching reasons.702 As Hua-Yi Lee and Thomas P. Bernstein have observed in a recent edited volume, the general consensus in the scholarship is that the evolution of Sino-Soviet relations strongly influenced Chinese attitudes toward the Soviet model.703 Accordingly, the causes for the disaffection are still largely framed either as an ideological dissonance, or an egotistical one, or both. They frequently center on how Mao and the CCP leadership reacted to Khrushchev’s 1956 speech denouncing Stalin’s personality cult, on Mao believing that in the post-Stalin era he ought to be regarded as the leader of the Socialist world, and also on the reception of events like the quashing of

Hungarian revolt that year.704 In his January 1956 report, “On the question of intellectuals,” Zhou Enlai recognized other reasons as well. The report acknowledged the continuing central role that Soviet aid needed to play in order for China to advance in science and technology, but buried in the middle were also clear words of caution:

On the question of the study of the Soviet Union, however, in the past there had also been such defects as undue haste, arbitrary learning, and mechanical application. Some comrades even arbitrarily rejected the

702 Li Bernstein 2010.

703 Bernstein and Li 2010: 3. Also see: Shen 2003: 410.

704 Bernstein and Li 2010: 14-15. The latest and most comprehensive treatment of the Sino-Soviet split is Lüthi 2008. Also see: Westad 1998.

287

achievements of the capitalist countries in science and technique. These defects should henceforth be avoided.705

Zhou’s warning is redolent of the possibility that the disaffection was widespread and substantive. The articles in the Bernstein and Li volume have already suggested some bottom up causes for disaffection and frustration—most notably the claim that the advice being given by the Soviets did not take into account local Chinese conditions. But it is the case of statistics, perhaps, that provides one of the clearest instances of a widespread and substantive disaffection with Soviet advice and expertise.

To review briefly from Part I, the Chinese adoption of the Soviet approach to statistics occurred in the early 1950s. In the first couple of years after 1949, statistical treatises written by Chinese statisticians freely reflected, both, Soviet as well as what were subsequently identified as bourgeois English-American approaches to statistics.

Starting in 1950, and gathering momentum from the following year, Anglo-American statistics was progressively critiqued and repudiated in favor of the Soviet understanding of statistics. By the time the SSB was established in 1952, the Soviet approach held sway, both in the academy as well as in the practical fields of statistical work. This Soviet approach attained even greater influence in the People’s Republic following the 1954 conference on statistics in the Soviet Union. At this conference, statistics was reaffirmed as a social science, in service to building a socialist society, bifurcating it—rhetorically and substantively—from the tainted, bourgeois, and socially unproductive pursuit of abstract mathematical statistics.

705 Zhou Enlai 周恩来. “On the Question of Intellectuals.” In Communist China 1955-1959—Policy Documents and Analysis, edited by Robert R. Bowie and John K. Fairbank, 128-144, 137. Cambridge: Harvard University Press, 1963. To be fair, this appears to be the only cautionary instance in the long report, which otherwise goes to great lengths to stress how valuable and appreciated Soviet technical assistance is.

288

In the People’s Republic, the practical implications of this definition were two- fold. Within the academy, the study of statistics stood divided. The discipline of statistics, now defined as a social science, was taught in statistics departments where the focus was primarily on procedure, methodology, and work. Mathematical statistics, that

“unproductive” and bourgeois pursuit, was banished to mathematics department. As a result, much like in the Soviet Union, the theoretical and practical arms of statistics were separated, denying the possibility of mutual cross-pollination or the discovery of new theories and methods through application, experimentation, and testing.706 Outside the academy, in the world of statistical work, the impact was just as significant, and resulted in essentially two methods for gathering statistical information: (1) the establishment of a periodical report system based on complete enumeration wherein from the village level upwards, paper reports were compiled and supplied to each successive higher level till they all reached the headquarters of the State Statistics Bureau in Beijing; and (2) where complete enumeration was deemed impractical, sample surveys employing the typical sampling methods.707 Accordingly, by 1957, the SSB was able to summarize:

“For the purpose of providing statistical data for drafting and examining national plan [sic], the complete enumeration periodical report system has been established in all the economic departments in the country. To supplement these, a sampling body of sixteen thousand farm households and another of six thousand industrial workers families have been selected for survey purpose.”708

706 In his speech on intellectuals, Zhou would criticize this too: “It is certainly unquestionable that theory must not be divorced from practice and we must fight against any theoretical study which is dissociated from practice. But the main tendency at present is the neglect of theoretical study.” Zhou Enlai 1963: 141.

707 The Chinese is dianxing diaocha 典型调查 or zhongdian diaocha 重点调查.

708 See: WJB 105-00530-05: 11-12. “Complete enumeration periodical report system” is how the Chinese themselves translated “quanmian tongji baobiao zhidu 全面]统计报表制度.”

289

The sampling spoken of here was purposive (also called typical) sampling. As already noted, this is a method of sampling in which the samples are not selected using any random or stochastic method. It was the precursor to random sampling and had had a long history of being regarded as a valid and worthwhile method. Even into the early 20th century it was officially accepted in many places and, as such, its employment into the

1950s was not altogether unusual.709

Within a few years, the constraints imposed by the methods of socialist statistics became unmistakable. For a country with limited technical resources and a limited supply of trained personnel, yet saddled with an enormous agricultural sector and the largest population in the world, the challenges posed by complete enumeration were legion. And they were made harder still by the employment of the periodical report system. Paper reports, full of tables and summaries, were generated at each level of administration.

Very soon, not only Beijing, but every administrative level was complaining of a flood of papers and numbers, and little capacity to make any sense of them. Excess issuing (滥发 lanfa) of reports was a problem that drew the attention of all levels of the statistical system. Even though repeated attempts had been made to tackle the problem, it remained a constant feature of statistical work into the second half of the decade.

The reliance on purpose sampling also generated problems. Where complete enumeration was not deemed possible, such as in accounting for total area under crops or calculating agricultural yield, the method of purposive sampling had been and was being employed. As Mahalanobis would himself later explain:

An attempt is made to select sample-units (agricultural fields, households, etc) which are typical or have characteristics close to the average…It is

709 Jelke Bethlehem 2009: 8.

290

now recognized that the method of “purposive sampling” is not scientifically correct. The only sound method is “random or probability sampling” which can be relied upon to supply truly representative results; and which alone makes it possible to calculate valid estimates of the sampling error.710

Chinese statisticians were, thus, frequently stuck with estimates that did not accurately represent any reality on the ground. Claiming that a particular village or tract of land was representative of an entire prefecture, and using numbers gained from a close analysis of its characteristics, only lead to inaccurate estimates. These estimates, upon collation with other inaccurate estimates, helped compound errors, and could eventually result in data that bore little relation to any reality on the ground.

By 1956, socialist statistics had exhibited weakness on at least three counts. First, by not taking into account the greater challenges posed by China’s larger population and predominantly agrarian economy, it demonstrated a rigidity that rendered it incapable of adapting to local conditions. Second, the insistence on using purposive sampling meant that the technology and scientific theories being employed were not the most advanced, or to use Mahalanobis’ even more blunt term, simply not correct. Finally, and most damningly, the methods had not been able to deliver the desired results, something best articulated by Wang Sihua himself, who noted that statistical work remained unable to address the needs of the Planning Commission and the demands of the nation’s leaders.

It is precisely at this moment that the statistical exchanges with India were re- established. That the Chinese were willing to explore solutions in, and in cooperation with, India was perhaps no longer as inconceivable as it was in 1951. Frustration with

Soviet expertise and methods had already led Zhou Enlai to demand that the People’s

Republic not neglect advances made in capitalist countries. India was a useful middle-

710 P.C. Mahalanobis. Some Impressions of a Visit to China 19 June – 11 July 1957: 5. PCMMA, Kolkata.

291

ground, where the intellectual and policy environment was open to a variety of perspectives. And Sino-Indian relations, especially in the cultural sphere, had blossomed since the early 1950s.

Sino-Indian contacts: Wrestling for Asia’s Soul711

In 1953, Frank Moraes, then editor of the Times of India and recently returned from a tour of the People’s Republic of China, wrote:

“By a quirk of fate Asia’s two most densely populated countries, which are also neighbors, are the testing grounds for two differing and contending political philosophies. If China proves that her system of government ensures economic security to the vast mass of her people without detracting greatly from their sense of freedom, Asia will be lost to Communism. If India, on the other hand, demonstrates that democratic government can ensure not only economic security but individual liberty, then Asia will be won to democracy. What India and China are today doing is wrestling for the political soul of Asia. That is why India’s role in Asian affairs is crucial and vital.”712

The fate of the ‘political soul of Asia,’ as Moraes called it, captured the imagination of many minds in the 1950s.713 Not least among them were Indians and Chinese themselves.

There was mutual suspicion, but also much admiration, at least from several individual

Indian visitors. Unfortunately, the documentary record of this mutual appraisal, and of

711 By contact here, I mean to specifically exclude diplomatic engagements and exchanges, which would undoubtedly have been legion. In addition to individual archival and published sources as identified, I have relied in large part on the following two surveys to reconstruct Sino-Indian exchanges of the 1950s: (1) Margaret W. Fisher and Joan V. Bondurant. “Review Article: The Impact of Communist China on Visitors from India.” The Far Eastern Quarterly 15, no. 2 (February 1956): 249-265; and (2) Herbert Passin. “Sino- Indian Cultural Relations.” The China Quarterly, no. 7 (July-September 1961), pp. 85-100. For more on Sino-Indian contact during the first half of the twentieth century and during the nineteenth century, see: Madhavi Thampi ed. India and China in the Colonial World. New Delhi: Social Science Press, 2005.

712 Frank Moraes. Report on Mao’s China. New York: The MacMillan Company, 1953, 207.

713 This was very clearly how the two new superpowers and ideological metropoles perceived things. The discourse of ‘losing China’ was particularly lively in the United States, with the concomitant realization that India and other countries could not also be lost.

292

the varied contacts that formed at least part of its basis, is rather lop-sided. The plentiful and varied writings of numerous Indian visitors, ranging from die-hard Communists through Communist sympathizers to those openly skeptical or even disdainful of

Marxism, make it hard to pinpoint an “Indian” view of China.714 They do, however, suggest an always enthusiastic and at times also serious engagement. Chinese accounts, especially the impressions of individual visitors are harder to come by, making it difficult to gauge the nature and variety of views that individual Chinese may have entertained about India.715

A survey of existing materials suggests that Sino-Indian contacts during the decade of the 1950s had three distinct phases. India was the sixteenth nation to recognize the PRC on the international stage in early 1950.716 Diplomatic relations were established with India’s ambassador to the Republic of China, K.M. Panikkar, carrying on in the recently renamed and re-designated capital of Beijing after 1949.717 Even so, the first phase, which lasted through much of 1951 was marked by limited contact, fueled in part by the first disagreements over border demarcations, the status of Tibet, and by Beijing’s

714 Passin claims that as many as 2000 Indians may have visited China over the course of the decade. See: Passin 1961: 91.

715 It can be argued that there was a greater coherence from the Chinese side than the Indian. This is in part supported by the variety of delegations from India that visited China, with people with varying agendas, some openly hostile to China. Less is known of visiting Chinese delegations, partly because they were not as prolific in reporting on their travels as their Indian counterparts, but also because they all appear to have been official in nature.

716 “On 1 April, 1950, India became the first non-socialist bloc country to establish diplomatic relations with the People’s Republic of China.” See: “India-China Relations,” Government of India, Ministry of External Affairs. Accessed May 27, 2014 at: http://www.mea.gov.in/Portal/ForeignRelation/China_January_2014.pdf. Also see: Dieter Heinzig The Soviet Union and Communist China, 1945-1950: The Arduous Road to the Alliance. M.E. Sharpe, 2004, 292.

717 On Panikkar’s experiences as an ambassador during the early PRC, see: K.M. Panikkar. In Two Chinas: Memoirs of a Diplomat. London, G. Allen & Unwin, 1955.

293

suspicions regarding India’s independence.718 The second phase, from 1951 to 1954, witnessed the beginning of mutual visits that were primarily political and cultural in nature and often included state and non-state participants in equal measure—whether as high diplomacy, as people to people contacts, or as part of cultural delegations. Starting in 1955, and carrying on through early 1959, these exchanges expanded to also include high powered state delegations on fact-finding missions and study tours. 1956, in particular, appears to have been the a high-tide of Sino-Indian exchanges.

The first important milestone in the Sino-Indian exchanges was an unofficial

Indian goodwill mission to China in September 1951. The Indians were invited by five

Chinese organizations to celebrate the second anniversary of the founding of the PRC.719

In return a Chinese cultural delegation consisting of fifteen scholars, artists, and scientists and led by Ding Silin (丁西林 Ding Xilin), Vice-Minister for Cultural Affairs visited

India later in the year. According to Herbert Passin, “this was the first Chinese delegation to visit a non-Communist country.”720 The following year a cultural delegation, led by

Vijaylakshmi Pandit, was sent by India to participate in China’s May-day celebrations.721

India-China/China-India Friendship Associations (ICFA;中印友好协会 zhong-yin

718 India’s independence was considered incomplete since it had merely replaced a foreign bourgeoisie ruling elite with a local one. In the Chinese world-view, there were communist states and those states that would surely, eventually become communist. And these states, more specifically the people of these states, needed all the help the Chinese could offer. Or so, at least, was how several Indian and Western commentators perceived this. See Passin 1961: 87-88, for more along these lines.

719 Bondurant and Fisher 1956 cover this visit, and the various reports and memoirs published by the delegation, in some detail in their review essay. The five “All China” organizations were: The Peace Council, The Federation of Labor, the Democratic Women's Federation, The Democratic Federation of Youth, and the Federation of Literature and Arts Circles.

720 Passin 1961: 85.

721 Bondurant and Fisher 1956; also see: NAI 10(8)-XP, 1952.

294

youhao xiehui) had been set up in each country by 1952, and these quickly became one of the principal organizers of mutual visits.722 In June 1954, Zhou Enlai visited India and in

October of the same year Nehru visited China. These two visits not only re-affirmed the principles of Panchsheela, the Buddhist inspired principles of mutual respect and friendship, but also resulted in the signing of agreements on trade and student exchanges.723

The Afro-Asian Conference at Bandung in the spring of 1955 further served to enhance cultural exchanges.724 In September, thirty-two Indian teachers and students representing ten Universities went on a four-week trip to China. They were led by C.P.

Ramaswamy Aiyer, the Vice-Chancellor of Benares Hindu University. That same month

Chinese Muslims en route to Mecca were greeted by Indian Muslims and Hindus in

Delhi.725 September also witnessed the visit to China of India’s Health Minister, Srimati

Rajkumari Amrit Kaur.726 In October, an Indian film festival was held in China, which reportedly had three million viewers spread across twenty cities. During the festival, an eleven-member Indian film delegation visited China and was even granted an audience

722 See, for instance: WJB 105-00101-02; 105-00101-03; 105-00101-04; and 105-00884-01. Also see NAI 1(41)-FEA, 1955, where K.M. Panikkar requests a grant of INR 5,000 to regularize the activities of the Indian ICFA, which he claims had descended into a moribund state by 1955, lacking both a staff and a proper office.

723 Passin 1961: 95. In that year, Passin notes, two Chinese were sent to India to study Hindi and one Indian was sent to China to study Chinese.

724 On Bandung and its significance, see: Christopher J. Lee, ed. Making a World after Empire: The Bandung Moment and Its Political Afterlives. Athens: Ohio University Press, 2010.

725 Passin 1961.

726 See: NAI 10(23)-FEA, 1955.

295

with the Chairman.727 Other exchanges included a trip by Dr. Raghu Vira through China and Inner Asia to collect material for the Indian Academy of Asian Culture, the visit of

10 Chinese students, mostly to study Hindi, and a Chinese exhibit at the India Industries

Fair in New Delhi.728

The post-Bandung bonhomie and euphoria, already evident during the second half of 1955, blossomed in 1956. In that year “alone, 26 Indian delegations to China were counted and 33 Chinese delegations to India.”729 But 1956 also marked a qualitative shift in the exchanges. As the Indian leader of an ICFA delegation to China noted,

“We are proud to record here another important fact that premier Chou-enlai was pleased to grant us an interview which lasted for over an hour. We discussed very frankly and unreservedly several problems common both to India and China and the general feeling on both sides was that it would be in the interests of both the countries if an exchange on a more lasting basis between scholars and Scientists or technicians of both countries is made possible. This of course, it was recognized, requires [sic] to be done at Government level.”730

This appears not to have been just lip service. As many as three Indian Planning

Commission delegations visited China in the summer of that year with the explicit aim of studying Chinese planning techniques and experiences. The first was actually to have been led by Mahalanobis himself, but owing to his unavailability on account of ill health, was led by his protégé Pitambar Pant, who served as the secretary to the Planning

Commission. The goal of this delegation, which arrived in China in July, was to study the

727 Passin 1961; the delegation was led by led by the thespian Prithviraj Kapoor, the father of the famous Indian actor and director Raj Kapoor, and by B.N. Sircar.

728 Passin 1961.

729 Passin 1961: 91. Passin himself cites another source: Kirkpatrick, op. cit., p. 359. This appears to be a phantom source, since Passin doesn’t mention Kirkpatrick before this.

730 NAI EA FEA_4(4)-FEA/56: “Report of the Leader, Cultural Delegation sent by the India-China Friendship Association to China in April-May 1956 under the leadership of Prof. P.V. Bapat,” p. 41.

296

Chinese economy and planning methods.731 That same month, two other delegations arrived in Beijing. The first was a seven member team that included experts on cooperative organizations and had come to study agrarian cooperatives. They visited nineteen producers’ cooperatives in eight provinces, spending altogether about two months in China.732 The second was a six-member agricultural team under Deputy

Minister M.V. Krishnappa whose goal was to study Chinese Agricultural Planning and techniques.733 The Chinese, in turn, sent a three member team to India in August to study

“various technique of using Hankoo [sic] in House Construction.”734 1956 was also the year when real substantive statistical exchanges were inaugurated, first by Zhou’s visit to the ISI on 9 December, which was then followed by the visit of the delegation led by

Wang Sihua that stayed at the ISI for about a month. Other exchanges included a Chinese

Women's Delegation, also in December; ten journalists from India and as well as various student groups.735 Zhou Enlai summed up this new, more substantive enthusiasm for engagement in a letter he sent Jawaharlal Nehru on the same day that he visited the ISI:

“Through the present visit, we have been able to know more [of] India and have realized that in many respects the Chinese people can learn from your people. The Chinese

731 See NAI EA_FEA_4(18)-FEA/56; more on this delegation and its possible importance below.

732 Government of India, Planning Commission. Report of the Indian Delegation to China on Agrarian Cooperatives. New Delhi: Government of India: May 1957.

733 Government of India, Ministry of Food and Agriculture. Report of the Indian Delegation to China on Agricultural Planning and Techniques. New Delhi: Government of India, 1956.

734 See: NAI 7(10)-FEA, 1955 (Secret).

735 NAI 4(48)-FEA, 1956; NAI EA_FEA_4(18)-FEA/56; and Passin 1961.

297

Government will consult the Indian Government through diplomatic channels about sending people over to study.”736

By the end of 1956, domestic concerns with an inefficient and ineffective statistical apparatus, combined with a greater engagement with India across all spheres, had made engaged statistical exchanges a possibility again.

Re-engagement

The foundation for the statistical exchanges that followed was laid during the visit of the first Planning Commission Delegation. By the time it arrived in Beijing on 12 July 1956, the Chinese may already have had some sense of the advanced statistical methods being employed in India. During his 1955 China tour, Nehru had informed Zhou Enlai of the work that Mahalanobis had been doing in the field of statistics and planning.737 Yet, it is unclear whether Zhou or the statisticians at the SSB had much detailed information.

When Bahadur Singh, Counselor in the Indian embassy in Beijing, approached the

Chinese in early April to inform them that Mahalanobis was interested in leading an

Indian Planning Commission delegation to China, the initial response from Xue Muqiao, the director of the SSB and a Vice-Chairman of China’s Planning Commission, was lukewarm. He scribbled the less than enthusiastic note, “seemingly will not be good to refuse,” and forwarded the file to his boss, Li Fuchun for his approval. Li, in turn, agreed

736 WJB 203-00084-01: 129.

737 A People’s Daily article published on January 26 1956 (India’s Republic Day) had praised India’s achievements in economic construction and mentioned Mahalanobis’ role in drawing-up second five-year plan. See: Feng Zhidan 冯之丹. “Yindu renmin de jingji jianshe chengjiu 印度人民的经济建设成就 (The Achievements in Economic Construction of the Indian People).” RMRB January 26 1956, 4.

298

with Xue and forwarded it to Zhou Enlai for his final approval. This Zhou gave on April

19th, instructing the relevant bureaus and ministries to make necessary preparations.738

A reception team to organize the delegation’s activities was set up, with Xue

Muqiao as its leader.739 Among the first things they did was to request more information about the delegation members. The Chinese embassy responded to their request by the third week of May. In addition to informing the Foreign Ministry of Mahalanobis’ credentials, they also pointed out that he was pretty open-minded 开明—having visited and drawn from the influences of both the socialist and capitalist worlds—and favorably disposed toward socialism.740 The other members—the statisticians M. Mukherjee and R.

Natarajan, and Pitambar Pant—were similarly evaluated. It was acknowledged that the delegation was comprised of highly qualified specialists. The Chinese should adopt a warm and friendly attitude toward them and introduce not just the People’s Republic’s recent achievements but also her short-comings and failings.741 It was also felt that, given

Mahalanobis’ fame and achievements, he ought to be invited to deliver a few lectures on important statistical questions.742 The delegation’s familiarity with statistics and planning approaches in the United States as well as the Soviet Union was especially attractive to the Chinese.743 Such expertise made the delegation ideal sources for learning about the

738 The phrase Xue used was: “似乎不好拒绝.” See WJB 105-00487-02: 5-8

739 Other members included: Zhang X, Yang Ying X, Wang Guangwei; see: WJB 105-00487-02: 5-8.

740 They said Mahalanobis was open to accepting socialist ideas.

741 The Indians had provided them with a list of 18 questions/subjects of interest, responses to which had begun to be drafted. This list was later supplemented by an additional 10 questions on July 27. The Indians were also interested in particular in ascertaining the nature and extent of Soviet aid provided to the People’s Republic. WJB 105-00487-0: 13-14, 21, and 32-34.

742 WJB 105-00487-02: 16.

743 WJB 105-00487-02: 10-12.

299

methods of capitalist countries. Accordingly, it was hoped that the delegation members would introduce not just statistics and statistical work in India, but also in the West.

Lectures at universities on these subjects were also envisaged. Finally, advice was sought on what important statistical materials (such as journals and books) from the United

Nations, South-east Asia, and the West were worthy of subscription.

As it transpired, Mahalanobis fell ill and had to withdraw his participation. The delegation was instead led by Pant and spent a total of forty days in the People’s

Republic.744 During that time they held over thirty-six special discussion meetings with officials, planners, and statisticians across several bureaus. They were based out of

Beijing, but they also traveled a fair bit across China, visiting the famous Anshan steel works and the cities of Tianjin, Changchun, Shenyang, Shanghai, and Guangzhou. Such travel “gave an opportunity to the team to see the work of planning and statistical organizations at the provincial and municipal level.”745 It is not clear to what extent the

Chinese demands to learn more about statistics outside the socialist bloc were met. The

ISI Annual Report from that year states that:

At the request of the State Planning Commission, the members of the team delivered a few lectures at the Planning Commission premises. Pitambar Pant spoke on Planning in India. R. Natarjan talked about Industries in India and M. Mukherjee spoke on the Statistical set up of India. M. Mukherjee gave another talk to a small group of technical workers on the method of estimation of national income in India and the logical basis of income estimation work in the West and in the East.746

744 They entered the PRC through Shenzhen on 8 July 1956, and departed on 17 August.

745 Indian Statistical Institute. Annual Report, April 1956 - March 1957. Kolkata, 1957, p. 84.

746 Indian Statistical Institute. Annual Report, April 1956 - March 1957. Kolkata, 1957, p. 84.

300

This is the only indication I have found that the Indians actually delivered talks to the

Chinese, though there must have been opportunities during the small discussion group meetings for informal exchanges on these subjects.

If they had not done so before, the benefits of mutual exchange and learning seem to have become evident by the end of the trip. Just before departing the People’s Republic, the delegation traveled up to the Communist Party summer resort town of Beidaihe 北戴

河 and spent two hours in conversation with Zhou Enlai. During the conversation Zhou stated that he was particularly impressed with Indian statistical methods in the field of agriculture. He lamented that the Chinese themselves had been using purposive sampling, which had led to non-scientific results. In this matter, India was more advanced 先进 than

China, and China ought to study India.747 He went on to add that the history of planning in both countries was very short. Their experience was extremely limited, and they had committed errors along the way. But he “hoped that henceforth [the two nations] could mutually learn form each other.”748 Pant’s suggestion that an Indian team visit the following year was warmly received. So was his suggestion that one to three Chinese planners be sent to the Indian Planning Commission for a period of six to twelve months, though the internal memos went on to note that Pant did not make clear what exactly he expected the Chinese planners to do.749 Finally, Pant also invited a one to three member team to Calcutta in December of that year to participate in the silver jubilee celebration of the ISI. Zhou not only confirmed Chinese participation but added that a study team would

747 WJB 105-00487-02: 5.

748 WJB 105-00487-02: 9.

749 WJB 105-00487-02: 30.

301

also be sent to the ISI. And then, for good measure, he added: “I myself hope that I [too] will be able to meet with Prof. Mahalanobis when I am in Calcutta in mid-December.”750

The stage had been set.

Silver Jubilee Assessments

As we have seen, Zhou Enlai did indeed personally visit the ISI and meet with

Mahalanobis on 9 December 1956.751 Two days after Zhou’s visit, a four member delegation arrived at the Indian Statistical Institute on 11 December 1956. The team was led by Wang Sihua, who was now one of the SSB’s deputy directors. Other members of the team were Hwang Chien To, the director of the SSB’s Agricultural Statistics

Department; Tai Shih Kuang (戴世光 Dai Shiguang), a professor at People’s University in Beijing; and Sun Shih Cheng, a research officer of the SSB.752 Their first task was to attend the ISI’s twenty-fifth anniversary celebrations. Wang later reported that the

People’s Republic and the Soviet Union were the only two countries to have sent official delegations. Representatives of twelve other countries—all “capitalist” ones—were present in a personal capacity, and included such notables as Ronald A. Fisher of the

United Kingdom and the Americans Jerzy Neyman and Morris H. Hansen.753 The

750 WJB 105-00487-02: 3.

751 Zhou was accompanied by His Excellency Mr. Ho Lung, Vice Premier of the State Council and Marshall of the People’s Republic of China, His Excellency Pan Tsu-Li, Ambassador For the People’s Republic of China in India, Mr. Chiso Kwan-hua, Assistant Minister of Foreign Affairs, the Chief Minister of West Bengal, His Excellency the Ambassador of India in China.

752 Samvadadhvam 1, no. 3 (March 1957): 20.

753 Wang Sihua. “Kaocha yindu tongji gongzuo de baogao 考察印度统计工作的报告 (Report on Indian Statistical Work).” Report delivered at the 6th Annual Statistical Work Conference in Beijing in September 1957, published in Tongji gongzuo zhongyao wenjian huibian di sanji 统计工作重要文件汇编第三辑 (Compilation of Important Documents Relating to Statistical Work—Volume III). Beijing: Statistical Press,

302

festivities included boat rides on the Hooghly River, paper presentations [check], and much feting of the Institute’s success. Wang “conveyed to Indian statisticians the fraternal greetings of the statisticians in his own country.” He lauded the achievements of the institute, both in theory and in its practical applications, reserving special praise for the advances in sample survey and agricultural statistics.

“In these fields, the Indian Statistical Institute has accumulated not only data for planning and policy making, but also a great amount of valuable experience and methods, which will not only be useful to India but also to other countries including China.”

Wang highlighted the similarities that existed between the two countries and said that

Chinese statisticians hoped to follow Chairman Mao’s instructions that one must learn from other countries by learning from Indian statisticians.754 With the niceties taken care off, the delegation then proceeded to spend over a month studying statistics and statistical methods.

Altogether, the delegation was in India from the 11 December 1956 to 16 January

1957. Within a month of their return to Beijing, Wang Sihua published a report in the principal statistical journal, Statistical Work 《统计工作》, summarizing the delegations findings, and putting forward its recommendations.755 Wang informed his readers that

November 1959, 90-97 (quote from 90). The other countries included: The Netherlands, United States, Lebanon, Thailand, Pakistan, Japan, Switzerland, and the United Kingdom. For details, see: Samvadadhvam 1, no. 3 (March 1957).

754 Samvadadhvam 1, no. 3 (March 1957): 16.

755 Wang Sihua 王思华. “Jieshao yindu tongji gongzuo, jiji kaizhan quanguo chouyang diaocha 介绍印度统计工作，积极开展全国抽样调查 (Introducing India’s Statisitcal Work, Actively Carrying out a National Sample Survey).” Tongji gongzuo 1957.06: 8-10, 17. The article was reproduced in Wang Sihua 1986: 95-102. Dai Shiguang 戴世光, Wang’s fellow delegation member, also wrote an article on India’s sample survey system but it was entirely factual in nature and did not seek to make recommendations for the People’s Republic. See: Dai Shiguang 戴世光. “Yindu de quanguo chouyang diaocha—shehui jingji

303

from the very beginning Prime Minister Nehru strongly promoted the setting up of a statistical organization for post independence India. Accordingly, in 1949 a National

Income Committee was set up, which subsequently discovered that data for its work was lacking or fragmentary. In response, to collect data on income the National Sample

Survey was established in 1950. The delegation spent most of their time in West Bengal and as a result was able to witness the ISI’s close relationship with state statistical organs.

Wang concluded that even though it remained an autonomous body, the ISI had in effect become the Indian Government’s principal organ for statistical research and progress. Its principal activities included planning and design of country-wide sample surveys, the consolidation and research of statistical data, and the training of personnel. After 25 years of development, it had 1,400 or so employees, of whom 850 worked on sample surveys.756

Overall Wang offered several key observations about statistical work in India.

The first, and perhaps most striking to the delegation, was the fact that statistical work was not centrally organized or carried out. In line with the federal structure of the country, each state’s statistical work was different. Eleven states carried out their work independently, while the work of three states was jointly coordinated. The center only collected, coordinated and then summarized material it received from the various state statistical offices.757 Another worrying feature was the inefficiency of the system—the

diaocha 印度的全国抽样调查—社会经济调查 (India’s National Sample Survey—A Socio-economic Survey).” Tongji gongzuo, 1957.09: 2-6.

756 Wang did not fail to mention that on account of ISI’s varied activities, it had become a de facto organ of the Indian Government’s international diplomatic activities.

757 According to Wang’s reckoning, this Central Statistical Office (CSO), also set up in 1950, had 250 workers, divided into six divisions: publications, training, coordination, standardization, planning, and national income.

304

rate of not filing reports was as high as ten to fifteen percent. Furthermore, the incompleteness of the data remained a problem. The national income estimate had a huge error that ranged from ten to thirty percent for different sectors of the economy. As a result, the principal method of collecting data remained sample surveys. Besides a decennial census, five-yearly farm and livestock census, annual census of twenty-nine heavy industries, all other kinds of data, relating to income, production, economic conditions in rural and urban areas, urban employment, etc., were collected using sample surveys. The delegation was impressed with the numbers involved in statistical work and the relative abundance of published material on sampling work. Statistical training too seemed to be more widespread than in China. Fifteen of the thirty-three national universities possessed departments of statistics or offered training programs and altogether 2,000 students were engaged in the study of statistics. The ISI itself trained advanced statisticians, focusing in equal measure on theory and practical application.

Error checking was also done meticulously and the use of machines was central to the process. The ISI already have 250 machines and a dedicated staff of 400 to man them.

Sample surveys received the delegation’s particular attention. In this area,

“especially as far as the technology [of sampling] was concerned, [the Indians] had carried out detailed research; and at the same time, with regard to checking the quality of data, [they] had conducted various kinds of studies and [gained much] experience.”758

Furthermore, most of India’s statistical work relied on random sampling 随机抽样调查, whose theoretical basis was mathematical statistics 数理统计, i.e. probability theory 几

率论. To explain the salience of sampling in particular, Wang described India as an

758 “印度在抽样调查方面，特别是关于技术方面，研究得很详细，并在资料质量的检查上，进行各种研究和经验.” Wang Sihua1986: 97.

305

agricultural nation with over 50 million households, drawing strong links to the reality as it obtained in the People’s Republic as well. To carry out complete enumeration would not only require a huge number of investigators 调查员 but would also be uneconomical

这是很不经济的. And the entire process from designing a complete enumeration 普查 to collection, analysis, and publication would take anywhere from one to two years.

Compared to this, sample surveys were much quicker and only need a small number of investigators who would be sent to a small number of sample households 调查户.759

Wang told his readers that the Indians believed sample surveys were cheaper, quicker, and more accurate 较经济，较快，又较正确 and that random sampling was the best way of avoiding bias 偏见. “They believe that if the scope of the survey is very large, the survey area very extensive, and if one wants to obtain data within a short period of time, then the only practical and economical means of achieving this is through sampling.”760

Wang identified three major characteristics of sample surveys in India. The first was that they were in widespread use, and often national in scope. This was significant because India stood far ahead of other capitalist countries in its employment of sampling technology 远非其他资本主义国家所可比拟的. National sample surveys had been carried out in: industry, agriculture, animal husbandry, household industry, transportation, commerce, services, as well as on urban workers’ livelihood and rural cost of living and income. A second characteristic was that sample surveys were continuously carried out,

759 Wang also spoke of errors of two types: i) 抽样误差 (particular to sample surveys) and ii) 非抽样误差 (an error that is common to all kinds of surveys, including complete enumeration). In India, the latter kind are relatively more, since personnel to do complete enumeration are relatively less-able, less well-trained. See: Wang 1986: 99.

760 “他们认为抽样调查是在调查范围很大，调查地区很广，而又要在短时期内取得资料的唯一经济可行的办法.” Wang Sihua1986: 99-100.

306

using the method of interpenetrating 交叉 samples. Accordingly, it was possible to check and cross-check results and easily discover errors. In addition, all methods of examination and verification, including purposive ones 典型研究, were used to raise the quality of the data. Finally, sampling work was closely tied to policy making in India, and so it received the attention and support of the government. To illustrate his point, he explained that when Bengal was struck by a famine in 1943, the ISI was able to investigate the effects of the famine in a detailed way. Wang cited other examples of the close links between policy and sampling as well. In all cases, he asserted, useful data and results were delivered in a timely 及时 fashion.

Having worked hard to establish the usefulness of sampling, Wang turned his attention to recommending it for his own country:

On the basis of all that we have gained from an inspection of statistical work in India, having consulted and followed the Soviet experience of combining sample surveys with the periodic reporting of statistical tables, and on the basis of the recent specific circumstances of statistical work in our country, we offer this suggestion: vigorously/energetically launch national scale sample surveys in our country, and extensively collect [data] on every aspect of social and economic activity.761 [verify translation of last phrase]

Wang bemoaned the recent status of statistical work in the PRC. On the one hand, statistical forms were too many to the point of overflowing 太多太滥, while on the other, the statistical materials so produced were unable to completely satisfy the needs of the party leadership. According to him, such an irrational situation needed to be swiftly remedied. And the only way to stop or limit the excess issuing of unscientific 不科学 or fiscally wasteful reports by various departments was to speedily adopt the use of

761 Wang Sihua 1986: 100-101.

307

scientific sample surveys. But China’s planning apparatus was still rather rudimentary, and its society complex. Wang conceded that the complete enumeration periodical report system remained useful for tracking production and circulation 流通. But when it came to economic activities on a national scale, especially those dealing with the complex issues of allocation and expenditures, pursuit of complete enumeration at the ground level only resulted in all the data and indicators being flooded in needless trivial details 十分繁琐.

The resultant workload was too onerous and was inconsistent with the principle of spending less material and human resources for greater results, i.e. it was inefficient. In conclusion, Wang declared that only by organizing specialized scientific sample surveys could the Chinese get twice the result with half the effort 事半功倍.762

“And next year in Peking”763

At about the time that that statisticians, planners, and leaders were reading Wang Sihua’s exhortation to adopt sample surveys, deliberations were initiated by the SSB to invite

Mahalanobis to the People’s Republic for a lecture and discussion tour. The following month, the Foreign Ministry informed its embassy in Delhi that the State Statistics

Bureau, Peoples University, and the Economic Research Institute of the Chinese

Academy of Sciences would jointly invite Mahalanobis to the People’s Republic.764

Several months of negotiations over dates and other details followed. Sometime towards

762 Wang Sihua 1986: 101-103.

763 Unless otherwise specified, the primary source material for this section is WJB 105-00530-06. This archival file consists of a series of 14 bulletins issued by the office in charge of handling Mahalanobis’ affairs (接待办公室), and covers his activities over the entire duration of his stay in the People’s Republic.

764 WJB 105-00530-05: 3.

308

the end of May (the letter was received 31 May 1957) the three designated host institutions sent a letter to the State Council reporting on plans for the impending visit.765

They stated that his wife and one other statistical expert would accompany Mahalanobis and identified two principal areas for the visiting experts to lecture on: problems relating to statistical sample surveys 统计抽样调查问题 and the achievements of mathematical statistics and its application in practical work 数理统计学的成就及在实际工作中的应

用问题. A memorandum from earlier in May, meant to be forwarded to Mahalanobis, provides further details into the kinds of questions the SSB wanted to raise:

1. The concrete procedures of the sample survey concerning crop yield, such as the criteria of selection of survey units and the adequate number of such units. 2. How can one guarantee that the farm budget statistics as a result of such sampling method will, on the one hand, represent the [sic] national or provincial characteristics, and on the other hand, satisfy the needs of the administrations on [sic] the county level? What are the ways and means of controlling and investigating the representative character of the sampling of such a coordinated and complex nature? 3. In order to study the economic activities such as the distribution of income of the agricultural producers’ cooperatives, sample survey is necessary. But how are the proper measures of sampling and the adequate number of sample units determined? How can such sampling be representative of the whole country and the various provinces and at the same time satisfy the requirements of the counties?766 (Original in English)

The memorandum further observed that sampling bodies of 16,000 farm households and

6,000 industrial workers families had been selected, but wanted to know “to what other fields should we extend such sample surveys in our country?” Another letter, dispatched on 28 May 1957 to the Foreign Ministry by Wang Sihua himself, articulated additional topics for forwarding to Mahalanobis: Mathematical Statistics, its concrete uses and its

765 WJB 105-00530-05: 13-18.

766 WJB 105-00530-06: 11.

309

development trends, India’s experience with planning nation-wide sample surveys, and

[Statistical] Quality control of industrial products in India.

In return, the SSB acknowledged that Mahalanobis hoped to resolve three requests of his own while in the People’s Republic. The first was to organize visiting students who would travel to India for study; the second was to have specialists sent to the Indian Statistical Institute to help with translating materials from the Chinese; and the third was that he was interested in understanding the training of health services personnel in China. The letter went on to acknowledge that Mahalanobis was not just India’s premier statistician, and the moving intellectual force behind its second five-year plan, but also that he enjoyed much international fame 享有国际名誉. In receiving him an attitude of sincere and genuine friendship was to be adopted, thereby advancing Sino-

Indian friendship generally, and closer ties in statistical science in particular. Ideological differences ought to be set aside. The guiding principle would be to “seek common ground while accepting differences” 求同存异原则. The Chinese would learn with an open mind 虚心 and both achievements and shortcomings were to be presented honestly.

A list of preparatory materials, including published statistics, was drawn up so that they could be presented to the visitors for context and background.767 Wang Sihua was designated the leader of the team responsible for the visitors’ reception.768

767 These included materials on National Income, statistical data on the achievements of socialist construction, Farm Family income, Nation-wide workers family income, and other materials that Mahalanobis may request, assuming they were deemed safe to share.

768 Other members included: Huang Jianta 黄剑拓 (Head of the Agriculture Statistics Department 农业司长), Di Chaobai 狄超白 (head of the Economics Research Institute of the Chinese Academy of Sciences), Qi Luming 祁鹿鸣 (Director of Renmin University’s Statistical Planning Department), Dai Shiguang 戴世光 (Renmin University professor) and six others.

310

Mahalanobis entered the People’s Republic via Shenzhen on the morning of 19

June 1957. By three in the afternoon of the same day the party arrived in the southern metropolis of Guangzhou (formerly Canton).769 Accompanying him were his wife and the noted statistician and specialist in sampling methodology, D.B. Lahiri. The following morning they visited the Sanyuanli Agricultural Production Cooperative within the city and “saw the system of maintaining the village statistical records, and got acquainted with the method of “purposive sampling” (on the Soviet model) which was being extensively used for estimating agricultural production and for family budget enquiries.”770 Later that day Cheng Yingzhong 陈应中, the director of the Guangdong

Statistical Bureau, and Guo Ling 郭凌, the director of the Guangzhou Statistics Bureau, gave Mahalanobis and Lahiri a short presentation on the nature of sampling work done in the province and city. They laid out the methods used to estimate agricultural production and rural and urban incomes and livelihood measures. Mahalanobis was immediately suspicious when told that the error in the sample surveys was barely two percent. He remarked that for such [purposive] sample surveys the error ought to be much higher, adding that in the Soviet Union it was of the order of fifteen percent, in the United States twelve percent, and in countries such as Germany and Japan it was above ten percent.

The only way to carry out reliable surveys was to use random sampling 随机抽样, he exclaimed.

769 Unless otherwise specified, the primary source material for Mahalanobis’ activities is WJB 105-00530- 06. This archival file consists of a series of 14 bulletins issued by the office in charge of handling his affairs (接待办公室), and covers the entire duration of his stay in the People’s Republic.

770 P.C. Mahalanobis. Some Impressions of a Visit to China 19 June – 11 July 1957: 1. PCMMMA, Kolkata.

311

Over dinner later that evening Mahalanobis urged his hosts not to be too dogmatic in their approach to statistics. He told them that India had a few statistical methods which could greatly benefit a newly reborn socialist nation like China. Being newly reborn had significant advantages he explained: “even though the Soviet Union was the first nation to establish socialism, many of its statistical methods are too old and not easy to change anymore.” Indeed, this was exactly like England, whose “railways had the longest history in the world, and yet today [they] are the oldest and [therefore] rather difficult to alter.”771

Through his visits to the Soviet Union and his numerous interactions with its leading statisticians and economists Mahalanobis was well aware of the status of statistical research and work there. Three years earlier, he had written to Pitambar Pant that, “one orthodox viewpoint in USSR has been to exclude mathematics from economics; to insist on statistics to be almost exclusively the handmaiden of economics; and to separate mathematical statistics as a separate (and somewhat abstract) subject under probability and mathematics.”772 Since then he had had time to further understand developments in the Soviet Union. In his unpublished report of his China tour, he explained:

This dichotomy became at one stage particularly rigid in USSR partly due to the conservation [sic] of older statisticians (who were mostly economists lacking the knowledge of statistical mathematics) and partly to the antagonism of a certain school of thought which objected to the use of “abstract” mathematical methods in biology and agriculture. Soviet mathematicians (among whom Kolmogorov is recognized all over the world as the greatest expert in the theory of probability of the present age, who has supplied the rigorous mathematical foundations of modern statistical theory) and mathematical statisticians have, of course, always recognized the superiority of “random or probability sampling” and have always advocated its use in sample surveys. During the last 3 or 4 years

771 WJB 105-00530-06: 3.

772 Letter from PCM to Pitambar Pant dated July 3 1954. Papers of Pitambar Pant (190 LII). NMML, New Delhi, p. 49.

312

this question is being increasingly debated in USSR. From published reports of the proceedings of scientific and statistical conferences as well as from personal discussions with Soviet statisticians I have the impression that opinion is turning increasingly in favour of “random or probability sampling.” Statistical practice in USSR, however, still continues to be based on “purposive sampling.”773 [Emphasis added.]

Whether Chinese statisticians were aware of this ferment in Soviet statistical circles is a moot point. Much like in the Soviet Union, statistical practice in the People’s Republic had not moved towards any reconciliation with mathematical statistics. But Mahalanobis’ presence in the People’s Republic suggests that the Chinese were aware of the limitations imposed by their methodology. And discussions over the following three weeks would indeed indicate to him that the Chinese were “keeping an open mind” to the possibilities offered by mathematical statistics.774

Xue Muqiao’s Briefing

The following day, on 21 June, Mahalanobis and company flew to Beijing, arriving in the capital at three in the afternoon. The welcoming committee at the airport was a veritable who’s who of statistical luminaries of the People’s Republic and included Xue Muqiao, the director of the SSB, as well as the familiar faces of Wang Sihua and Dai Shiguang.

Wang and Dai personally accompanied the visitor’s to the Beijing Hotel, where they were to be put up. Over the course of the next day Mahalanobis and Lahiri met with representatives of the three host institutions, to decide upon the goals and tasks for the coming weeks. In addition to offering his knowledge and experience, Mahalanobis was

773 P.C. Mahalanobis. Some Impressions of a Visit to China 19 June – 11 July 1957: 5. PCMMA, Kolkata. Later in his trip he would also inform his Chinese hosts that ever since 1947 the Soviet Union had not attended any sample surveys activities organized by the International Statistical Institute. They just didn’t attach importance to mathematical statistics. See: WJB 105-00530-06: 15.

774 P.C. Mahalanobis. Some Impressions of a Visit to China 19 June – 11 July 1957: 6. PCMMA, Kolkata.

313

keen to learn about all aspects of socialist construction in the People’s Republic. He naturally had an interest in statistics and the statistical system, but his long list of questions and topics touched upon all aspects of planning, training of personnel, and national income. He was also keen to get his hands on a copy of the Peoples Republic’s

‘Twelve Year Plan for Scientific Research,’ which had been announced the previous year.775 When it came to discussing his lectures, Mahalanobis wanted to know the general level of the audience, so that he could better calibrate the technical parts of his explanations. A formal welcoming party was thrown for the visitors on the evening of the

23rd with an even bigger and more impressive list of Chinese statistical luminaries in attendance. In his welcome speech at the party, Xue Muqiao stressed the advances India had made in developing and applying sample surveys. It was precisely this experience that the People’s Republic wanted to mine, and therefore, had invited Mahalanobis, or

Prof. Ma, as he came to be called, to visit. The usual pleasantries from both sides followed. At the end of the evening, D.B. Lahiri turned to his interpreter and expressed the hope that future dinners would be smaller in scale so that more meaningful conversations could be had. The strongly political flavor 政治味道 of such parties had made him uncomfortable.776

The delegation’s work began in earnest on 24th June with a briefing by Xue

Muqiao on statistical work in the People’s Republic.777 Xue explained that over the

775 This last request seems to have been denied. See: WJB 105-00530-06: 77. In general, a decision had been taken to freely share those materials and data that were already in the public domain. Requests for data that were not would be handled on a case by case basis. Statistical data was still considered secret in the People’s Republic, and as a result, many of Mahalanobis’ requests were eventually denied.

776 WJB 105-00530-06: 10.

777 Much of the day on Sunday, June 23 was lost as Mahalanobis came down with a stomach ailment. Lahiri took the opportunity to visit the Great Wall at Badaling and the Ming Tombs.

314

course of the preceding five years an integrated statistical system modeled on that of the

Soviet Union had been set up and reached down to the level of village cooperatives all over the People’s Republic. With a staff of about 650 personnel, the SSB headquarters served as the apex body of this system and was responsible for collection and auditing of all statistical data and its subsequent delivery, as required, to various government agencies. Below it, at the provincial level, were statistical offices that coordinated and guided work across the roughly 2,200 counties 县 in the country. Under the counties were the village 乡 administrations, which coordinated the statistical work of about 750,000 village cooperatives. At this level, much of the work was carried out by part-time workers.

In line with the Soviet model, the scope of statistics was “comprehensive” and covered population, agriculture, industry, transport, trade, consumption, health, education, finance, national income, employment, and so on. Unlike the Soviet Union, though, the provincial statistical bureaus were under the administrative control of the Provincial People’s

Councils and their expenses were paid out of provincial budgets. Xue explained that this made the Chinese system more administratively decentralized but he still felt that it was

“probably as integrated as the Soviet system.”778

Xue then proceeded to explain the methodology employed by the SSB. The preferred choice was complete enumeration. This worked rather well in several sectors.

In the industrial sector, 3,000 units sent monthly returns and another 60,000 units submitted annual report. In construction, about a 1000 of the most important projects reported monthly, with the rest doing so on an annual basis. Reports from the railways, shipping, and trucking were submitted monthly, as were government purchase and the

778 P.C. Mahalanobis. Some Impressions of a Visit to China 19 June – 11 July 1957: 1-3. PCMMMA, Kolkata.

315

consumption of important commodities. The agricultural sector, however, posed a much larger problem. Data of interest included area sown, yield per unit area, “outturn” of important agricultural crops, livestock, and so on. The provincial bureaus had their hands full trying to completely enumerate the 750,000 village cooperatives. Delays were the norm, and the accuracy of the results was not always high. As a result, in 1955, an initial attempt was made to calculate the yield of crops per unit area over a large part of the country using sample survey techniques. In addition, sample surveys had been carried out in order to better assess the standard of living of rural, working class, and urban households. For instance, a rural survey in 1955 had covered 15,000 farm families across twenty-three provinces and an urban survey in 1956 had included 6,000 families. Xue clarified that the technology used was purposive sampling, and concluded by observing that through increased contact with India, and in particular following the Wang Sihua delegation’s visit to India, the Chinese had become particularly interested in learning from India’s expertise and experience in carrying out sample surveys.779 At the end of the briefing, Lahiri , apparently so frustrated at the People’s Republic’s refusal hitherto to adopt more advanced sample survey techniques, pointedly told his interpreter that he couldn’t fathom why the Chinese lacked confidence 信心不足 in sample surveys. 780

Mahalanobis would later claim that after this “brilliant survey,” both he and Lahiri had a much clearer sense of why the Chinese “were so deeply interested in sample surveys and mathematical statistics.”781

779 Much of the description in the above two paragraphs is from: P.C. Mahalanobis. Some Impressions of a Visit to China 19 June – 11 July 1957: 3-4. PCMMA, Kolkata.

780 See: WJB 105-00530-06: 16.

781 P.C. Mahalanobis. Some Impressions of a Visit to China 19 June – 11 July 1957: 1. PCMMMA, Kolkata.

316

“Quite a heavy programme”

With their brief firmly established, Mahalanobis and Lahiri set about preparing their lectures. Over the course of the following two weeks Lahiri and Mahalanobis led eight group discussions on sample surveys and delivered a total of nine lectures. The group discussions usually lasted from three to four hours and were frequently quite technical in nature. Of the nine lectures, Mahalanobis delivered five and Lahiri four. In the lead-up to each lecture, both of them had to prepare a synopsis, which was translated and pre- circulated, leading Mahalanobis to remark: “this was quite a heavy programme.” The lectures were usually day long affairs, divided into two three-and-a-half hour sections.

Mahalanobis delivered two lectures on “Sample Surveys” at the SSB, one on “Planning in India” at the Planning Commission, and two on “Recent Developments in the Theory and Applications of Mathematical Statistics” at Beijing and Peoples Universities. For most of these lectures the audience, comprising as many as 300 people, included statisticians, economists, senior mathematical students, and members of the mathematical staff of the two universities.782 Lahiri’s four lectures were similar in format, though he spoke exclusively on “Applications of the Theory of Mathematical Probability in Sample

Surveys.” Two of these were delivered at the SSB, one at Peoples University, and one at the Chinese Academy of Sciences. Lahiri’s lectures, in particular, were especially technical and the audience was restricted to thirty to thirty-five persons familiar with statistical work and knowledgeable in mathematics.783

782 Ni Jiaxun, then an undergraduate at Peoples University recalled attending one of these lectures. According to him, most in the audience did not follow much of what was said. He remembered it as an important event at the University, however.

783 P.C. Mahalanobis. Some Impressions of a Visit to China 19 June – 11 July 1957: 6-7. PCMMMA, Kolkata. On Lahiri’s lectures, the SSB Bulletin from July 5 stated that there were over twenty people

317

On the 24 June, after Xue finished his briefing, Mahalanobis had offered a three stage strategy for the discussions. Each successive stage would only be entered once he had received the approval of his audience. If it became necessary, he even offered to extend his stay in Beijing and cancel sightseeing tours to create additional time for discussions and lectures. In the first stage he wanted to introduce the general principles of sampling 一般原则的介绍. If everyone was satisfied with this discussion, he would proceed, in the next stage, to discuss these principles with much greater thoroughness 详

尽的讨论这些原理. The final stage would be more practically oriented: starting with a particular province they would try and draft a detailed and concrete sampling program 从

一省开始试拟详细具体的抽样调查方案.784 Having clarified this plan, Mahalanobis began stage one. He spoke about what a sample 样本 was, about errors 误差, the survey object 调查对象, the differences between random sampling 随机调查 and purposive sampling 典型调查, outlined their histories, and so on. In discussing the overall structure of statistical work, he stressed the importance of scientific and intellectual freedom and the need to engage with other disciplines since statistics was essentially an applied science. He also spoke of the importance of training. By this he meant training at all levels, and not just middle school students who could become statistical workers. Just as important was the training of statistical researchers and instructors. To make his point clearer he drew comparisons with the operation of a hospital. “The medical school must maintain an intimate relationship with the hospital. Similarly, statistical research organs

present and that the “discussion was a little too specialized, a little too technical in nature.” (谈得比较专门一些，技术性大一些). See: WJB 105-00530-06: 62-63.

784 WJB 105-00530-06: 18.

318

must remain in touch with the organs carrying out actual statistical work.”785 Based on the experience gained in India, he estimated that if the Chinese were to conduct a national sample survey with the sole purpose of satisfying the center’s requirements, then all they needed were 1,200 surveyors. If, at the same time, they also wanted to satisfy the needs of the various provinces, then 2,500 would be required.786

On the morning of 27 June, Mahalanobis delivered his first lecture on Sample

Surveys at the SSB and followed that up with a second one the next afternoon. On both days the audience numbered around 300. At the end of the second day Xue reported that even though the lecture was at times too specialized, it was still very helpful for the entire audience.787 The next day, which happened to be Mahalanobis’ birthday, both he and

Lahiri continued group discussion on sample surveys at the SSB.788 The previous day

Mahalanobis had also handed over a much more detailed list of questions regarding planning and statistical work in the People’s Republic. This list contained a total of seventy-three items and covered topics such as the organization of long term planning, the structure of planning, the technology used and its improvement, population and labor, gross domestic product and national income, production of goods, natural resources, trade, capital goods industries, the training of scientific personnel, and so on. The final item was potential Sino-Indian cooperation on planning and statistical matters.789 After a day of rest, preparation, and sightseeing, the discussions continued on the afternoon of the 1st of

785 WJB 105-00530-06: 22.

786 WJB 105-00530-06: 20.

787 WJB 105-00530-06: 31-35.

788 WJB 105-00530-06: 36-37.

789 WJB 105-00530-06: 38-48.

319

July at the State Planning Commission. On the 2nd of July Mahalanobis delivered a lecture on the Indian planning experience. This lecture was attended by over 250 people from the State Planning Commission, Economic Commission, Peoples University,

Chinese Academy of Sciences, and the SSB. Mahalanobis also delivered lectures to similarly large audiences at Beijing and People’s Universities on 5 and 6 July on the subject of “The development of mathematical statistics and its applications in the social sciences.”790 Mahalanobis and Lahiri then held independent group discussions over the next few days, visiting the three host institutions as well as the Bureau of Public Health.

The principal content of all the lectures and group discussions—the lecture synopses, lecture transcripts, subsequent questions and answers, as well as the minutes from the group discussion—was compiled, translated, edited, and published in a year later in May 1958 by the Statistical Press. 791 The foreword of this publication granted that there were specific points on which the Indians and the Chinese did not see eye to eye, but advocated that in spite of those differences, the Chinese needed to pay greater attention to the science and technique of sample surveys. The book was produced, therefore, as a reference for future research and discussions on mathematical statistics and sampling. Its distribution, however, was most likely limited to statistical bureaus and offices across the country. Whether it entered curricula at teaching institutions or was used in any other specific way remains unclear.792

790 WJB 105-00530-06: 64-67.

791 State Statistics Bureau. Yindu mahalanuobisi jiaoshou fanghua tongji baogaoji 印度马哈拉诺比斯教授访华统计报告集 (Report on the visit to China of the Indian Statistician, Prof. Mahalanobis). Beijing: China Statistical Press, 1958.

792 Underneath the title were the characters neibu faxing (restricted publication).

320

The discussions as well as the visits to statistical organs and various bureaus left a rather favorable impression on Mahalanobis.793 K. Natwar Singh, then a young attaché in the Indian Embassy, met Mahalanobis on July 5th and found him “a forthright and interesting man,” who was “pessimistic about the implementation of our Five Year Plan.”

Mahalanobis told him “that the Chinese were ahead of us [Indians]. Their implementation was first rate.”794 Indeed Mahalanobis found much to admire in what he saw. He wrote later that he felt their lectures and discussions,

Were really appreciated and might lead to some fruitful results. We were deeply impressed by the seriousness of the purpose with which the Chinese discussed these questions and by the amount of details into which the entered to clear up obscure points and to formulate a concrete program of work.”795

The Chinese had repeatedly told him that China had a “clean sheet of paper before her in the field of statistics, and was therefore in the fortunate position of adopting the best methods from all over the world.”796 This may have been particularly appealing to his catholic mind. Mahalanobis was also generally trusting of the quality and accuracy of the published and unpublished statistical materials he had been shown, noting:

“We were favorably impressed not only by the size of the statistical organization but also by the wide range and internal consistency (which we had occasion to test in some cases) of Chinese statistics. Changes in production, income, employment, etc., from year to year (1952-56) should

793 Mahalanobis was impressed with the People’s Republic in general. See: P.C. Mahalanobis. Some Impressions of a Visit to China 19 June – 11 July 1957: 13. PCMMMA, Kolkata.

794 K. Natwar Singh is an Indian politician and a former Cabinet Minister. Before joining politics, he was a member of the Indian Foreign Service for 31 years. Among his earliest assignments was a three year stint in Beijing, from 1956 to 1958. For more, see: K. Natwar Singh. My China Diary 1956-88. New Delhi: Rupa & Co., 2009, 65-66.

795 P.C. Mahalanobis. Some Impressions of a Visit to China 19 June – 11 July 1957: 11. PCMMMA, Kolkata.

796 P.C. Mahalanobis. Some Impressions of a Visit to China 19 June – 11 July 1957: 7. PCMMMA, Kolkata.

321

be quite reliable and comparable. The statistical position (with the exception of sample surveys and mathematical statistics) seems to be appreciably better than that in India in respect of coverage, availability, and accuracy of data required for purposes of planning and current policy decisions. It is, therefore, possible to use Chinese official statistics with confidence for a general appraisal of economic conditions in China.”797

Meeting Vice-Premiers and the Premier

Mahalonobis had the opportunity to express his admiration for the general progress the

People’s Republic appeared to be making to its leaders on two separate occasions. On

July 3rd Mahalanobis and Lahiri met with the Vice-Premiers Li Fuchun and Bo Yibo 薄

一波.798 Li and Bo were also chairmen of the Planning and Economic Commissions respectively and Li was also the Minister for Statistics. At the meeting, which began at

9:50 in the morning and lasted for approximately an hour and a half, Mahalanobis told the Vice-Premiers that on general statistical work he had little advice to offer. He regarded China as being far ahead of India when it came to building a socialist society, and statistical work as a concomitant part of that, was also more advanced than in India.

But Mahalanobis did want to specifically offer advice on linking planning and statistical work more closely, and promoted the widespread adoption of sample surveys through the setting up of a dedicated sampling bureau. He reiterated the advantages of sampling— that it was faster, more accurate, its error easier to ascertain, and cheaper. With a national scale sample survey in place, he claimed one would only need four to eight weeks to get a sense of economic development and socio-economic conditions across the country. On specific techniques, he recommended the use of interpenetrating samples 交叉抽样 and

797 P.C. Mahalanobis. Some Impressions of a Visit to China 19 June – 11 July 1957: 12-13. PCMMMA, Kolkata.

798 WJB 105-00530-06: 53-60.

322

probability sampling 机率抽样. Given China’s 2,200 counties, Mahalanobis estimated that the sampling bureau could begin with a staff of about 2,000. Work could initially begin in five or six provinces, and as it intensified each county would be assigned two surveyors on average, resulting in total staff numbering between four and five thousand.

Transcribers, computers (humans), and statisticians would, of course, also be required.

Mahalanobis urged that a training institute along the lines of the ISI be set up, so that not just statistical workers, but also a highly specialized cadre of statisticians could be trained.

This institute would also foster closer ties with the Chinese Academy of Science, and more generally, between statistics and the disciplines of mathematics and economics—

“this was the way of the future.”

Li and Bo were very receptive to Mahalanobis’ suggestions. They requested that he bring them up during his conversation with the Premier later in the week as well. Both

Li and Bo were particularly keen to see sample survey work begin as soon as possible.

And yet, lingering doubts about sampling technology remained. Li specifically enquired whether “on the matter of the accuracy of sample surveys, Prof. Ma could once more carry out discussions with the SSB.”799 Bo wondered if India could send a specialist or two to help with a sample survey of cotton production in Hebei that he wanted to see carried out in September or October. Mahalanobis replied that September target date was too soon and wouldn’t allow enough time for planning the survey. Better preparation would be to carry out pilot surveys on the paddy crop. He also recommended that they immediately send one or two people to the ISI so that planning for such work could begin.

799 WJB 105-00530-06: 53-60, 57.

323

Mahalanobis also took the opportunity to reiterate his specific proposals for enhanced Sino-Indian cooperation in statistics and planning. He requested that the

Planning Commission and the SSB arrange for the regular forwarding of technical data to the Indians. He suggested that a small “China Unit” be set up at the ISI. This group would help translate into English important Chinese publications, articles and technical data, teach Chinese to a few statisticians and professors, and maintain direct links with

Beijing. To man this unit he requested one or two people be sent over, and assured that all their costs would be borne by the Indian side, and that they would be treated as guests.

In return, he offered to happily share materials for the teaching of statistics, and also impart training to those Chinese who visit the ISI. The hope was that three to four people would be sent to India for further study and that this would later become an annual practice. Indian specialists in statistics on sample surveys, mathematical statistics and other areas, could also be sent to help in further research, training or the drafting of technical schemes.800 Li and Bo agreed with this advice and indicated that the details could be worked out with Xue Muqiao and Wang Sihua, both of whom were also present at the meeting.

About a week later, on 9 July, Mahalanobis, his wife, and Lahiri were invited to dinner with Zhou Enlai.801 Xue Muqiao and Wang Sihua were also present, as were the

Indian ambassador, R.K. Nehru and Mrs. R.K. Nehru. Mahalanobis repeated much of what he had told Li and Bo. He was pleasantly surprised to discover that Zhou had been kept informed of his activities and had “studied some of the rather technical points in my

800 WJB 105-00530-06: 47-48.

801 WJB 105-00530-06: 68-71.

324

recommendations as he asked some searching and critical questions on the methods of assessing the accuracy of the results of sample surveys.”802 The Premier agreed in principle with most of Mahalanobis’ suggestions. The setting up of an institute might take some time, he observed, since the People’s Republic still lacked enough trained personnel.

But he welcomed Indian help in the organization of statistical work and research, and expressed interest in continuing the exchange of personnel, expertise, and experience. He also told Mahalanobis that while it may be difficult to send experts immediately to lecture on Chinese statistics and planning, the two SSB staffers that would be sent later in the year could help get the work of the Chinese Unit started. The dinner was a pleasant affair and was subsequently reported upon in the People’s Daily with an accompanying photograph of the group.803 As the trip wound down, both sides threw farewell parties, the Chinese one much grander than the smaller affair the Indians organized at the Beijing

Hotel. Continuing cooperation was the buzzword. On 12 July Mahalanobis and company left the People’s Republic the same way they had entered, via Shenzhen, assured that exchanges and future cooperation had been put on a solid footing.

Domestic Difficulties

A couple of months after Mahalanobis’ return to India, Wang Sihua addressed the

Conference on National Statistical Work. His speech, titled “Report of An Inspection of

Statistical Work in India,” started by referring his audience to his article from earlier in

802 P.C. Mahalanobis. Some Impressions of a Visit to China 19 June – 11 July 1957: 10. PCMMMA, Kolkata.

803 Renmin ribao 10 July 1957, p. 1 (“周总理接见印统计学家马哈拉诺比斯”).

325

the year.804 Wang stated that he had already covered the general outlines of statistical work in India in that piece and would not revisit trodden terrain. Instead, he wanted to focus exclusively on sample surveys. His report was less enthusiastic, more guarded, and more strategic than his February article.

Wang now took pains to stress that Mahalanobis was aware not only of Soviet statistics, but that he also knew that Chinese statistics were largely a product of learning from Soviet experience. Furthermore, Mahalanobis, he assured his listeners, actually thought very highly 评价还很高 of the Chinese periodical report system. It was only in those instances where original records 原始记录 were missing, or where the numbers were too large and too many, such as in matters relating to the people’s livelihood, that the use of sample surveys was recommended. As a socialist country, Wang acknowledged, the People’s Republic should of course use a system of complete enumeration as the main methodology, but it was also important to study other methods, something which had not been done sufficiently before. He then addressed the doubts, fears, and criticisms head on:

“There are those who are suspicious: Does our visit to India mean that we have changed our policy of learning from the Soviet Union? Such an understanding is in error. Our India visit and Prof. Ma’s [Mahalanobis’] coming to the People’s Republic absolutely do not mean that we have abandoned [a policy of] learning from the Soviet Union. In reality, sample surveys have already been in use in the Soviet Union from an early time. It can be confidently stated: in the past, today, and in the future, we want to learn from the Soviet Union. But, at the same time, we must also study those methods of capitalist countries that are relatively good, especially the methods of sample surveys and the problems relating to statistical technology [check translation of this]. As far as method is concerned, not just India, if other capitalist countries have even better methods, we can

804 Wang Sihua 1986: 90-97.

326

study those as well. Our study’s goal is to provide better service in the construction of a socialist society.805

Wang then went on to outline the nature of sampling work in the Soviet Union, providing evidence of its use for decades. He reminded everyone that Soviet specialists were the ones to design the rural livelihood survey 农家收支调查. It was only after this that he began a detailed discussion of India’s sampling work, concluding with a list of four advantages. The first of these was that the scale and nature of sample surveying in India were larger and more expansive than in England, the United States, and other capitalist countries. The second was that Indians had made new contributions to the technology of sample surveys, namely the methods of continuous sampling 连续抽样 and interpenetrating samples 交叉抽样. Thirdly, they had carried out detailed comparative research and experimentation on various survey methods. Finally, the contents of the surveys were tailored to meet practical requirements. [Check last bit for whether meaning has been translated properly].806

What was known to Wang and many of his comrades in Beijing had to be communicated more effectively, and subtly, to the various regional representatives at the conference, and through them to the vast numbers of statistical workers across the country. This was that sample surveys were a general category that included purposive sampling as well as random sampling. And what was new and particularly relevant to the

People’s Republic were the Indian random sampling methods, not the prevailing Soviet methods. Wang adopted a two-stage strategy to bring this point home, without actually

805 Wang Sihua 1986: 91.

806 Wang Sihua 1986: 94.

327

stating it. He first differentiated sample survey from complete enumeration, highlighting advantages already familiar to us: sample surveys were cheaper, required fewer personnel, and that since their scale was smaller than complete enumeration, they could accordingly accommodate many more indices and much greater complexity.807 Having established the usefulness of sample surveys, Wang proceeded to differentiate what he called subjective

主观 sampling from random or stochastic 随机 sampling. For this purpose he again resorted to a laundry list of advantages: (1) if the survey design is good, it can completely reflect every characteristic of the totality, (2) random sampling avoids the subjective influence of the surveyor, and is therefore more accurate, (3) subjective sampling only allows for one or two indices 指标 unlike random sampling which allows all indices to be used [check this translation], (4) random sampling is more objective, (5) random sampling is a simple and easy method, and (6) random sampling can provide accurate estimates of sampling error, and allows for the estimation of the accuracy of the overall survey.808 But he did not advocate the abandonment of other forms of sampling. As long as a method was scientific it ought to be used. What then was the definition of scientific?

For Wang, “[a method] was scientific only if it could truly reflect the objective real situation.”809 Thus, subjective sampling, under the right conditions of the Party’s Mass

Work Line 群众工作路线, could be scientific too. He further comforted his audience by invoking the authority of Mahalanobis himself:

In the past some comrades believed that the difference between purposive surveys and sample surveys rested on whether it was random or not. But,

807 Wang Sihua 1968: 95.

808 Wang Sihua 1986: 95-96.

809 Wang Sihua 1986: 96.

328

on the basis of Prof. Ma’s introduction [to sampling], even within sampling there can exist subjectivity. Setting aside random sampling, to varying degrees both mechanical sampling and sampling by classification have subjective elements. Therefore, I believe the principle difference between sample surveys and purposive surveys lies in that sample surveys use a small number of objects [对象—find more accurate translation for this] to calculate the entire object. That is, the entire sample must represent the whole and the totality [全及总体]. [On the other hand,] the principal goal of purposive surveys is not to analogize for the totality.810

Wang pointed out that the method of purposive (or typical) surveys had not been effective in the past because the selection of typical cases 选典 had not reflected the objective reality. The task, as it lay before them, was to carry out small scale experiments to find out which method would be most suitable.

The transformation from February to September was stark. From the vigorous promotion of establishing nation-wide sample surveys, the mood had shifted to one of qualified, even muted, endorsement. After several years of denigrating ‘capitalist’ statistics, Wang had uncharacteristically written favorably about them in February. This speech, seven months later, suggests he did not strike enough of a conciliatory tone in the earlier article and, therefore, had to make amends. Indeed, Wang’s speech at the conference can be read as an exercise in persuasion meant to assuage reactionary cadre who feared not just ideological compromises, but perhaps even more importantly, being shifted out of their comfort zones and established statistical practices and habits. Larger domestic political currents also no doubt played their part. Starting with the Hundred

Flowers movement in 1956 and through May 1957 when free expression by intellectuals was encouraged, the political climate had been relatively conducive to the exploration of new possibilities. In June 1957 the Anti-Rightist campaign began, leading to the

810 Wang Sihua 1986: 96.

329

narrowing of those possibilities, and no doubt, to the much more cautious tone that we find Wang adopting in September. A couple of months later the xiafang [下放] movement began with cadres being sent to factories and villages to work with their hands.

The movement was later extended to intellectuals as well. Under the circumstances, advocating capitalist methods could no longer just be a matter of their scientific-ness.

The Study Tour of 1958811

It was in this somewhat confused milieu, where intellectuals and city-dwellers faced the prospect of manual labor in the countryside, that Wu Hui 吴辉 and Gong Jianyao 龚鉴尧 found themselves on a mission to India. Wu, born in 1926 [verify], and Gong, in 1927, had received degrees in statistics from Chongqing and Nanjing Universities respectively in 1950. They were part of a whole generation that had been introduced to and trained under an inclusive understanding of statistics, which drew from texts and materials from the United States, Britain, as well as the Soviet Union. After 1949, they had to recalibrate their training to serve the newer and narrower definition of statistics within the People’s

Republic. By the mid-1950s both were working at the SSB headquarters in Beijing. Gong was assigned to the agricultural statistics division and Wu to the industrial statistics one.

In 1957 they were among ten young cadres shortlisted for additional testing to determine who would be sent to India. Each shortlisted candidate was made to write an essay demonstrating their English language skills. In the end, Wu and Gong’s facility with the language coupled with their specializations in the key areas of focus—industrial and agricultural statistics—won out.

811 Unless otherwise noted, this section draws primarily on oral history interviews conducted with Gong Jianyao and Wu Hui. I met Wu in Beijing on 8 September 2011. I met Gong, who lives in Guangzhou, a week later on 14 September 2011.

330

Wu Hui and Gong Jianyao arrived in Calcutta on January 4 1958. They were immediately invited by Mahalanobis to travel to Madras (present day Chennai) to attend a conference on statistical work and planning [find details]. Upon their return to Calcutta, they began their studies at the Indian Statistical Institute. Their official status was that of

Visiting Scholars (访问学者 fangwen xuezhe). At that time, the ISI had several tens of students form various Asian countries who were receiving training at the International

Statistical Education Centre (ISEC). As visiting scholars Wu and Gong’s status was different from these students and they did not attend classes. Instead, they were given study spaces and permitted to conduct their own independent reading and research. The primary focus of their work was the study of Indian sampling techniques. This they began by reading all previous documents and publications relating to the ISI’s statistical sampling activities, which were held in the ISI library. From there, they moved to a close study of the theory and methods of sampling. In the process, they studied simple random sampling, stratified random sampling, cluster sampling, double sampling, multi-stage sampling, and other sampling methods. Periodically, they would meet with Mahalanobis,

Lahiri or C.R. Rao and discuss progress made and problems encountered.812 D.B. Lahiri, in particular, was their principal guide and interlocutor throughout their studies. Exposure to the latest Western texts on statistics and sampling rounded off their readings.

812 Both Wu and Gong recalled Mahalanobis fondly, observing that he left a rather deep impression on them. In a 2001 article published in the China Statistics, Gong went so far as to claim that two professor Mas had played a hugely inspirational and instrumental role in his life. The first, he told his readers, was Professor Ma Yinchu 马寅初 (1882-1982), the well-known economist and one time President of Beijing University, persecuted during the late 1950s for his championing of birth control. The second Prof. Ma was none other than Mahalanobis himself. See: Gong Jianyao 龚鉴尧. “Shidao qinghuai 师道情怀 (Sentiments from a Career in Teaching).” China Statistics 中国统计, (May 2001): 52-54.

331

While the bulk of their work was theoretical or centered on book-learning and based at the ISI, Wu and Gong also had the opportunity to gain first hand experience in the practice of sampling by participating in two sample surveys. The first of these was a land use sample surveys where they witnessed the Mahalanobis devised method of cutting circular crop samples in the fields. The second was a family budget survey carried out in Calcutta. Other activities included sitting in on meetings of the National Sample

Survey (NSS) where sampling teams would report their recent experiences, problems, and discuss solutions for the next round of the NSS. Both Wu and Gong also had the opportunity to interact with the various Indian and foreign statisticians present at the ISI during 1958. Both remembered, in particular, the visit of the Soviet statistician A. Ezhov, who had served as the first Soviet statistical expert to the PRC from 1950 to 1952.813

Gong also became close friends with C.R. Rao, who at that time ran the ISI’s training school [verify]. During the year, both of them visited Bombay, Pune and Delhi, as well as several paddy and jute sample survey sites across the country.814 While in Delhi, they also met with Pitambar Pant, who was still serving on the Indian Planning Commission.

During the exactly one year and five days that Gong and Wu spent in India, events within the People’s Republic had pushed statistics, as they had just about everything else, in new directions. Gong recalled reading about the Great Leap Forward in the papers, and receiving letters from friends and family describing the push towards communes. Wu also remembered discussing the agricultural growth results of the Great

813 For more on Ezhov, see Part I Chapter II.

814 Gong Jianyao 龚鉴尧. “Woguo chouyangfa yanjiu yu shijian de fazhan 我国抽样法研究与实践的发展 (Development of the Research and Practice of Sampling in our Country).” Xi’an tongji xueyuan xuebao 西安统计学院学报 9, no. 2 (1994): 7-14, 10.

332

Leap Forward, claiming that Gong was openly suspicious of them. Statistics too, they were informed in time, was changing. In August 1958 an “on-the-spot meeting” 现场会

议 was held in the town of Baoding, just south of Beijing.815 At this meeting, statistical work carried out by the SSB over the previous seven years was repudiated, and a call was made for a Great Leap Forward in statistics as well. The logic was straightforward:

[With] the collectivization of peasant production and consumption, all [statistical] data could basically be obtained [directly] from the communes. The peasant family revenue and expenditure survey, which had been set up on the basis of surveys of sample households, had fulfilled its historical obligation/role [历史使命].816

In one fell swoop, established statistical systems, including the purposive sampling already carried out, as well as future potential surveys based on random sampling, were all consigned to history. The Sino-Indian exchanges and much touted statistical cooperation too were to become a victim of this development. Towards the end of their

India stint, Wu and Gong had worked hard to draft a lengthy report of their activities in

India and their recommendations for future statistical work in China. Upon their return in early 1959, it “too had to be rewritten and was converted into a simple report and submitted to the SSB’s leadership to read and dismiss [参阅了事].”817

815 See for instance: “Zhaokai quanguo tongji baoding xianchang huiyi de baogao 召开全国统计工作保定现場會議的報告 (Report on Convening an on-the-spot National Statistical Work Meeting at Baoding)” in Vol. 3 of Tongji gongzuo zhongyao gongzuo huibian 统计工作重要文件汇编. Beijing: Statistical Press, 1959.

816 Gong Jianyao 1994: 10-11. During my interview with him, Gong recalled how bad an effect the Baoding meeting had on statistical work. More on the Baoding meeting in the Epilogue.

817 Gong Jianyao 1994: 11. Both Gong Jianyao and Wu Hui would play major roles in subsequent years in statistical work and were principal actors in the post 1978 reform of statistics as a discipline and profession.

333

Things Fall Apart

Just as the reengagement had been facilitated by domestic and international factors in

1956, in 1959 again, a combination of domestic and international factors drove the two nations and statistical exchange apart. The previous section has already alluded to the changes within China that had a deleterious impact on statistical work. On the international stage, events were to overtake the Sino-Indian statistical exchanges as well.

Disagreements over the Sino-Indian border, which had existed since the start of the decade began to dominate diplomatic exchanges as each side hardened its position. The ensuing frosty relations no doubt served to reduce some of the spirit of Bandung and endanger the Panchshila principles.

It is unclear to what extent Mahalanobis and the ISI had cottoned on to these differences or to the major upheavals within the People’s Republic during 1958. In

October of that year, he lobbied the Chinese to send a couple of planners with expertise in heavy industries to help with the drafting of India’s third Five Year Plan.818 He lamented that the half-hearted development of heavy industries in India during the years of the second Five Year Plan had been a big mistake and wanted to take advantage of

China’s direct experience 直接的经验 in this area. The Chinese Embassy in New Delhi duly forwarded Mahalanobis’ request to the Foreign Ministry in Beijing and sought further instructions. The Foreign Ministry, in turn, asked the Planning Commission for advice on the matter. In doing so, they also added that were Chinese experts to be sent, their primary goal would be to “expand our country’s influence.”819 After additional

818 WJB 105-00647-02: 1.

819 WJB 105-00647-02: 4.

334

deliberation involving the Foreign Ministry and the Planning Commission it was eventually agreed to send a three-man team. A telegram sent to the Chinese Embassy in

New Delhi on 26 February 1959 listed the members of this team: the Deputy Director of the Planning Commission’s Long term Comprehensive Planning Bureau, Chen Xian;

Deputy Director for the Comprehensive Department of the Capital Construction Planning

Bureau, Wang Yilin; and the recently returned Wu Hui, who would serve as the team’s interpreter. The telegram made clear that these experts would only serve as consultants who would introduce the People’s Republic’s recent general experience in heavy industries. Under no circumstances would they advise the Indians on their problems. The

Chinese Embassy in New Delhi wrote back a few days later on March 6, stating that

Mahalanobis welcomed the Chinese experts and hoped they could visit from April to

June. Pitambar Pant and M.M. Mukherji, both of whom had been part of the 1956

Planning Commission delegation, and were now involved in comprehensive planning, would receive them and arrange their activities.820

The archival trail turns cold after this. On 10 March an uprising occurred in Lhasa and by the end of the month the Dalai Lama and several tens of thousands of his followers had escaped to India. They were granted asylum and allowed to settle in the northern hill resort of Dharamshala. At a scientific conference held in Moscow in May, a little over a month after the Dalai Lama’s flight from Lhasa, Mahalanobis bumped into the Chinese physicist Zhou Peiyuan 周培源 and the two engaged in a long conversation.821 Zhou later filed a report of this meeting with the Foreign Ministry.822

820 WJB 105-00647-02: 5-9.

821 Zhou and Mahalanobis were part of a delegation of the Association of International Scientists 世界科学工作者协会代表团 that had written to Khruschev requesting a cessation of nuclear testing. See: “赫鲁晓

335

Eight of the report’s nine pages were devoted to a discussion of Tibet and geopolitics.

Only on the last page did Zhou dwell on a few scientific points that had also come up. As for the three-member Chinese team, it never visited India.

夫复函世界科学工作者协会代表团苏联愿立即签订停止核试验协定各国人民正殷切期望国际局势有所缓和美国一些人物却大叫大嚷要恢复核试验.” Renmin ribao 13 August 1959, p. 4.

822 WJB 105-00656-02.

336

CHAPTER IX

Epilogue: A ‘Great Leap’ in Statistics

337

1958 marked a watershed in the history of the early PRC. The execution of the first five- year plan had been a success, meeting and in many cases exceeding the targets that had been set. And yet, instead of a second five-year plan that built upon the first one, China was launched down the path of the Great Leap Forward (GLF) 大跃进. At its most basic level, the GLF was a radically different approach to economic and social organization. It retained plan-like characteristics of setting specific production and agricultural targets, but did so by implementing a far-reaching reorganization of labor and production technologies. In the agricultural sector, farming was collectivized and the people’s communes 人民公社 so created held several thousands of households that ate in collective kitchens and labored in nationalized fields and back-yard blast furnaces.

Similar labor-intensive practices were also introduced in the industrial sector as China sought to overtake England’s steel production within fifteen years.823

Much of the impetus behind the GLF came from Mao’s disavowal of the Soviet

Union and her ‘advanced experience.’ At the Chengdu Conference in March 1958, “Mao applied the label of ‘dogmatism’ to previous Chinese policies towards heavy industry, planning, banking, and statistics, explaining that because China lacked native experience and competence in these areas, she had rigidly copied Soviet practices regardless of their

823 There is a large literature on the Great Leap Forward. For succinct summaries, see: Riskin 1987, Meisner 1999. On the effect it had in the countryside, see: Kimberly Ens Manning and Felix Wemheur, eds. Eating Bitterness—New Perspectives on China’s Great Leap Forward and Famine. Vancouver: University of British Columbia Press, 2011. There is a large body of ever-expanding scholarship on the famine that ensued in the wake of the GLF. For a trenchant review of some of the more recent works, see: Anthony Garnaut 2013. “Hard Facts and Half-Truths: The New Archival History of China’s Great Famine. China Information 27, no. 2 (2013): 223-246.

338

appropriateness to Chinese conditions.”824 Like almost everything else, statistics too did not escape the consequences of a rejection of Soviet practices.

Between 26 June and 8 July the State Statistics Bureau convened a national ‘on- the-spot conference 现场会议 in the northern city of Baoding 保定 to promote its vision of a great leap in statistics. Consisting of regular meetings as well as extended site-visits, the Baoding conference had two explicit aims: one, to criticize statistical practice of the past few years, especially given the new dictum of ‘politics in command’ 政治挂帅; and two, to promote statistical work done in the province of Hebei 河北 as the exemplar of a new approach to statistics.825 After two days of meetings on 26 and 27 June, the delegates visited an exhibition of statistical work 统计工作展览, which had been jointly curated by the SSB and several provincial and municipal statistical bureaus, including Beijing,

824 Riskin 1987: 114. For more on Mao’s critique of the Soviet Union and Soviet experience, see: Mao Zedong 毛泽东. A Critique of Soviet Economics by Mao Tsetung (translated by Moss Roberts and annotated by Richard Levy, with an introduction by James Peck). London: Monthly Review Press, 1977. James Peck’s introduction, in particular, examines the evolution of Mao’s views, which eventually came to consider Soviet experience “utterly inadequate as a positive model for China’s drive to build socialism.” See Mao 1977: 24.

825 For summaries of the meeting, see: (1) State Statistics Bureau. “Ganqing, chongtian—quanguo shengshi tongji gongzuo baoding xianchang huiyi shengli bimu 干劲冲天、信心百倍—全国省市统计工作保定现场会议胜利闭幕 (Energetic and Brimming with Confidence—The National Provincial and City Level On-the- Spot Meeting in Baoding Comes to a Successful Conclusion).” Tongji gongzuo 统计工作 1958.14: 25; (2) “结合现实斗争搞统计—统计工作者决心依靠群众三个月改变工作面貌.” Renmin ribao 1 August 1958, p. 6.; and (3) DSJ: 59-60.

The DSJ notes that discussions on organizing a meeting in Baoding were initiated in the Spring of 1958, and on May 4 the State Planning Commission gave the SSB the go-ahead on holding the meeting. The Beijing statistics bureau had already sent delegations to study statistical work in Hubei in April of 1958. For more, see: BMA 133-001-00120.

339

Shanghai, Shanxi, and Sichuan.826 After that, approximately twenty delegations were organized and dispatched across Hebei to various factories, construction sites, schools, streets and districts, townships, communes, and [work] teams to carry out week-long site visits. In all, nearly one-hundred grass-roots statistical units were visited. The conference was reconvened in Baoding on 6 July and the following three days witnessed statements by twenty-four delegates, representing a range of provinces, districts 区, and municipalities.827 In their speeches, various delegates promised to not only catch-up with

Hebei’s reforms but to also completely reform statistical work within three months, and thereby be able to report such good news as a gift to the nation at the upcoming National

Day on 1 October.828 The delegates brought the meeting to a close on 8 July by sending a telegram to Mao Zedong in which they publicly guaranteed to meet that three month goal.829

The Baoding Conference was also followed by two additional conferences in

Beijing in 1958 and 1959. Restricted to provincial, municipal, and regional bureau chiefs,

826 See: “Yige fengfu duocai de ‘tongji gongzuo zhanlanguan 一个丰富多彩的 ‘统计工作展览’ (A Richly Colored Exhibition of Statistical Work).” Tongji gongzuo 统计工作 1958.14: 28-29. The exhibition contained approximately 1,000 items displayed in over ninety rooms spread across three floors.

827 The statements delivered by representatives from Shanxi, Henan, Shaanxi, Guizhou, and Heilongjiang were republished as individual articles in the Tongji gongzuo’s 14th issue in 1958. An additional twenty-one statements from representatives were summarized and reproduced in the order they were delivered in an article in the same issue: Hunan, Guangdong, Jiangsu, Anhui, Zhejiang, Beijing, Guangxi, Shanghai, Hubei, Liaoning, Jiangxi, Sichuan, Inner Mongolia, Gansu, Fujian, Xinjiang, Tianjin, Qinghai, Shangdong, Yunnan, and Hebei. See: “Ge sheng qu shi tongji gongzuo yuejin zhishi 各省区市统计工作跃进指示 (Instructions for a Great Leap Forward in Statistical Work at the Province, Regional, and Municipal Level).” Tongji gongzuo 统计工作 1958.14: 26-28.

828 “学河北, 赶河北, 国庆节报捷、献礼.” See: State Statistics Bureau (“Ganqing, chongtian—quanguo shengshi tongji gongzuo baoding xianchang huiyi shengli bimu”), 1958.14: 25.

829 “Xiang mao zhuxi baozheng—kuzhan sange yue, shixian quanguo tongji gongzuo de da gaige, da yeujin! 向毛主席保证—苦战三个月，实现全国统计工作的大改革，大跃进! (A Guarantee to Chairman Mao—Three Months of Hard Work to Achieve Reforms in National Statistical Work, [an d bring about] A Great Leap Forward!).” Tongji gongzuo 统计工作 1958.14: 1.

340

these conferences sought to discuss the progress made and problems encountered during the post-Baoding rectification 更正 of statistics. The first of these meetings took place four months after the Baoding meeting, and lasted from 25 October to 3 November.

Summary materials produced in its wake, and meant for internal circulation, reiterated that the Baoding Conference had created a new situation 新的局面 in statistical work.830

By the time of the next meeting, which took place from 11 to 21 April 1959, the SSB had a new director in Jia Qiyun 贾启允.831 The Deputy Director of the SPC, Jia Tuofu 贾拓

夫 also participated in the meeting.832

In his speech at the Baoding Conference, Xue Muqiao had rhetorically asked:

“What is a Great Leap Forward in Statistics?”833 For Xue, a great leap in statistics was marked by two key characteristics.834 The first, in line with Mao’s general critique, was based upon a rejection of dogmatism. Statistics was not an end unto itself but rather a tool

工具 in service to promoting political and productive struggles 为政治斗争、为生产斗

争服务. And so, in a dramatic turnaround, Xue rejected the primacy of the very principles that he and the SSB had been espousing over the past several years. Xue now explained that a Great Leap Forward in statistics could not be achieved merely by

830 See: BMA 133-001-00105: 1-10. In his speech at the meeting, Xue Muqiao would claim that Hebei’s statistical reforms heralded the start of a new era in the history of Chinese statistics 开辟一个新的时期. See: Xue Muqiao. “Kuzhan sange yue, gaibian quanguo tongji gongzuo mianlin—shixian quanguo tongji gongzuo dayuejin! 苦战三月, 改变全国统计工作面貌—实现全国统计工作大跃进! (Three Months of Hard Work, Change the Face of Statistical Work—Realize a Great Leap Forward in National Statistical Work!).” Tongji gongzuo 统计工作 1958.14: 2-7.

831 At this point, it is not clear why Xue Muqiao was removed from his post. For more on Jia Qiyun, see: DSJ: 6, 472.

832BMA 133-001-00121: 1-8.

833 Xue Muqiao 1958: 3.

834 Xue Muqiao 1958: 3-5.

341

ensuring the correctness, timeliness, and completeness of periodical reports. And yet, he warned that there were many who continued to hold on to such antiquated beliefs. What they practiced was nothing but ‘statistics for statistics sake’ 为统计而统计. Another common mistake was to regard vigor and dynamism 轰轰烈烈 in statistical work as a

Great Leap Forward in statistics. Unless one also reflected on how one’s statistical work contributed to a high tide 高潮 in production, such vigor was of little use. Xue proceeded to equate both these types of errors, observing that statistical workers of the ‘vigor and dynamism’ brand often ridiculed their ‘statistics for statistics’ comrades. In truth, there was no difference between the two groups; much like Mencius’ allegory about the soldiers who had retreated fifty steps mocking those who had retreated one hundred steps, both groups were guilty of misunderstanding the true nature of a great leap in statistics.

The second characteristic of a great leap in statistics was the mobilization of the masses to actively participate in statistical work. In this way, Xue explained, the theory of the mass line could be implemented in statistics: “Statistics was much like other kinds of work: it cannot rely solely on dedicated organizations or dedicated personnel, nor can it rely on a vertically integrated leadership structure.” Instead, each level of administration was required to participate in statistical work, and even more significantly, the masses themselves could not remain indifferent 漠不关心 to statistical work.835 The idea of mass participation in statistical work was subsequently promoted through the phrase “the entire party and all the people [must] attend to statistics” 全党全民办统计.836

835 “统计工作也象其他工作一样，不能单纯依靠专职的机构和专职的干部来办，也不能单纯的依靠垂直领导思想...群众更对统计工作漠不关心.” See: Xue Muqiao 1958: 4.

836 See, for instance, the editorial in The People’s Daily: “全党全民办统计.” Renmin ribao, 13 August 1958, p. 3.

342

Within a short period of time after the Baoding Conference, publications and reports that sought to distill and spread the message of the Conference or claim a leap in the effectiveness of statistical work began to make an appearance. In August 1958, the

Hebei Statistical Press published The Experience of a Great Leap in Statistics.837 That volume was followed by The High Tide of Statistical Work in Baoding City.838 Early the following year, the Fujian Planning Committee published a volume on how to carry out statistical surveys, explaining that the establishment of People’s Communes 人民公社 required new procedures for statistical work.839 Internal reports began to offer descriptions of what these changes looked like. A summary of the first post-Baoding meeting in Beijing produced by the SSB in November 1959 noted:

Since the Baoding Conference, statistical work has emerged on a new scale. Now, statistical authorities at all levels, from the central to the local, can within a few hours, or within a few tens-of-hours, master statistical data on all aspects of industrial and agricultural production and construction at the national, provincial and regional levels. [And] within three or five days, they can conduct one-time comprehensive surveys, and basically fulfill in very quick time the party leadership’s various data requirements.840

The SSB had strived long years to achieve the goal of meeting the party leadership’s requirements, but here, the Great Leap Forward had helped realize that goal in mere months. Or so it was claimed.

837 Hebei Statistics Bureau. Tongji gongzuo yuejin jingyan 统计工作跃进经验 (The Experience of a Leap Forward in Statistical Work). Hebei People’s Press 河北人民出版社, 1958.

838 Hebei Statistics Bureau . Baodingshi tongji gongzuo de gaochao 保定市统计工作的高潮 (The High Tide of Statistical Work in Baoding). Beijing: State Statistical Press, 1958.

839 Fujian Planning Committee 福建计划委员会. Zenyang kaizhan tongji diaocha yanjiu gongzuo 怎样开展统计调查研究工作 (How to Commence Statistical Survey Research). Fujian People’s Press 福建人民出版社, 1959.

840 BMA 133-001-00105: 2.

343

Regardless of whether such claims were true or not, and they were most certainly

‘true’ only if a host of issues including accuracy and representativeness were kept in abeyance, the very ability to make them pointed to a fundamental shift in the terrain of statistical work. The down-grading of the periodical report system and the promotion of mass participation in statistics effectively bypassed the tussle between socialist statistics and its non-socialist alternatives in favor of a return to Maoist mass science, expressed via the promotion of the Maoist method of social investigation. Discussed in Part I

Chapter I, this form of typical sampling had been largely ignored by the SSB through much of the 1950s. After the Baoding Conference, it was elevated to the status of orthodox method. 841 A March 1960 article explained that:

Typical surveys are an innovative survey method utilized in our nation under particular historical and social conditions; it is the concrete manifestation in statistical survey methodology of Chairman Mao’s synthesis of Marxist-Leninist theory with the practice of revolution in China.842

The authors of the article went on to list the various qualities of typical sampling. They claimed it could solve those problems that exhaustive enumeration could not and it could also help address problems where it was not necessary to carry out exhaustive enumeration. In addition, typical sampling also served to supplement the results of exhaustive enumeration. It was flexible and lightweight 灵活轻便; it helped save on manpower and reduced problems 省人省事, thereby doing more with less 事半功倍.

841 Consider, for instance, the occurrence of “typical sampling” 典型调查 in the title of articles in Tongji gongzuo during the 1950s: seven articles for the five-year period from 1953 to 1957, two articles in 1958, and nine in 1959 alone.

842 Feng Guoxi 冯国熙, et al. “Mao zhuxi lun dianxing diaocha 毛主席论典型调查 (Chairman Mao’s Discussion on Typical Surveys).” Caijing kexue 财经科学 1960.3: 76-81. The quoted section is reproduced from Garnaut 2013: 239.

344

Using it, new things could be discovered or developed in a timely fashion. It was thus like a dagger 短剑, which dissected problems in the national economy and across society, and offered specific analyses of complex social phenomena. In so doing, it also promoted continuous improvements in the style and methods of leadership. Finally, such typical sampling was an important way of cultivating 培养 red and expert 红又专 cadre, and thus avoided problems in [earlier] statistical work, which was divorced not just from reality but also from the masses 脱离实际、脱离群众.843

Rebadged as Mao’s synthesis, such typical surveys quickly became the dominant method for collecting information.844 As Yang Bo explained in the first of a series of articles titled “How to Carry Out Typical Sampling,” which were published in the first five issues of Statistical Work in 1959:

In recent times, there has emerged an extremely valuable new phenomenon among statistical leaders at all levels: [they] often dispatch large numbers of cadres deep into the countryside, factories, mines, shops, basic construction sites, and other grass-roots units to conduct highly specified, on-the-ground research work and thereby grasp the new circumstances and problems of the development of the national economy…[such activity] is the distinctive marker of changes in the style of statistics leadership and in work methods. 845

This state of affairs continued unabated until 1961. In June of that year, Wang Sihua 王

思华 was appointed the SSB’s third director and under him attempts were made to re- establish statistical work along the lines of the pre-GLF years. Even so, the influence of

843 Feng et al., 1960: 79.

844 See: Yang Bo 杨波. “Zenyang zuo dianxing diaocha 怎样做典型调查 (How to Carry out Typical Surveys).” Tongji gongzuo 统计工作 1959.01: 36-37. For a brief discussion of this methodology in the context of the reliability of famine related data, see Garnaut 2013: 238-240. For a discussion of the importance of non-exhaustive forms of surveying 非全面调查, see Feng et al. 1960: 76-77.

845 Yang Bo 1959: 37.

345

mass science, with its emphasis on anti-expertise and anti-professionalization, persisted within the practice of socialist statistics through the rest of that decade and the decade of the 1970s. It is this notion that has come to inform our understanding of statistics during the Mao era.846 But as this dissertation has shown, the picture till 1958 was quite different, far more contested, and open to very different paths.

It was not until 1979 that a debate about socialist statistics was re-kindled. That year, statistician and Renmin University professor Dai Shiguang 戴世光 published an article in the journal Economic Research in which he set about demolishing what he called the theoretical cloak 外衣 of Soviet socialist statistics. The article was republished in the journal Statistics 统计, the descendant of Statistical Work 统计工作, and ignited a major reform movement within the world of Chinese statistics.847 But that is the subject of another dissertation.

846 See, for instance: Liu 2009.

847 Dai Shiguang 戴世光. “Jiji fazhan kexue de tongjixue—wei woguo zaori shixian sige xiandaihua fuwu 积极发展科学的统计学—为我国早日实现四个现代化服务 (Actively Develop Scientific Statistics—In Service to the early realization of the four modernization in our country).” Jingji yanjiu 经济研究 1979.02: 63-67. Republished in Tongji 统计 1979.01: 23-27.

346

REFERENCE MATTER

BIBLIOGRAPHY

ARCHIVES AND LIBRARIES

In China:

Beijing Municipal Archives (BMA) Foreign Ministry Archives of the People’s Republic of China, Beijing (WJB) Institute of Economics, Chinese Academy of Social Sciences, Beijing Institute of Modern History, Chinese Academy of Social Sciences, Beijing National Library of China, Beijing Peking University Library, Beijing Resource Room, Department of Statistics, Renmin University Tsinghua University Archives, Beijing Tsinghua University Library, Beijing

In India:

Indian Council for World Affairs, New Delhi Institute for Economic Growth, New Delhi National Archives of India (NAI), New Delhi National Library, Kolkata Nehru Memorial Museum and Library (NMML), New Delhi P.C. Mahalanobis Memorial Museum and Archives (PCMMMA), Kolkata Tata Institute of Fundamental Research (Archives), Mumbai

In the United States:

Sripati Chandrasekhar Papers, University of Toledo Hoover Institution—Library and Archives, Stanford University

(ORAL HISTORY) INTERVIEWS

Debkumar Bose Kolkata, August 14 2009 V.P. Dutt New Delhi, August 30 2009 Gong Jianyao 龚鉴尧 Guangzhou, September 14 2011

347

Lin Fude 林富德 Beijing, 22 April 2011 Meera Kosambi Pune, August 20 2012 Ni Jiaxun 倪加讯 Beijing, March 29 2011 and September 7 2011 Narayan Sen Singapore, July 23 2012 Wu Cangping 邬沧萍 Beijing, March 11 2011) Wu Hui 吴辉 Beijing, September 8 2011

JOURNALS AND NEWSPAPERS

Caijing yanjiu《财经研究》 Guangming ribao《光明日报》 Jingji zhoubao 《经济周报》 Renmin ribao《人民日报》 Tongji gongzuo (Northeast Statistics Bureau) 《统计工作》 Tongji gongzuo tongxun (State Statistical Bureau) 《统计工作通讯》 Tongji gongzuo (State Statistical Bureau) 《统计工作》 Xin jianshe《新建设》 Zhongguo gongye 《中国工业》

Ananda Bazaar Patrika The Hindu Indian Statistical Institute—Annual Report(s) Samvadadhvam Sankhya The Statesman The Times of India

BOOKS AND ARTICLES

Aird, John S. “Sources of Demographic Data on Mainland China.” American Sociological Review 24, no. 5 (1959): 623-630.

___. “China’s Population, Census and Vital Statistics.” Journal of the American Statistical Association 56 (1961): 44.

___. Estimating China’s Population. The Annals of the American Academy of Political and Social Science 369 (1967): 67-72.

348

Alonso, William and Paul Starr, eds. The Politics of Numbers. New York: Russell Sage Foundation, 1987.

Anderson, T.W., K.L. Chung, and E.L. Lehmann. “Pao-Lu Hsu 1909-1970.” The Annals of Statistics 7, no. 3 (1979): 467-470.

Andreas, Joel. Rise of the Red Engineers: The Cultural Revolution and the Origins of China’s New Class. Stanford: Stanford University Press, 2009.

Anon. “Jinian Wang Sihua tongzhi shishi wuzhou nian 纪念王思华同志逝世五周年 (Commemorating the Fifth Anniversary of the Passing of Comrade Wang Sihua).” Tongji gongzuo 统计工作 (China Statistics) 1983.07: 5-6, 14.

___. “Huiwang: lishi shang de dongbei tongjiju 回望: 历史上的东北统计局 (Looking Back: The History of the Northeast Statistics Bureau).” Tongji gongzuo 中国统计 (Statistical Work), 2010.8: 22-23.

Asen, Daniel. “Dead Bodies and Forensic Science: Cultures of Expertise in China, 1800- 1949.” Columbia University, PhD dissertation, 2012.

Asimov, Isaac. The Foundation Trilogy. London: The Folio Society, 2012.

Ba, Jin 巴金. Ba Jin quanji: di shiwu juan 巴金全集—第十五卷 (The Collected Works of Ba Jin: Vol 15). Beijing: Renmin wenxue chubanshe, 1986-.

Bernstein, Thomas P. and Hua-yu Li, eds. China Learns from the Soviet Union, 1949- Present. New York: Lexington Books, 2010.

Bethlehem, Jelke. “The Rise of Survey Sampling.” Discussion Paper (09015). The Hague: Statistics Netherland, 2009.

Br ard, Andrea. “Robert Hart and China’s Statistical Revolution.” Modern Asian Studies, 40, no. 3 (2006): 605–629.

Br ard, Andrea and Wang Youjun 王幼军. “Tongjixue zai jindai zhongguo de yanbian— cong shehui tongji dao shuli tongji 统计学在近代中国的演变—从社会统计到数理统计 (The Evolution of Statistics in Modern China—From Social Statistics to Mathematical Statistics).” Shanghai jiaotong daxue xuebao (zhexue shehui kexue ban)上海交通大学学报 (哲学社会科学版), 18, no. 3 (2010): 61-67, 74.

Boorman, Howard L. and Richard C. Howard, eds. Biographical Dictionary of Republican China, 2 Vols. New York: Columbia University Press, 1967.

Bowie, Robert and John K. Fairbank, eds. Communist China 1955-1959—Policy Documents and Analysis. Cambridge: Harvard University Press, 1963.

349

Buck, John Lossing. Land Utilization in China, Atlas: A Study of 16,786 Farms in 168 Localities, and 38,256 Farm Families, in Twenty-Two Provinces in China, 1929- 1933. Chicago: University of Chicago Press, 1937.

Buck, Peter. American Science and Modern China, 1876-1936. Cambridge: Cambridge University Press, 1980.

Chatterjee, Partha. “Development Planning and the Indian State.” In The State Development Planning and Liberalisation in India, edited by T J Byres, 82-103. New Delhi: Oxford University Press, 1998.

Cheek, Timothy and Tony Saich, eds. New Perspectives in State Socialism in China. Armonk, NY: M.E. Sharpe, 1997.

Chen, Da 陈达. Population in Modern China. Chicago: University of Chicago Press, 1946.

Chiang, Yung-chen. Social Engineering and the Social Sciences in China, 1919-1949. Cambridge: Cambridge University Press, 2001.

Chen Jiading and Zheng Zhongguo. “Academic Achievements of Professor P.L. Hsu.” Review essay on the website of the International Chinese Statistical Association, 2010. Accessed online: http://www.icsa.org/awards/PLHsu2.pdf.

Chen, Theodore Hsi-en Chen. “The Thought Reform of Intellectuals.” In Government of Communist China, edited by George P. Jan, 477-485. San Francisco: Chandler Publishing Company, 1966.

Chen, X. R., K.T. Fang, and C.C. Yang, eds. The Development of Statsistics: Recent Contributions from China. Harlow, Essex: Longman Scientific and Technical, 1992.

Chu, Jennings P. (See entry for Zhu Zhunyi below).

Chung Lin and Hsia Lu. “How We Made the Census. China Reconstructs, no. 4 (December 1955): 9-11

Coale, Ansley J. Rapid Population Change in China 1952-1982. Washington, D.C.: National Academy Press, 1984.

Cohn, Bernard. An Anthropologist among Historians. Oxford: Oxford University Press, 1990.

Commercial Press. Xinhua zidian—hanyying shuangjie 新华字典—汉英双解 (Xinhua Dictionary with English Translation). Beijing: Commercial Press, 2000.

350

Comte, Auguste. A General View of Positivism [Discours sur l'Esprit positif 1844] London, 1856.

Connelly, Matthew. Fatal Misconception: The Struggle to Control World Population. Cambridge: Belknap (Harvard University Press), 2008.

Cressey, G.B. Cressey. “The 1953 Census of China.” The Far Eastern Quarterly 14, no. 3 (1955): 387-388.

Cullather, Nick. The Hungry World: America’s Cold War Battle Against Poverty in Asia. Cambridge: Harvard University Pres, 2010.

Dai Shiguang 戴世光. “Yindu de quanguo chouyang diaocha—shehui jingji diaocha 印度的全国抽样调查—社会经济调查 (India’s National Sample Survey—A Socio- economic Survey).” Tongji gongzuo 《统计工作》1957.09: 2-6.

___. “Jiji fazhan kexue de tongjixue—wei woguo zaori shixian sige xiandaihua fuwu 积极发展科学的统计学—为我国早日实现四个现代化服务 (Actively Develop Scientific Statistics—In Service to the early realization of the four modernization in our country).” Jingji yanjiu 《经济研究》 1979.02: 63-67. Republished in Tongji 统计 1979.01: 23-27.

___. Dai Shiguang wenji 戴世光文集 (Dai Shiguang—Collected Works) Beijing: Renmin University Press, 2008.

Daston, Lorraine. Classical Probability in the Enlightenment. Princeton: Princeton University Press, 1988. de Bary, William Theodore and Richard Lufrano, eds. Sources of Chinese Tradition, Vol. 2. New York: Columbia University Press, 2000.

Desai, Meghnad. “Development Perspectives: Was there an Alternative to Mahalanobis?” In India's Economic Reforms and Development: Essays for Manmohan Singh edited by Isher Judge Ahluwalia, 43-52. New Delhi: Oxford University Press, 2012.

Desrosières, Alain. The Politics of Large Numbers—History of Statistical Reasoning. Translated by Camille Naish. Cambridge: Harvard University Press, 1998.

___. “How Real Are Statistics? Four Possible Attitudes.” Social Research 68, no. 2, (Summer 2001): 339-355.

Diamant, Neil J. Revolutionizing the Family—Politics, Love, and Divorce in Urban and Rural China, 1949–1968. Berkeley: University of California Press, 2000.

351

___. “Policy Blending, Fuzzy Chronology, and Local Understandings of National Initiatives in Early 1950s China.” Frontiers of History in China 9, no. 1 (2014): 83-101.

Doel, Ronald E. and Thomas Söderqvist, eds. The Historiography of Contemporary Science, Technology, and Medicine—Writing Recent Science. London: Routledge, 2006.

Donnithorne, Audrey. China’s Economic System. New York: Praege, 1967.

Ekbladh, David. The Great American Mission: Modernization and the Construction of an American World Order, 1914 to the Present. Princeton: Princeton University Press, 2010.

Elman, Benjamin. On Their Own Terms: Science in China, 1550-1900. Cambridge: Harvard University Press, 2005.

Erickson, Paul, Judy L. Klein, Lorraine Daston, Rebecca Lemov, Thomas Sturm, and Michael Gordin. How Reason Almost Lost its Mind—The Strange Career of Cold War Rationality. Chicago: University of Chicago Press, 2013.

Eyferth, Jacob. Eating Rice from Bamboo Roots—The Social History of a Community of Handicraft Papermakers in Rural Sichuan, 1920–2000. Cambridge: Harvard University Press, 2008.

___. “Learning to Count, Learning Not to Be Counted: Numbers, Governance, and Resistance in Rural China.” Unpublished Paper Presented at the Workshop on “Nation and Enumeration in Modern China,” Columbia University, New York, 26 October 2013.

Ezhov, A. Soviet Statistics. Translated by V. Shneerson. Moscow : Foreign Languages Publishing House, 1957.

___. Industrial Statistics ( т т т п о ыш о т ). Calcutta: Statistical Publishing House, 1958.

___. Organisation of Statsitics in the U.S.S.R. Moscow: Progress Publishers, 1967.

Feng Guoxi 冯国熙, et al. “Mao zhuxi lun dianxing diaocha 毛主席论典型调查 (Chairman Mao’s Discussion on Typical Surveys).” Caijing kexue 《财经科学》 1960.3: 76-81.

Fei Xiaotong. Peasant Life in China: A Field Study of Country Life in Yangtze Valley. New York : E. P. Dutton, 1939.

352

___. Earthbound China: A Case Study of Rural Economy in Yunnan. Chicago: University of Chicago Press, 1945.

Feng Zhidan 冯之丹. “Yindu renmin de jingji jianshe chengjiu 印度人民的经济建设成就 (The Achievements in Economic Construction of the Indian People).” Renmin ribao January 26 1956.

Feyerabend, Paul. Against Method. Verso Press, 1993.

Fisher, Margaret W. and Joan V. Bondurant. “Review Article: The Impact of Communist China on Visitors from India.” The Far Eastern Quarterly 15, no. 2 (February 1956): 249-265.

Fisher, Irving. “The Best Form of Index Number.” Quarterly Publications of the American Statistical Association 17, no. 133 (March 1921): 536.

___. The Making of Index Numbers: A Study of Their Varieties, Tests, and Reliability. Boston: Houghton Mifflin Compan, 1922.

Friedman, Edward, Paul G. Pickowicz, and Mark Selden. Revolution, Resistance, and Reform in Village China. New Haven: Yale University Press, 2005.

Focus: Science in Modern China.. Isis 8, no. 5. Chicago: University of Chicago Press, 2007.

Fong, H.D. “Recollections of Early Research on Chinese Economy.” The Tsing Hua Journal of Chinese Studies 清华学报 3 & 4, no. 1 (1963): 69-86.

Foucault, Michel. “Governmentality.” In The Foucault Effect: Studies in Governmentality, edited by G Burchell, C Gordon and P Miller. London: Harvester Wheatsheaf， 1981.

Friedman, Milton. Money Mischief –Episodes in Monetary History. Orlando: Harcourt, Brace & Company, 1992.

Fujian Education Press. Ershi shiji zhongguo xueshu dadian: jingjixue—xiace 20 世纪中国学术大点：经济学—下册 (Key Features of Scholarship in Twentieth Century China: Economics—Volume 2). Fujian Education Press 福建教育出版社, 2005.

Fujian Planning Committee 福建计划委员会. Zenyang kaizhan tongji diaocha yanjiu gongzuo 怎样开展统计调查研究工作 (How to Commence Statistical Survey Research). Fujian People’s Press 福建人民出版社, 1959.

Fukuyama, Francis. “What is Governance?” Center for Global Development, Working Paper 314, January 2013.

353

Gamble, Sydney. Ting Hsien: A North China Rural Community. New York: International Secretariat Institute of Pacific Relations, 1954.

___. North China Villages: Social, Political, and Economic Activities Before 1933. Berkeley: University of California Press, 1963.

Gao, James Z. The Communist Takeover of Hangzhou: The Transformation of City and Cadre, 1949–1954. Honolulu: University of Hawai‘i Press, 2004.

Gao, Mobo C.F. Gao Village—A Portrait of Rural Life in Modern China. London: Hurst and Company, 1999.

Garnaut, Anthony. “Hard Facts and Half-Truths: The New Archival History of China’s Great Famine. China Information 27, no. 2 (2013): 223-246.

Gong Jianyao 龚鉴尧. “Woguo chouyangfa yanjiu yu shijian de fazhan 我国抽样法研究与实践的发展 (Development of the Research and Practice of Sampling in our Country).” Xi’an tongji xueyuan xuebao《西安统计学院学报》 9, no. 2 (1994): 7-14.

___. Shijie tongji mingren zhuanji 世界统计名人传记 (Biography of Famous Statisticians of the World). Beijing: Statistical Press, 2000.

___. “Shidao qinghuai 师道情怀 (Sentiments from a Career in Teaching).” China Statistics 《中国统计》, (May 2001): 52-54.

Gordin, Michael D. “Statitique et revolution en Russie: Un compromisie impossible (1880-1930), and: L’anarchie bureaucratique: Statistique et pouvoir sous Staline.” Kritika: Explorations in Russian and Eurasian History 5, no. 4 (Fall 2004 (New Series)): 803-810.

Gore, A.P. “P. V. Sukhatme: A 'Social' Statistician.” Economic and Political Weekly 32, no. 6 (Feb. 8-14, 1997): 257-258.

Government of India, Ministry of Food and Agriculture. Report of the Indian Delegation to China on Agricultural Planning and Techniques. New Delhi: Government of India, 1956.

Government of India, Planning Commission. Report of the Indian Delegation to China on Agrarian Cooperatives. New Delhi: Government of India, May 1957.

Greenhalgh, Susan. Just One Child—Science and Policy in Deng's China. Berkeley: University of California Press, 2008.

354

De Grief A, Alexis and Mauricio Neito Olarte. “What We Still Do Not Know About South-North Technoscientific Exchange—North-centrism, Scientific Difussion, and Social Studies of Science.” In The Historiography of Contemprary Science, Technology, an Medicine—Writing Recent Science, edited by Ronald E. Doel and Thomas Söderqvist, 239-259. London: Routledge, 2006.

Ramachandra Guha. An Anthropologist Among Marxists and Other Essays. New Delhi: Permanent Black, 2001.

Guo Gengji 过庚吉. “Fandui sanbu zichan jieji tongji sixiang dusu—ping jinguobao zhu he zou yi ren zhu 反对散布资产阶级统计思想毒素——评金国宝著《工业统计学原理》和邹依仁著《工业统计》 (Combat the Spread of the Toxin that is Bourgeois Statistical Thought—A Review of Jin Guobao’s Principles of Industrial Statistics and Zou Yiren’s Industrial Statistics).” 1952.

Gui Shizuo 桂世祚. “Ping Zou Yiren, Jin Guobao liangwei xiansheng de gongye tongjixue zhuzuo 评邹依仁、金国宝两位先生的工业统计学著作 (Criticisms of Zou Yiren and Jin Guobao’s Works on Industrial Statistics).” Tongji gongzuo 《统计工作》, (1956: 24-27).

Hacking, Ian. “Biopower and the Avalanche of Printed Numbers.” Humanities in Society 5 (1982): 279–95.

___. The Taming of Chance. Cambridge: Cambridge University Press, 1990

___. “Making up People.” Essay based on “Tenth Annual British Academy Lecture” of April 2006; full text of talk in the Proceedingsof the B.A.. London Review of Books, 28, no. 16 ( 17 August 2006): 7-11.

Halberstam, David. The Best and the Brightest. New York: Random House, 1972.

Hanyu dacidian Press. Hanyu dacidian—di jiu juan 汉语大词典—第九卷. Beiing: Hanyu dacidian Press.

Harris, Seymour E. “Appraisals of Russian Economic Statistics—Introduction.” The Review of Economics and Statistics 29, no. 4 (November 1947): 213-214.

Hebei Statistics Bureau. Tongji gongzuo yuejin jingyan 统计工作跃进经验 (The Experience of a Leap Forward in Statistical Work). Hebei People’s Press 河北人民出版社, 1958.

___. Baodingshi tongji gongzuo de gaochao 保定市统计工作的高潮 (The High Tide of Statistical Work in Baoding). Beijing: State Statistical Press, 1958.

355

He Sheng 荷生. “Ping “tongjixue yuanli--diyijuan” 评《统计学原理》(第一卷) (Review of Principles of Statistics (Volume I)”).” Jingji zhoubao 《经济周报》 (Economic Weekly) 12, no. 3 (1950).

Heinzig, Dieter. The Soviet Union and Communist China, 1945-1950: The Arduous Road to the Alliance. M.E. Sharpe, 2004.

Hershatter, Gail. The Gender of Memory—Rural Women and China’s Collective Past. Berkeley: University of California Pres, 2011.

Hsu, Kai-yu, ed. Literature of the People’s Republic of China. Bloomington and London: Indian University Press, 1980.

Hu Daiguang 胡代光. “Ganqing zichan jieji tongji sixiang de liudu 肃清资产阶级统计思想的流毒 (Eliminate the Pernicious Influence of Bourgeois Statistical Ideology).” Tongji gongzuo《统计工作》(1955.10)

Hu, Danian. China and Albert Einstein—The Reception of the Physicist and His Theory in China, 1917-1979. Cambridge: Harvard University Press, 2005.

Huang Yingwei 黄英伟. “Gongfenzhi de zhidu jingjixue fenxi 公分制的制度经济学分析 (Analysis of the Institional Economics of the Work-Point System).” Ershiyi shiji shuangyue kan 《二十一世纪双月刊》, no. 131 (June 2012a): 87-96.

___. Gongfenzhi xia de nonghu laodong 工分制下的农户劳动 (Rural Labour under the Work-Point System. Beijing: China Agriculture Press, 2012b.

Immerwahr, Daniel. The Village Ideal: The United States, Community Development, and the World. Cambridge, Mass.: Harvard University Press, 2014.

International Statitical Institute. Introduction—International Statistical Conferences, September 6-18, 1947, Washington, D.C. — Proceedings of the International Statistical Conference, Volume I and II. Calcutta: Alka Press, 1951.

Iriye, Akira. “Internationalizing International History.” In Rethinking American History in the Global Age, edited by Thomas Bender, 48-60. Berkeley: University of California Press, 2002.

Jaffe, A.J. “A Review of the Censuses and Demographic Statistics of China.” Population Studies 1, no. 3 (February 1948): 308-337.

Jay, Anthony and Jonathan Lynn. The Complete Yes Minister. London: BBC Books, 1989.

Jiang Xuepei 江学培. “Beijing daxue jinian—zhuming shuxuejia, tongjixuejia, Xu Baolu jiaoshou 北京大学纪念—著名数学家、统计学家许宝騄教授 (Beijing

356

Univeristy Memorial—The Famous Mathematician and Statistician, Professor Xu Baolu).” Tongji 《统计》, (1981.01): 39.

Jin Guobao 金国宝. Tongjixue dagang 统计学大纲 (An Outline of Statistic)s. Shanghai: Commercial Press, 1934.

___. “Dui ‘Tongjixue dagang’ de ziwo pipan 对“统计学大纲”的自我批判 (Self- Criticism on ‘An Outline of Statistics’).” Tongji gongzuo 《统计工作》, 1957.3: 28-32.

Kaple, Deborah A. “Soviet Advisors in China in the 1950s.” In Brothers in Arms: The Rise and Fall of the Sino-Soviet Alliance, 1945-1963, edited by Odd Arne Westad, 117-140. Stanford: Stanford University Press, 1998.

Katzenstein, Peter and Stephen Nelson. „Worlds in Collision: Uncertainty and Risk in Hard Times.” In Politics in the New Hard Times: The Great Recession in Comparative Perspective, edited by Miles Kahler and David Lake, 233-52. Ithaca: Cornell University Press, 2013.

Kendall, D.G. “Andrei Nikolaevich Kolmogorov, 25 April 1903-20 October 1987.“ Biographical Memoirs of Fellows of the Royal Society 37 (1991): 300–26.

Kevles, Daniel J. In the Name of Eugenics: Genetics and the Uses of Human Heredity. Cambridge, MA: Harvard University Press, 1985.

Kirby, Maurice W. Operations Research in War and Peace: The British Experience from the 1930s to 1970. London: Imperial College Press, 2003.

Kirby, William C. “Continuity and Change in Modern China: Chinese Economic Planning on the Mainland and on Taiwan, 1943-1958.” Australian Journal of Chinese Affairs 24 (July 1990): 121-141.

Knight, Nick. “Mao ZeDong's on Contradiction and on Practice: Pre-Liberation Texts.” The China Quarterly, no. 84 (Dec., 1980): 641-668.

Knorr-Certina, Karin. Epistemic Cultures—How the Sciences Make Knowledge. Cambridge, London: Harvard University Press, 1999.

Kotz, S. a. E. S. “Lenin as Statistician: A Non-Soviet View.” Journal of the Royal Statistical Society. Series A (Statistics in Society) 153, no. 1 (1990): 73-94.

Krotevich, Sergei K. “Vsekitayskaya perepis’ naseleniya 1953 g. (The All-China Population Census of 1953).” Vestnik Statistiki (Journal of Statistics), no. 5 (September-October 1955): 31-50.

357

___. Zhongguo 1953 nian quanguo renkou diaocha 中国 1953 年全国人口调查 (The All-China Population Census of 1953). Beijing: Statistical Press, 1956.

Krüger, Lorenz, Lorraine J. Daston, Michael Heidelberger, eds. The Probabilistic Revolution - Vol. 1: Ideas in History. Cambridge: MIT Press, 1987.

Kulp, Daniel. Country Life in South China: The Sociology of Familism. Volume 1. New York City: Bureau of Publications, Teachers College, Columbia University, 1925.

Lam, Tong. A Passion for Facts: Social Surveys and the Construction of the Chinese Nation-State, 1900–1949. Berkeley: University of California Press, 2011.

Lampland, Martha and Susan Leigh Starr, eds. Standards and their Stories: How Quantifying, Classifying, and Formalizing Practices Shapes Every Life. Ithaca, NY: Cornell University Press, 2009.

Lampland, Martha. False Numbers as Formalizing Practices. London: LSE Centre for the Philosophy of Natural and Social Science Contingency and Dissent in Science Technical Report July 2009. Accessed: http://www.lse.ac.uk/CPNSS/research/concludedResearchProjects/ContingencyDi ssentInScience/DP/DPLampland0709Online.pdf

Lardy, Nicholas R. and Kenneth Lieberthal, eds. Chen Yün's Strategy for China's Development: A Non-Maoist Alternative. Armonk, NY: M.E. Sharpe, 1983

Latour, Bruno. Science in Action, How to Follow Scientists and Engineers through Society. Cambridge: Harvard University Press, 1987.

Lee, Christopher J., ed. Making a World after Empire: The Bandung Moment and Its Political Afterlives. Athens: Ohio University Press, 2010.

Lee, Hong Yung. From Revolutionary Cadres to Party Technocrats in Socialist China. Berkeley: University of California Press, 1991.

Lei, Sean Hsiang-Lin. “How Did Chinese Medicine Become Experiential? The Political Epistemology of Jingyan.” Positions: East Asia Cultures Critique 10.2 (2002): 333-364.

Lenin, Vladimir I. Lenin Collected Works, Vol. 22, 33, and 35. Moscow: Progress Publishers, 1960 (Reprinted 1977-1978).

Li, Choh-Ming. Economic Development of Communist China. Berkeley: University of California Press, 1959.

358

___. “Communist China’s Statistical System: 1949-1957.” The American Economic Review (Papers and Proceedings of the Seventy-Third Annual Meeting of the American Economic Association) 51, no. 2 (May 1961): 499-517.

___. The Statistical System of Communist China. Berkeley: Univ. of California Press, 1962a.

___. “Statistics and Planning at the Hsien Level in Communist China.” The China Quarterly, No. 9 (Jan. – Mar., 1962b): 112-123.

Li, Deng 李澄. “Jinian Wang Sihua tongzhi 纪念王思华同志 (Commemorating Comrade Wang Sihua).” Renwen zazhi 《人文杂志》 (The Journal of Humanities(Bimonthly)), no. 6 (1984): 64-65, 68.

Li, Lillian M. Fighting Famine in North China—State, Market, and Environmental Decline, 1690s-1990s. Stanford: Stanford University Press, 2007.

Li Yuren. “Wo dui tongji gongzuo de renshi 我对统计工作的认识 (My Understanding of Statistical Work).” Tongji gongzuo 《统计工作》 1955.10: 53

Li Zhenzhou 李振周. “Chedi pipan yu kefu women nongye tongji gongzuo zhong de youqing baoshou sixiang 彻底批判与客服我们农业统计工作中的右倾保守思想 (Thoroughly Critique and Overcome Rightist Conservative Ideas in our Agricultural Statistics Work).” Tongji gongzuo 《统计工作》(1956.4).

Lieu, D.K. 刘大钧 (Liu Dajun). “Statistical Work in China.” Presented at the Nineteenth Session of the International Statistical Institute, Tokyo, Japan, 1930.

___. “Collecting Statistics in China.” The American Statistician 2, no. 6 (December 1948): 12-13.

Lin Yueh-hwa. The Golden Wing: A Sociological Study of Chinese Familism. Greenwood Press, 1947.

Lindsay, Michael. Notes on Educational Problems in Communist China, 1941-1947. New York: Institute of Pacific Relations, 1950.

Liu, Chang 刘畅 and Zhang Yun 张云. Ershi shiji zhongguo de tongjixue 二十世纪中国的统计学 (Statistical Education in 20th Century China). Beijing: Dangjian duwu, 2000.

Liu, Lydia H. Translingual Practice—Literature, National Culture, and Translated Modernity, 1900-1937. Stanford: Stanford University Press, 1995.

359

Liu, Xin. The Mirage of China: Anti-Humanism, Narcissism, and Corporeality of the Contemporary World. New York: Berghahn Books, 2009.

Lu Ch’i 陸棨 (Lu Qi). “Willow Village Revisited.” In Literature of the People’s Republic of China, edited by : Hsu Kai-yu, 729. Bloomington and London: Indian University Press, 1980.

Lü , Xiabo and Elizabeth Perry, eds. Danwei: the Changing Chinese Workplace in Historical and Comparative Perspective. New York: M.E. Sharpe, 1997.

Lüthi, Lorenz. The Sino-Soviet Split: Cold War in the Communist World. Princeton: Princeton University Pres, 2008.

Ma Feng 马烽. “San nian qian zao zhidao 三年前早知道 (I Knew Three Years Ago).” In Literature of the People’s Republic of China, 565. Bloomington and London: Indian University Press, 1980.

MacFarquhar, Roderick and John K. Fairbank, eds. The Cambridge History of China, Vol. 14: The People's Republic, Part 1: The Emergence of Revolutionary China, 1949- 1965. Cambridge: Cambridge University Press, 1987.

MacFarquhar, Roderick. The Politics of China 1949-1989. Cambridge: Cambridge University Press, 1993.

MacKenzie, Donald A. An Engine, Not a Camera: How Financial Models Shape Markets. Cambridge: MIT Press, 2006.

Mahalanobis, P.C. “Why Statistics?” Sankhya: The Indian Journal of Statistics 10, no. 3 (September 1950): 195-228.

___. “Statistics Must Have a Purpose.” Samvadadhvam 1, no. 1 (July 1956): 3-10.

Mann, Michael. “The Autonomous Power of the State: Its Origins, Mechanisms and Results.” European Journal of Sociology 25, no. 2, (November 1984): 185-213.

Manning, Kimberly Ens and Felix Wemheur, eds. Eating Bitterness—New Perspectives on China’s Great Leap Forward and Famine. Vancouver: University of British Columbia Press, 2011.

Mao, Tse-Tung 毛泽东. Selected Works of Mao Tse-Tung, Volume I. Peking: People’s Publishing House, 1952.

___. Report on an Investigation of the Peasant Movement in Hunan. Peking: Foreign Language Press (3rd edition; first edition 1953), 1965.

360

___. Selected Works of Mao Tse-Tung, Volume III. Beijing: People’s Publishing House, 1966.

___. Selected Works of Mao Tse-Tung, Volume V. Beijing: People’s Publishing House, 1977.

___. A Critique of Soviet Economics by Mao Tsetung (translated by Moss Roberts; annotated by Richard Levy, with an introduction by James Peck). London: Monthly Review Press, 1977.

Martin, Bernard. Strange Vigour. Port Washington, NY: Kinnikat Press， 1970.

Marx, Karl. Capital, Volume 1: A Critique of Political Economy (Translated by Samuel Moore, Edward Bibbins Aveling, and Ernest Untermann , edited by Frederick Engels). London: Modern Library, 1906.

Mazur, Mary G. “The United Front Redefined for the Party-State: A Case Study of Transition and Legitimation,” In New Perspectives in State Socialism in China, edited by Timothy Cheek and Tony Saich, 51-75. Armonk, NY: M.E. Sharpe, 1997.

McCormick, Ted. William Petty: And the Ambitions of Political Arithmetic. Oxford: Oxford University Pres, 2009.

Meisner, Maurice. Mao’s China and After—A History of the People’s Republic. New York: Free Press (Simon and Schuster), 1999.

Memorandum Outlining a Plan of Economic Development for India. Penguin Edition, 1944.

Meng, Yue. “Hybrid Science versus Modernity: The Practice of the Jiangnan Arsenal.” East Asian Science, Technology, and Medicine 16 (1999): 13-52.

Mills, Harriet C. “Thought Reform: Ideological Remolding in China.” In Government of Communist China, edited by George P. Jan, 486-498. San Francisco: Chandler Publishing Company, 1966.

Mitchell, Timothy. Rule of Experts: Egypt, Techno-Politics, Modernity. Berkeley: Univ. of California Press, 2002.

Moraes, Frank. Report on Mao’s China. New York: The MacMillan Company, 1953.

Mullaney, Thomas. Coming to Terms with the Nation: Ethnic Classification in Modern China. Berkeley: University of California Press, 2011.

361

Nathan, Andrew. “China’s Work-Point System: A Study in Agricultural “Splittism.” Current Scene—Developments in Mainland China 2, no. 31, (April 15, 1964): 1- 13.

National Bureau of Statistics 国家统计局. Zhonghua remin gongheguo tongji dashiji 1949-2009 中华人民共和国统计大事记 1949-2009 (Chronicle of Events in Statistics in China). Beijing: China Statistics Press, 2009.

Nixon, J. W. A History of the International Statistical Institute 1885-1960. The Hague: International Statistical Institute, 1960.

Orleans, Leo A. Professional Manpower and Education in Communist China. Washington, D.C.: U.S. Government Printing Office, 1961.

___. “Chinese Statistics: The Impossible Dream.” The American Statistician 28, no. 2 (May, 1974): 47-52.

Ostrovitianov, K.V. and J.M. “The Discussion on Statistics Summed Up.” Soviet Studies 6, no. 3 (Jan. 1955): 321-331.

Oxford University Press. The Oxford Dictionary of Statistical Terms 2003. Oxford: Oxford University Press, 2003.

___. Oxford English Dictionary (Online Edition). Oxford: Oxford University Press.

Panikkar, K.M. In Two Chinas: Memoirs of a Diplomat. London, G. Allen & Unwin, 1955.

Parsons, Talcott. “The Professions and Social Structure.” Social Forces 17, no. 4 (May, 1939): 457-467.

Passin, Herbert. “Sino-Indian Cultural Relations.” The China Quarterly, no. 7 (July- September 1961), pp. 85-100.

Patriarca, Silvana. Numbers and Nationhood: Writings Statistics in Nineteenth-Century Italy. Cambridge: Cambridge University Press, 1996.

Pickowics, Paul G. and Jeremy Brown, eds. Dilemmas of Victory--The Early Years of the People's Republic of China. Cambridge: Harvard University Press, 2007.

Poovey, Mary. A History of the Modern Fact: Problems of Knowledge in the Sciences of Wealth and Society. Chicago: University of Chicago Press, 1998.

Porter, Theodore. The Rise of Statistical Thinking, 1820-1900. Princeton, Princeton University Press, 1986.

362

___. “Chance Subdued by Science. Poetics Today 15, no. 3 (Fall 1994): 467-478.

___. Trust in Numbers: The Pursuit of Objectivity in Science and Public Life. Princeton: Princeton University Press, 1996.

Rao, C.R. “Prasantha Chandra Mahalanobis, 1893-1972.” Biographical Memoirs of Fellows of the Royal Society 19 (December 1973): 454-492

___. “ISI Honorary Member Interviews: Professor C.R. Rao.” International Statistical Institute Newsletter 29, no. 2 (2005)

Rawski, Thomas G. “Studies of China’s Economy.” In A Scholarly Review of Chinese Studies in North America. Asia Past & Present: New Research from AAS, 2013.

Renmin University. Tongjixue yuanli jiangyi—shangxia ce 统计学原理讲义—上下册 (Lectures on Statistical Principles—Volumes 1 and 2). Beijing: People’s University Press, 1956.

___. Tongjixue jiangyi 统计学讲义 (Lectures in Statistics), Six Volumes. Beijing: Renmin University Press, 1959.

___. Xiaoyou jiniance 校友纪念册 (Alumni Record). Beijing: People’s University Press, 2007

Riskin, Carl. China’s Political Economy—The Quest for Development since 1949. Oxford: Oxford University Press, 1987.

Rogaski, Ruth. Hygienic Modernity: meanings of health and disease in treaty-port China. Berkeley: University of California Press, 2004.

Rosensweig, Phil. “Robert S. McNamara and the Evolution of Modern Management.” Harvard Business Review, December 1, 2010.

Rowe, William T. “ Bao Shichen (1775-1855), an Early Nineteenth-Century Agrarian Reformer.” Unpublished paper delivered at the Modern China Seminar, Columbia University, New York, December 15 2012.

Rudra, Ashok. Prasanta Chandra Mahalanobis: A Biography. New Delhi: Oxford University Press, 1996.

Schmalzer, Sigrid. The People’s Peking Man: Popular Science and Human Identity in Twentieth-Century China. Chicago: University of Chicago Press, 2008.

Schneider, Laurence. Biology and Revolution in 20th Century China. Rowan and Littlefield, 2003.

363

Schrader, Charles S. History of Operations Research in the United States Army, Vol. 1 1942-1962. Washington, D.C.: US Army, 2006.

Schran, Peter. The Development of Chinese Agriculture, 1950-1959. Urbana: University of Illinois Press, 1969.

Schumpeter, Joseph. Ten Great Economists: From Marx to Keynes. London: Allen and Unwin, 1952.

Schurmann, Franz. Ideology and Organization in Communist China. Berkeley: University of California Press, 1968.

Scott, James C. Seeing Like a State: How Certain Schemes to Improve the Human Condition Have Failed. New Haven: Yale University Press, 1998.

___. The Art of Not Being Governed—An Anarchist History of Upland Southeast Asia. New Haven: Yale University Press, 2009.

Shen, Zhihua 沈志华. Sulian zhuanjia zai zhongguo (1948-1960) 苏联专家在中国 (Soviet Experts in China). Beijing: Xinhua Press, 2003.

Shi Qing 石青. “Pipan jiben jianshe tongji gongzuo zhong tuoli shiji de xianxiang批判基本建设统计工作中脱离实际的现 (Criticisms Regarding Divorcing from Reality the Statistical Work in Capital Construction).” Tongji gongzuo 《统计工作》 (1956.19).

Shi Tai 施泰. “Jin Guobao 金国宝 (1893-1963).” Tongji yu yuce 《统计与预测》, no. 6 (1994): 60-61.

___. “Zou Yiren 邹依仁.” Tongji yu yuce 《统计与预测》, no. 6, (2001): 52-53.

Shwed, Uri and Peter Bearman. “The Temporal Structure of Scientific Consensus Formation.” American Sociological Review 75, no. 6 (2010): 817-840.

Si Gengsheng 司更生. “Fanlan chengzai de diaocha tongji biao 泛滥成灾的调查统计表 (Swamped by Statistical Tables).” RMRB, September 12 1953, 3.

Singh, K. Natwar. My China Diary 1956-88. New Delhi: Rupa & Co., 2009

Smith, Aminda. Thought Reform and China's Dangerous Classes Reeducation, Resistance, and the People. Rowman and Littlefield, 2013.

Smith-Peter, Susan. “Defining the Russian People: Konstantin Arsen'ev and Russian Statistics Before 1861.” History of Science 45, part 1 (2007): 47-64.

364

Song Tao 宋涛 ed. 20 shiji zhongguo xueshu dadian: jingjixue—xiace 20世纪中国学术大点：经济学—下册 (Major Themes in Twentieth Century Chinese Academics: Economics—Volume 2). Fujian Education Press 福建教育出版社, 2005.

Song, Xuguang 宋旭光. “Xin zhongguo jingji tongji fazhan de lishi kaocha 新中国经济统计发展的历史考察 (Historical Investigation of the development of Economic Statistics in New China).” Caijing wenti yanjiu 《财经问题研究》 (Research on Issues in Finance and Economics), no. 3 (2013): 21-26.

Spence, Jonathan D. The Search for Modern China. New York: W.W. Norton and Company, 1990.

State Statistics Bureau. Guanyu yijiu wusan niandu guomin jingji fazhan he guojia jihua zhixing jieguo de gongbao 关于一九五三年度国民经济发展和国家计划执行结果的公报 (Bulletin Concerning the Results of National Economic Development and Plan Implementation for 1953). Beijing: Statistical Press, 1953.

___. Guanyu yijiu wusi niandu guomin jingji fazhan he guojia jihua zhixing jieguo de gongbao 关于一九五四年度国民经济发展和国家计划执行结果的公报 (Bulletin Concerning the Results of National Economic Development and Plan Implementation for 1954). Beijing: Statistical Press, 1954.

___. Tongji gongzuo zhongyao wenjian huibian 《统计工作重要文件汇编》 (Compilation of Important Documents Relating to Statistical Work). Beijing: Statistical Press, 1955, 1957, 1959.

___. Guanyu yijiu wuliu niandu guomin jingji jihua zhixing jieguo de gongbao关于1956 年度国民经济计划执行结果的公报(Bulletin Concerning the Results of National Economic Plan Implementation for 1956). Beijing: Statistical Press, 1956.

___. Jiceng nongye tongji gongzuo jingyan huibian 基层农业统计工作经验汇编 (Compilation of Experiences Conducing Grassroots Agricultural Statistics Work). Beijing: China Statistical Press, 1956.

___. Re’ai women de tongji gongzuo 热爱我们的统计工作 (Ardently Love our Statistical Work). Beijing: China Statistical Press, 1956.

___. Tongji lilun yiban wenti jianghua 统计理论一般问题讲话 (Common Problems in Statistical Theory). Beijing: China Statistical Press, 1956.

___. Zenyang zuo gongye tongji fenxi gongzuo 怎样做工业统计分析工作 (How to Carry out Analysis of Industrial Statistics). Beijing: China Statistical Press, 1956.

___. Tongji ganbu de zixue 统计干部的自学 (The Self-Study of Statistical Cadre). Beijing: Statistical Press, 1957.

365

___. “Ganqing, chongtian—quanguo shengshi tongji gongzuo baoding xianchang huiyi shengli bimu 干劲冲天、信心百倍—全国省市统计工作保定现场会议胜利闭幕 (Energetic and Brimming with Confidence—The National Provincial and City Level On-the-Spot Meeting in Baoding Comes to a Successful Conclusion).” Tongji gongzuo 《统计工作》1958.14: 25.

___. Guanyu fazhan guomin jingji de diyi ge wunian (1953 nian dao 1957 nian) jihua zhixing jieguo de gongbao 关于发展国民经济的第一个五年 (1953年到1957年) 计划执行结果的公报(Bulletin Concerning the Results of the First Five-year (1953-1957) Plan for the Development of the National Economy). Beijing: Statistical Press, 1958.

___. Guojia tongjiju guanyu 1958 nian guomin jingji fazhan qingkuang de gongbao 国家统计局关于1958年国民经济发展情况的公报 (Bulletin Concerning the National Economic Developments in 1958). Beijing: Statistical Press, 1958.

___. Jisuanji de susuanfa 计算机的速算法 (The Methods of Rapid Calculations using Calculating Machines). Beijing: Statistical Press, 1958.

___. Woguo gongye tongji gongzuo de jingyan 我国工业统计工作的经验(The Experience of Industrial Statistical Work in our Country). Beijing: China Statistical Press, 1958.

___. Yindu mahalanuobisi jiaoshou fanghua tongji baogaoji 印度马哈拉诺比斯教授访华统计报告集 (Report on the visit to China of the Indian Statistician, Prof. Mahalanobis). Beijing: China Statistical Press, 1958.

___. Weida de shinian 伟大的十年 (Ten Great Years). Beijing: People’s Press, 1959.

___. “Communique of the Major Statistics from the first national population census as related by the State Statistical Bureau of the People's Republic of China (November 1, 1954).” Chinese Sociology and Anthropology 16, no. 3-4 (Spring- Summer 1984): 62-5.

Stiffler, Douglas. “Building Socialism at Chinese People’s University: Chinese Cadres and Soviet Experts in the People’s Republic of China, 1949-1957.” University of California, San Diego, PhD Dissertation, 2002.

Strauss, Julia, ed. “The History of the People's Republic of China, 1949-1976.” The China Quarterly Special Issue. Cambridge: Cambridge University Press, 2007.

366

Thampi, Madhavi ed. India and China in the Colonial World. New Delhi: Social Science Press, 2005.

Malcolm Thompson. “The Birth of Chinese Population: A Study in the History of Governmental Logics.”University of British Columbia, PhD Dissertation, 2013.

Thompson, Roger, tr.. Report from Xunwu. Stanford: Stanford University Press, 1990.

Tie Hua 铁华 and Liu Xin 刘新. “Lun tongjixue de duixiang ji qi kexue jichu—ping Gou Shisheng zhu 论统计学的对象及其科学基础—评勾适生著 <统计学原理> (Discussion of the Object and Theoretical Basis of Statistics—A Review of Gou Shisheng’s Principles of Statistical).” Xin jianshe 《新建设》 (New Construction) 4, no. 5 (1951).

Tobin, James. Essays in Economics: Macroeconomics. Chicago: Markham, 1972; reprint, Cambridge, MA: MIT Press, 1987.

Trescott, Paul B. Jingji Xue: The History of the Introduction of Western Economic Ideas into China, 1850-1950. Hong Kong: Chinese University Press, 2007.

U, Eddy. “The Making of Zhishifenzi: The Critical Impact of Registration of Unemployed Intellectuals in the Early PRC.” The China Quarterly, no. 173 (2003): 100-121.

United Nations Statistical Commission. “The Preparation of Sampling Survey Reports.” Statistical Papers, Series C, No. 1. Lake Success, New York, 1950.

___. “Recommendations for the Preparation of Sampling Survey Reports (Provisional Issue).” Statistical Papers, Series C, No. 1, Rev. 2. New York, 1964.

Vogel, Ezra. “From Revolutionary to Semi-Bureaucrat: The “Regularisation” of Cadres.” The China Quarterly, no. 29 (Jan. – Mar. 1967): 36-60.

___. Canton Under Communism: Programs and Politics in a Provincial Capital, 1949- 1968. Cambridge: Harvard University Press, 1969.

Wakelam, Randall Thomas. The Science of Bombing: Operational Research in RAF Bomber Command. Toronto: University of Toronto Press, 2009.

Wang Feixian 王飞宪 . “Guanyu jiben tongjixue yuanli jiaocai de yixie yijian 关于几本统计学原理教材的一些意见 (A Few Recommendations for Textbooks of Basic Statistics).” Tongji gongzuo 《统计工作》(1956.16).

367

Wang Jianzhen 王健真. “Dui zichan jieji zhishu ‘lilun’ de pipan 对资产阶级指数‘理论’ 的批判 (Criticisms of the Bourgeois Theory of Index Numbers).” Tongji gongzuo 《统计工作》(1955.11).

Wang Sihua 王思华. Zibenlun yanjiu congshu 资本论研究丛书. Harbin: New China Bookstore 新中国书局, 1949.

___. Zibenlun jieshuo 资本论解说. Beijing: Sanlian Publishers 三联出版社, 1950.

___. “Tongji gongzuo the renwu 统计工作的任务 (The Tasks of Statistical Work).” In Tongji gongzuo—di yi xuanji 统计工作—第一选集 (Statistical Work—The First Anthology), 7-14. Shenyang: Northeast Statistical Press, February 1952.

___. “San nian lai dongbei tongjiju gongzuo zongjie 三年来东北统计局工作总结 (Summary of Statistical Work Performed by the Northeast Statistical Bureau over the Past Three Years).” Tongji gongzuo—di wu xuanji 统计工作—第五选集 (Statistical Work—Fifth Anthology), Northeast Finance and Economics Press, January 1953, 1-10.

___. “Jieshao yindu tongji gongzuo, jiji kaizhan quanguo chouyang diaocha 介绍印度统计工作，积极开展全国抽样调查 (Introducing India’s Statisitcal Work, Actively Carrying out a National Sample Survey).” Tongji gongzuo 《统计工作》, 1957.06: 8-10, 17.

___. Wang Sihua tongji lunwenji 王思华统计论文集 (Collected Statistical Works of Wang Sihua). Beijing: China Statistics Press, 1986.

Wang Sili 王思立. “Pipan wo zai tongjxue zhuzuo zhong de zichan jieji sixiang 批判我在统计学著作中的资产阶级思想 (Criticism of the Presence of Bourgeois Ideology in my Statistical Work).” Tongji gongzuo 《统计工作》 1957.10.

Ward, Michael. Quantifying the World: UN Ideas and Statistics. Bloomington: Indiana University Press, 2004.

Westad, Odd Arne, ed. Brothers in arms: the rise and fall of the Sino-Soviet alliance, 1945-1963. Stanford: Stanford University Press, 1998.

Wu, Cangping 邬沧萍. Wu Cangping zixuanji 邬沧萍自选集 (Personally Selected Writings of Wu Cangping). Beijing: Renmin University Press, 2007.

Wu Hui 吴辉. “Zhongguo de tongji diaocha fangshi 中国的统计调查方式 (China’s Statistical Survey Methods).” In Zhongguo shehui zhuyi tongji gongzuo de jianli yu fazhan 中国社会主义统计工作的建立与发展 (The Establishment and Development of Socialist Statistical Work in China), edited by the China State

368

Statistical Bureau’s Institute of Statistical Science. Beijing: China Statistical Press, 1985.

Xin Hua 辛华, ed. Eyu xingming yiming shouce 俄语姓名译名手册 (Handbook of Translations of Russian Names). Beijing: Shangwu yinshuguan 商务印书馆 (Commercial Press), 1982.

Xu Qian 徐前 and Liu Xin 刘新. “Guanyu zichan jieji tongji lilun de pipan 关于资产阶级统计理论的批判 (Criticisms of Bourgeois Statistical Theory).” Xin jianshe 《新建设》, no. 1 and 2 (1953). The authors noted that the article was drafted by them in December 1952.

___ . “Guanyu zichan jieji tongji lilun de pipan 关于资产阶级统计理论的批判 (Criticisms of Bourgeois Statistical Theory).” Tongji gongzuo tongxun 统计工作通讯 (Statistical Work Bulletin), (1955.08).

Xu Qian 徐前. “Tongji xueshuo fazhan bianhua de lunkuo 统计学说发展变化的轮廓 (An Outline of the Developments and Changes in Statsitical Theory).” In Tongjixue yuanli ziliao huibian 统计学原理资料汇编 (Compilation of Materials on Statistical Theory). Beijing: Zhongyang guangbo dianshi daxue chubanshe, 1983.

Xu Youchun徐友春. Minguo renwu dacidian 民国人物大辞典 (Biographical Dictionary of Republican China). Hebei People’s Press, 1991.

Xue Muqiao 薛暮桥. “Di yige wunian jihua qijian woguo tongji gongzuo de chubu jingyan he jinhou renwu 第一个五年计划期间我国统计工作的初步经验和今后任务 (The Preliminary Experience and Recent Tasks of Statistical Work during our country’s First Five-Year Plan Period).” Tongji gongzuo 《统计工作》 1957.21: 1-21

___. “Kuzhan sange yue, gaibian quanguo tongji gongzuo mianlin—shixian quanguo tongji gongzuo dayuejin! 苦战三月, 改变全国统计工作面貌—实现全国统计工作大跃进! (Three Months of Hard Work, Change the Face of Statistical Work— Realize a Great Leap Forward in National Statistical Work!).” Tongji gongzuo 《统计工作》1958.14: 2-7.

___. Xue Muqiao—huiyilu 薛暮桥—回忆录 (Xue Muqiao—Memoirs). Tianjin: Tianjin People’s Press, 2006.

Yang Bo 杨波. “Zenyang zuo dianxing diaocha 怎样做典型调查 (How to Carry out Typical Surveys).” Tongji gongzuo 《统计工作》 1959.01: 36-37.

369

Yang Jianbai 杨坚白. “Sulian zhuanjia bangzhu women jianli le kexue de tongji gongzuo 苏联专家帮助我们建立了科学的统计工作 (Soviet Experts have Helped us Establish Scientific Statiatical Work).” Renmin ribao (People’s Daily), February 05, 1953: 3.

Yang, Kuisong 杨奎松. “The Sino-Soviet Alliance and Nationalism: A Contradiction.” Parallel History Project on NATO and the Warsaw Pact, The Cold War History of Sino-Soviet Relations, June. Originally published in Social Sciences in China (English Edition), Spring 2005.

Yates, F. and K. Mather. “Ronald Aylmer Fisher 1890-1962.” Biographical Memoirs of Fellows of the Royal Society, 9 (1963): 91–129.

Yu Yue 于越. “会计当而已矣.” Tongji yanjiu 《统计研究》 (Statistical Research), (1986.01).

Yugow, Alexander. “Economic Statistics in the U.S.S.R.” The Review of Economics and Statistics 29, no. 4 (November 1947): 242-246.

Zarkovic, S.S. “Note on the History of Sampling Methods in Russia.” Journal of the Royal Statistical Society. Series A (General) 119, no. 3 (1956): 336-338

Zelin, Madeleine. The Magistrate’s Tael: Rationalizing Fiscal Reform in Eighteenth Century Ch'ing China. Berkeley: University of California Press, 1984.

Zhang Minru 张敏如.“ Dui ‘nongye tongjixue’ yishu de jidian yijian 对‘农业统计学’一书的几点意见 (Some Suggestions Concerning the Book “Agricultural Statistics”).” Tongji gongzuo《统计工作》 (1956.12).

Zhang Zhiji 张知几 and Zou Yiren 邹依仁, eds. Dazhong tongji 大众统计 (People’s Statistics).Shanghai: Xin zhishi chubanshe 新知识出版社 (New Knowledge Press), 1953.

Zhou Enlai 周恩来. “On the Question of Intellectuals.” In Communist China 1955- 1959—Policy Documents and Analysis, edited by Robert R. Bowie and John K. Fairbank, 128-144. Cambridge: Harvard University Press, 1963.

Zhu Junyi (Jennings P. Chu) 朱君毅. “The Independently Controlled Statistical System of the Chinese Government.” In Proceedings of the International Statistical Conferences, Vol. II, 92-97. Washngton, D.C.: U.N. World Statistical Confernce, 1947.

___. Minguo shiqi de zhengzhi tongji gongzuo 民国时期的政治统计工作 (Government Statistical Work during the Republican Period). Xiandai tongji zhishi congshu 现代统计知识丛书, 1988.

370

Zou Yiren 邹依仁. “Guanyu zichanjieji tongji xueshu sixiang de ziwo pipan 关于资产阶级统计学术思想的自我批评 (Self-Criticism Regarding Bourgeois Statistical Thought).” Tongji gongzuo 《统计工作》 1956.8.

___. “Lun tongji shi yimen shehui kexue 论统计是一门社会科学 (Discussion of Statistics as a Social Science).” Caijing yanjiu 《财经研究》1957.04.

Zu Ting’an 祖廷安. “Siben gongye tongji shuji 四本工业统计书籍 (Four Books on Industrial Statistics).” Tongji gongzuo 《统计工作》(1958.01): 25.

371