Aggregate Data and Study of Political Development

Swarthmore College Works

Political Science Faculty Works Political Science

2-1-1969

Aggregate Data And Study Of Political Development

Raymond F. Hopkins Swarthmore College, [email protected]

Follow this and additional works at: https://works.swarthmore.edu/fac-poli-sci

Part of the Political Science Commons Let us know how access to these works benefits ouy

Recommended Citation Raymond F. Hopkins. (1969). "Aggregate Data And Study Of Political Development". Journal Of Politics. Volume 31, Issue 1. 71-94. DOI: 10.2307/2128561 https://works.swarthmore.edu/fac-poli-sci/160

This work is brought to you for free by Swarthmore College Libraries' Works. It has been accepted for inclusion in Political Science Faculty Works by an authorized administrator of Works. For more information, please contact [email protected]. Aggregate Data and the Study of Political Development Author(s): Raymond F. Hopkins Source: The Journal of Politics, Vol. 31, No. 1 (Feb., 1969), pp. 71-94 Published by: The University of Chicago Press on behalf of the Southern Political Science Association Stable URL: http://www.jstor.org/stable/2128561 Accessed: 24-08-2015 18:34 UTC

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/ info/about/policies/terms.jsp

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

The University of Chicago Press and Southern Political Science Association are collaborating with JSTOR to digitize, preserve and extend access to The Journal of Politics. http://www.jstor.org

This content downloaded from 130.58.65.13 on Mon, 24 Aug 2015 18:34:02 UTC All use subject to JSTOR Terms and Conditions AGGREGATEDATA AND THE STUDY OF POLITICAL DEVELOPMENT*

RAYMONDF. HOPKINS

SwarthmoreCollege

I N THE LAST DECADE, the rapid emergenceof new less developed states has precipitatedinterest among political scientists in problemsof "politicaldevelopment." Going beyond mere historical accounts of the emergenceof new governinginstitutions, various scholarshave attemptedto spell out the basic processesinvolved in politicalchange and the creationof government.'Largely ignor- ing traditionalindicators such as constitutions,scholars have recent- ly suggestedvariables such as capacityfor responsiveness,division of political tasks, participationin decision-makingand provision of welfareas key indicatorsof politicalgrowth and the formation of healthyand viable polities. A few researchers,influenced by thesenew formulations,have collectedcross-national data on traits which seem related to political development.Their effortshave been aimed at refining,specifying, or testingrelationships among suggestedindicators and causal variables.Although an important focus of all types of studies has been on new dimensionsof political development,it is not clear to what extent their ideas about politicaldevelopment are valid, useful,or in agreement. My purpose in this paper is to comparethe resultsof some recenttheorizing about politicaldevelopment with effortsto collect data on the subject.A basic problemin advancingour understand-

*I wish to thank Hayward R. Alker, Jr., who provided helpful counsel; Ronald D. Brunner, Robert 0. Keohane, J. Roland Pennock and Bruce M. Russett, who read and commented on drafts of this article; and the Yale Political Data Program with its grant from the National Science Foundation for its financialsupport. 'See for example, the differencebetween a purely historical account of political development as in J. Clagett Taylor, The Political Development of Tanganyika (Stanford: Stanford University Press, 1963), and attempts to understandthe processes and problems of development,as in David E. Apter, Ghana in Transition (New York: Atheneum,1963), Gabriel A. Almond and James S. Coleman, The Politics of Developing Areas (Princeton: Princeton UniversityPress, 1960), and Lucian W. Pye, Politics, Personality and Nation Building (New Haven: Yale UniversityPress, 1962). ,[ 71 ]

This content downloaded from 130.58.65.13 on Mon, 24 Aug 2015 18:34:02 UTC All use subject to JSTOR Terms and Conditions 72 THE JOURNAL OF POLITICS [Vol. 31 ing of politicsis the gap betweentheory and empiricalevidence. I hope to contributeto the closingof this gap by suggestingsome inadequaciesin both recenttheory and data collectionsrelevant to politicaldevelopment. My discussionof neithersubject is exhaus- tive,but by analyzingcomparatively both theoryand data, I hope some of the problemsin the currentliterature will be highlighted. I began my analysis by selectingfour major data studies: Tke Cross-PolitySurvey by ArthurS. Banks and Robert B. Tex- tor; World Handbook of Political and Social Indicatorsby the Yale Political Data Program(Bruce M. Russett,Hayward R. Al- ker, Jr.,Karl W. Deutsch, and Harold D. Lasswell); the Dimen- sionalityof Nations Project by Rudolph J. Rummel,and studies of political developmentby Phillips Cutright.2Variables with ex- plicitlypolitical denotationswere selected fromthese studies for analysis. These data were then subjectedto a series of statistical refinementsand a factoranalysis was performedon the resulting data. Schemata fromthree sources were then selected for exam- ination: Gabriel Almond,the ComparativePolitics Committeeof the SSRC, and Samuel Huntington.3 'See ArthurS. Banks and Robert B. Textor,The Cross-PolitySurvey (Cambridge:MIT Press,1963), hereafterreferred to as B & T; Bruce M. Russett,Hayward R. Alker,Jr., et al., WorldHandbook of Politicaland Social Indicators(New Haven: Yale UniversityPress, 1964), hereafterreferred to as YPDP, for the Yale PoliticalData Program;Rudolph J. Rummel,Harold Guetzkow,Jack Sawyer,and RaymondTanter, Dimensionality of Nations (forthcoming),hereafter referred to as DON; and PhillipsCutright, "National Political Development:Measurement and Analysis,"American Sociological Review (April,1963), pp. 253-264,and PhillipsCutright, "Political Structure, EconomicDevelopment, and National Social SecurityPrograms," American Journalof Sociology(March, 1965), pp. 537-550,hereafter referred to as Cut- right. 'See GabrielA. Almond,"Political Systems and PoliticalChange," Ameri- can BehavioralScientist, June, 1963; GabrielA. Almond,"A Developmental Approachto PoliticalSystems," World Politics, January, 1965; and Gabriel A. Almondand G. BinghamPowell, Jr., Comparative Politics: A Developmental Approach(Boston: Little,Brown and Co., 1966). The Social ScienceResearch Committee'sComparative Politics Committee ideas are summarizedin Lucian W. Pye, Aspectsof PoliticalDevelopment (Boston: Little,Brown and Co., 1966), pp. 45-48 and 62-65. The themesof politicaldevelopment which Pye notesreflect, according to him,"the work of severalmembers of the Com- parativePolitics Committee of the Social ScienceResearch Council, including in particular,Leonard Binder,James S. Coleman,Joseph LaPalombara and MyronWeiner." These themesrun sporadicallythrough the earliervolumes in theStudies in PoliticalDevelopment series. This series,published by Prince- ton UniversityPress, includes the followingsix volumes: 1.Communications

This content downloaded from 130.58.65.13 on Mon, 24 Aug 2015 18:34:02 UTC All use subject to JSTOR Terms and Conditions 1969] THE STUDY OF POLITICAL DEVELOPMENT 73

Before discussingthe results of the factor analysis, I shall brieflyreview the analytical views on political developmentproposed by these authors.This reviewwill then forma basis upon whichto compareand contrastthese theoreticaldimensions of de- velopmentwith the resultsof the factoranalysis.

GabrielAlmond asserts that all politicalsystems may be compared at threelevels of functionalanalysis. These are: capabilities, conversionfunctions, and systemmaintenance and adaptive functions. Each of theselevels can subsequentlybe brokeninto several distinctcategories or aspects. He suggests,for example,that capa- bilitiesanalysis can be brokeninto fivefunctional types: regulative, extractive,distributive, symbolic, and responsive.4This complex framework,however, only servesto embrace,and perhaps confuse but not to describethe basic dimensionsof development.In his book with Powell,Almond concludes that politicalfunctions at all levels can be measuredaccording to threecriteria of development; these are autonomy,differentiation, and secularization.5Autonomy is linked to pluralismwhich is linked to democracyin Almond's schema. The notion put forthis that "subsystemautonomy" is a characteristicof development,and hence in societieswhere political,economic or religioussubsystems are capable of dominating

and PoliticalDevelopment, edited by LucianW. Pye, 1963; 2. Bureaucracyand PoliticalDevelopment, edited by JosephLaPalombara, 1963; 3. PoliticalMod- ernizationin Japanand Turkey,edited by RobertE. Ward and DankwartA. Rustow,1964; 4. Educationand PoliticalDevelopment, edited by JamesS. Coleman,1965; 5. Political Cultureand Political Development,edited by Lucian W. Pye and SidneyVerba, 1965; 6. PoliticalParties and PoliticalDe- velopment,edited by JosephLaPalombara and Myron Weiner,1966. The third theoreticaldiscussion selected for review is Samuel P. Huntington, "PoliticalDevelopment and PoliticalDecay," WorldPolitics (April, 1965), pp. 386-430. 'Similarly,conversion functions fall intosix categories:1) interestarticula- tion, 2) interestaggregation, 3) rule making,4) rule application,5) rule adjudication,and 6) communication.System maintenance and adaptivefunc- tionsinvolve two functionalprocesses: socialization and recruitment. 'In theirwords, "classification of politicalsystems is a developmentalone in whichthe variables of structuraldifferentiation, autonomy and secularization are relatedto otheraspects of the functioningof particularclasses of political systems-theirconversion characteristics, capabilities, and systemmaintenance patterns,"Almond and Powell,op. cit.,p. 300.

This content downloaded from 130.58.65.13 on Mon, 24 Aug 2015 18:34:02 UTC All use subject to JSTOR Terms and Conditions 74 THE JOURNAL OF POLITICS [Vol. 31 all social life, as for example occurs in totalitariansocieties, de- velopmentis retarded.Differentiation, Almond's second dimension of development,refers to the specializationof role functions.In more developed political systems,different roles, such as sheriff, judge, and tax collector,replace the single role, such as headman or chief,in performingsuch basic functionsas apprehendingcrim- inals, metingout justice,and raisingrevenues. The thirdcriterion of development,secularization, is an index based on the instru- mentalismor rationalityof the politicalculture. Secular goals and ends-meanscalculations increasingly predominate in thereflections of policy makers.The developmentof bureaucracieswith an accom- panyingincrease in recordkeeping, codification of laws, and for- mal proceduresfor evaluation of policies are indicatorsof the spread of rational choice, calculation, and control. Almond's functionalistapproach, therefore, has identifiedthree dimensions of political development:autonomy (pluralism), differentiation(role specialization),and secularization(bureaucratization), which are distinctif not uncorrelatedmeasures.6 A second list of dimensionsof developmenthas been drawn up by the Social Science Research Council's ComparativePolitics Committeeand is summarizedby Lucian Pye.7 The dimensions they suggest are: first,equality, which includes widespreadpar- ticipationand suffrage,equality and universalityof law, and political role recruitmentbased on achievement; second, capacity, which includes wider scope of political performance,effectiveness in executingpublic policy, rational administrationand secular policy orientation;and third,differentiation, which occurs when political officesand agencies have distinctand limitedfunctions, roles are assigned to specificfunctions, and complex structures and parts of the politicalsystem are integrated.There are obvious similaritieswith Almond's typology. To beginwith, differentiation is suggested by both as a dimensionof development.Capacity (SSRC), a broaderrubric than secularization,subsumes Almond's secularizationcriteria along with the notionof scope or magnitude

'Almond and Powell apparentlybelieve that differentiationand secularization are highlycorrelated, as indicated by their combinationof them into a single scale for categorizationof some countryexamples. Op. cit., p. 308. The paren- thetical descriptionsof each of the dimensionsabove are suggestedas possible synonymsfor Almond's terms. 7Pye, Aspects,pp. 45-48.

This content downloaded from 130.58.65.13 on Mon, 24 Aug 2015 18:34:02 UTC All use subject to JSTOR Terms and Conditions 1969] THE STUDY OF POLITICAL DEVELOPMENT 75 of politicalperformance. Finally, the SSRC dimensionof equality generallyparallels Almond's autonomy,but measures individual ratherthan systemdifferences. While both these remainingdimen- sions seem to have liberal democracyas their touchstone,the SSRC committeeemphasizes the sharingof powerthrough broader participation,widespread suffrage, and the eliminationof privilege, while Almond stresses the independenceof societal subsystems (the traditionally"liberal" value). Samuel Huntingtonbegins his discussionof political develop- mentwith the argumentthat a distinctionshould be made between modernizationand politicaldevelopment. The latter,he suggests,is oftenfalsely equated with the introductionof moderntechnology and social processes,for example,mass communicationsand ur- banization.Huntington classifies under four categories-rationaliza- tion,national integration, democratization, and mobilization-those characteristicsoften erroneously associated with development. Hunt- ingtonargues that political development should be conceivedof as a process independentof modernization,and definesit as "institu- tionalizationof political organizationand procedures."8While ad- mittingthat the strengthof politicalorganizations and procedures depends upon both a) their "scope of support" and b) their "level of institutionalization,"Huntington discusses only the level of institutionalizationas the core of political development.He thus ignoresthe developmentcharacteristics suggested by "scope of support,"an omissionwhich I will mentionagain later. He views institutionalizationas movementalong four continua: 1) adaptability-rigidity,2) complexity-simplicity,3) auton.omy-subordination, and 4) coherence-disunity.9 There are two obvious similaritiesand one contrastin this list of fourcriteria compared with the two earlierlists. Both the continua of complexity-simplicityand autonomy-subordinationare related to differentiation.Complexity, for example,requires "differentiationof separatetypes of organizationalsubunits" and autonomy relatesto "the extentto whichpolitical organizations and procedures exist independentlyof othersocial groupings."Moreover, Hunting- ton notes the interrelationof these two continuaor dimensions: "the complexityof a political systemcontributes to its autonomy

"Huntington,op. cit., p. 393. 9Ibid.,pp. 394-401.

This content downloaded from 130.58.65.13 on Mon, 24 Aug 2015 18:34:02 UTC All use subject to JSTOR Terms and Conditions 76 THE JOURNAL OF POLITICS [Vol. 31 by providinga variety of organi'zationsand positionsin which individualsare prepared for the highestoffices."'0 Huntington's coherencedimension is relatedto the SSRC's aspect of differentiation consi'stingof the integrationof complexstructures. In contrastto thesesimilariti'es, Almond and Huntingtonuse "autonomy" withnearly opposite emphases. Almond sees greaterdevelopment in societieswhich are not dominatedby a politicalsystem, but have subsystemsindependent and autonomousof political control,thus capable of articulatingtheir independent interests. Huntington, on the otherhand, is concernedthat the political subsystem should have greaterstrength, and thusbe autonomousor freefrom the influence and controlof military,economic, or internationalinterests. Both argue for equal ratherthan dominantinfluence for the "political" sphereof a society,but they reach this agreementfrom opposite concerns. Huntington,as I noted, passes over quickly, apparentlyex- cludingas a developmentcriterion, the scope of politicalsupport, which he describesas the extent to which a "a large segment of the populationis politi'callyorganized."" Almond, likewise, does not suggestscope, magnitude,or degreeof participationas a cri- terionfor development.However, his notionof state-building,discussed elsewhere,is related to these concepts.'2 In contrast,the SSRC Committeeincludes this quality under two of their three dimensions,first as participationunder the equalitydimension, and second, as scope of politi'calperformance under capacity. The differencesamong the authorslargely reflecttheir vary- ing concerns.Huntington, for instance,is clearly anxious about stability.The result is that he analyzes characteristicswhich he believes are associatedwith politicalinstitutions (such as parties, bureaucracies,legislatures, and courts) that are free fromcoups, internalviolence, and otherconflicts associated with disorder,dis- integration,and decay. Almondand the SSRC Committee,despite some differenceswe have noted,have similarteleologies. The We- ber-Parsonsvision of modernization(pattern variables) and various ingredientsof the democraticmodel informboth their lists of

"Ibid., p. 402. "Huntington, op. cit., p. 394. "Almond and Powell, op. cit., pp. 35-36.

This content downloaded from 130.58.65.13 on Mon, 24 Aug 2015 18:34:02 UTC All use subject to JSTOR Terms and Conditions 1969] THE STUDYOF POLITICALDEVELOPMENT 77 developmentdi"mensions.'3 Criteria stressing achievement, specialization,secularization, rationality and universalityof thelaw, clearly the influenceof Weber and Parsons,are encapsulatedby theirdimen- sions of differentiation,secularization, and capacity. Almond'sin- clusion of autonomyor pluralismand the SSRC's elaborationof equality are clear evidence, I believe, of effortsto include at least some facet of democraticprocesses in lists of development criteri'a.Thus the goals of stability,and alternatively,of democracy and rationality,are the valued outcomesof developmentwhich shape the alternativemeasures of developmentproposed by these authors. Certainlythe exhaustivenessand, perhapsalso, the consistency and mutual exclusivenessof the categoriesin each of these three typologiesmight be questioned.Moreover the conceptsemployed may embracemore than one empiricallydistinct phenomenon or it may be that two dimensionsfrom one list are so highlycor- relatedin the real world as to make theiroperational distincti'on misleading.In the face of such an increasingnumber of new and differentterms, concepts and typologies,social scientistsare wi'se to weigh carefullythe usefulnessof each new productof theory construction.The purposeof this summary,however, has not been to dissector criticizethese threelists of developmentcriteri'a ex- haustively.'4My purposeis ratherto see how well or convenient- ly politicalvariables commonly associated with development cluster or groupthemselves in some fashionor patternresembling dimensions proposedby these prominenttheorists. The procedurein the constructionof developmentcriteria em- ployedby the theoristsis to tryintuitively to isolateand aggregate salientdifferences from a comparativeanalysis of politicalsystems

"3SeeMax Weber, The Theory of Social and Economic Organization,edited by Talcott Parsons (New York: The Free Press, 1964), pp. 56-77, 115ff; Talcott Parsons and Edward A. Shils, editors, Toward a General Theory of Action (Cambridge: Harvard University Press, 1951) and Talcott Parsons, The Social System (New York: The Free Press, 1951). "4Otherlists mighthave been selected for review; for example, a similarlist suggestedby Joseph LaPalombara, ap. cit., pp. 39ff.which includes 1) struc- tural differentiation,2) magnitude,3) achievementorientation, and 4) degree of secularizationmight have been discussed. Or, a very differentlist, urging the measurementof outputs of a political system could have been included. See, for example,J. Roland Pennock, "Political Development,Political Systems and Political Goods," World Politics (April, 1966) pp. 415-434, who suggests indices of 1) security,2) welfare,3) justice and 4) liberty.

This content downloaded from 130.58.65.13 on Mon, 24 Aug 2015 18:34:02 UTC All use subject to JSTOR Terms and Conditions 78 THE JOURNAL OF POLITICS [Vol. 31 in order to constructa typology. This enables the plottingof politicalsystems along the typology's traits or dimensionsof development. Such an analytical-deductiveapproach is basic to social scienceresearch and those engagedin gatheringdata have a boot- less task withoutthis theoreticaldirection.15 After theoristshave conceived approaches, dimensions,and causes of developmentin this fashion,data orientedresearchers in- formedby these theorieshave organizeda numberof relevant measures. But thereis still an incompletestep in the circulation of ideas about development.The final step is to generatefeed- back, to see how well data collectorshave been measuringtraits similar to those suggestedby the most recent theories,and to suggestwhat changes or new directionsin data collectionor in theoreticalperspectives may be required.The next sectionof this paper will attemptto promotesuch a dialogue,employing, in contrast to the theorists,an inductive,data based approach to the constructionof politicaldevelopment measures.

II Factor analysisallows one to identifya set of dimensionsunderlying those political data which are available for the study of comparativedevelopment. The results are useful for evaluating various theoriesto the extentto which the variablesused in the factoranalysis are valid and are relevantto politicaldevelopment. The real merit of this techniqueis as an approach to simpli- fying,clarifying, and criticizingthe presentarray of data being used by studentsof political development.Table I presentsthe results of such a factoranalysis.1- It contains the orthogonally 1"Banksand Textor,op. cit.,for instance, draw heavily upon thefunctional categoriesproposed by GabrielAlmond and JamesS. Coleman,op. cit. Russett, et al., op. cit.,likewise are influencedby the theoreticalconsiderations of such menas Deutsch,Lerner and Pye. Conffictdata, which is of interestto Hunt- ington,see op. cit.,p. 402 and 416, is includedas a major domainof Rum- mel'sdata catalogue. "6Thecalculations were done at the Yale ComputerCenter using the YCC factoranalysis program and the IBM 7094. The principalcomponent technique was selectedand orthogonal(varimax) rotation was performed.Since the rotatedfactor structure was quitesimilar to the set of unrotatedfactors, and yetprovided clearer indications of the constituentvariables of a particular factor,it was selectedfor presentation. For a discussionof factoranalysis see R. J. Rummel,"Understanding Factor Analysis," Journal of ConflictResolu- tion (December,1967) pp. 444-480and HarryHarman, Modern Factor Anal- ysis (Chicago: Universityof ChicagoPress, 1967).

This content downloaded from 130.58.65.13 on Mon, 24 Aug 2015 18:34:02 UTC All use subject to JSTOR Terms and Conditions 19691] THE STUDY OF POLITICAL DEVELOPMENT 79 rotatedfactor loadings on 43 developmentvariables. These numbers, called "loadings," indicate the correlationof each variable with each underlyingdimension or factorwhich is extractedin the analysis.17 In orderto interpretand label the factors,those variables whichhad "high" loadings,generally .50 or above, as indicatedin Table 1, have been examined. The 43 variables-measuredacross 85 countries-wereselected for several reasons. The initial criterionwas that each variable be distinctivelypolitical. Thus data on communicationflows, eco- nomicgrowth, trade (as in YPDP and DON), and historicaland areal variables (in B & T) were excluded. DON variables were selected fromthe domains of internalpolitics and domesticcon- flictand YPDP variablesfrom the sectionon "governmentand politics." Criteriafor eliminatingmissing data and non-normaldistri- butions were used to complete the selection of the variables."8 The eight factorspresented in Table 1 account for 73.1 percent of the variance among the variables.19The communality (h2) of all variables is above .50 with the exceptionof four "The eightfactors in Table 1 are presentedin the descendingorder of varianceexplained (which is somewhatdifferent than the orderof the factors beforerotation.) 18Thesecond criterion for screening variables was missingdata. Beginning with 133 countriesand territories,the same as in the YPDP Handbook,the numberof observationswas reducedto a final85, afterdropping countries with20 or moremissing variables. Similarly, variables with less than 60 ob- servationswere dropped. The thirdcriterion was distribution.DON variables (73 and 75), legitimacyand defenseexpenditure, were dropped from the analysis because theirnon-normal distributions could not be normalized.The 85 countriesincluded in the analysisare U.S., U.K., Canada, Ireland,West Germany,France, Belgium, Netherlands, Italy, Switzerland,Austria, Sweden, Denmark,Norway, Finland, Iceland, Portugal, Spain, Greece,Turkey, Yugo- slavia, U.S.S.R., Poland, E. Germany,Czechoslovakia, Hungary, Rumania, Bulgaria,Albania, Mexico, Guatemaha,El Salvador, Honduras,Nicaragua, Costa Rica, Panama, Cuba, DominicanRepublic, Haiti, Venezuela,Colombia, Ecuador,Brazil, Peru, Bolivia,Paraguay, Uruguay, Argentina, Chile, Israel, Ethiopia,Egypt, Saudi Arabia,Iraq, Syria,Lebanon, Jordan, Tunisia, Iran, Indonesia,Malaya, Pakistan,India, Ceylon,Burma, Thailand,Philippines, Japan, South Korea, Taiwan, China (mainland),Australia, New Zealand, South Africa,Ghana, Madagascar, Afghanistan, South Vietnam,Nepal, Cam- bodia,Laos, Liberia,Mongolia, Luxembouirg. 18 Factorswith eigenvalues greater than 1.0 wererotated. In all, tenfactors, accountingfor 81.1 percentof the variance,met this test; but onlyeight of these ten rotatedfactors seemed worth discussing.The two factorsnot presentedaccounted for 4.1 percentand 3.9 percentof the varianceand each had one highloading variable. These were number of assassinationsand theratio of governmentrevenue to expenses,respectively.

This content downloaded from 130.58.65.13 on Mon, 24 Aug 2015 18:34:02 UTC All use subject to JSTOR Terms and Conditions 80 THE JOURNAL OF POLITICS [Vol. 31

w _ oo (7el Ch % 0% 0% (- &- _q m w inCe w 0% O 4, 4. %q A:

0% 0%os 00 t- %C In 0% em in) oo oo m %Oeq %Co

O1 W E > N O. O0. O.~~CN N O. V- O.cOl rA + 11 qt 0 C4in 00 in "4 %C qdw 0 %O 0% O %Ome+w0 ?; > OI OI VI VI O.OOO O . .O HO U NbXwOoeN+Nbw02O

ooin ? ., e., m (7? t7 0 i n m in 00r ?> cl

~~~~~~~~~~~C,400 in 0 e4 %a _ t-_1 I% 00 2 NoU

z e + e o' e ce) O m b ?~~~~~~~~~~~~O e N N .. C

00 <

O I t; :~~~~~zz z z Zz Z;i z z z z z z z z z zZi zSzO

or. ? ti + e o b oo oo ~~~t- N %. C% 4 C-. C-. -4 'w U- %o >~~~~~~~~~C,-1 _" C', c" cn C-. C-' C-. "t "t In U- t- + - % m_

P4 z"l Q) >t P P4 v ;;, (Sv < ?. ? ^? ? $ 4 Pz U) $ 4 > > _, N Xt~~~~~~~~~~~~~- v; -4 0 ti . .woo _. &J _ z__ _qNN6

This content downloaded from 130.58.65.13 on Mon, 24 Aug 2015 18:34:02 UTC All use subject to JSTOR Terms and Conditions 1969] THE STUDY OF POLITICAL DEVELOPMENT 81

Q !0

d- O. . O. O. O. O. . . . O. . 0.w

. 40 e me + o CMO.ow7e4 O.0\ 0 "i _4O 0% 0% 0% 0 O . .. 00 00 m O e4 - "...... O. O. .S .o .M O O. O. .M_1 O.

in+t-_ "i w+t t-o 1O in1 00 0% 0% 00 C i %O t0 Ct ~~~~~~O O. .HO .O

C 1 1 1 1 1

_1~~~~~e o CdtoomCd ooed ed tt +, _ -O HNO.No.o. O. O._, o

Z O ZO OZCZ CZ gpio4 gpi CZ CZ z

_/, _ -4 I/ z

I24o od Ooe 2 > 0 >) , v MMSgX - QQQ

V;vttA &4 U) OtO * . . . . \0 00* . . . . . * ...... 4- U) ~.bO \ b U) 0S 0 (U Cl w CM 0 e4 :t S e '. eW e;. +

This content downloaded from 130.58.65.13 on Mon, 24 Aug 2015 18:34:02 UTC All use subject to JSTOR Terms and Conditions g2 THE JOURNALOF POLITICS [Vol. 31

DON measures: legalityof governmentchange, number of political parties, number of assassinations,and ratio of government revenueto expenses.Otherwise, the eight factorsolution accounts for a majorityof the variati'onfound in the selected variables. The firstfactor, which accounts for nearly 30 percentof the varianceamong the variables,correlates highly with variables from all the data sourcesexcept the YPDP. Amongthe high loading vari'- ables are Press Freedom,Party Competition, Existence of Opposition Groups,Interest Articulation by Associationalrather then Non-asso- ciational Groups,Elitism (negativeloading), Multi'-partysystems, and Cutright'smeasures of democraticPolitical Development. The Banks and Textorvariables, owing to the codingprocedure, indicate thereverse condition of thei'rvariable names.22 This dimensionI label "Power Sharing."23It relatesto the concernfor the establishment of democraticinstituti'ons and suggeststhat a pluralisticsociety which shares decision-makingamong various groups is of major interestamong the data collectors.It may be that this factor's predominanceis largelya productof preferencesfor Western demo- craticprocedures shared by both some data collectorsand theory-

20Followingeach variablename is a notationas to its sourceand its origi- nal identificationnumber (in parentheses).Specifically see Russett,et al., op. cit., pp. 56-101; Rummel, et al., op. cit., and Rummel, "Dimensionality of NationsProject: VariableDefinitions, Data Sourcesand Year," (manuscript, July, 1964); Banks and Textor, op. cit., pp. 67-114; and Cutright op. cit. The Cutrightvariable 24 is fromhis 1963study and was takenfrom the DON data, #79a. Variable43 was obtainedfrom a set of raw code sheetssupplied by Cutrightfor his 1965study and is based on figuresfor 1961. 21Thepercent of variance"explained" by a factoris 100 x 2a; 2/N,whaere ai is the loadingof the jth variableon the factorand N is the numberof variables(43 in this case). "Common"variance would generallyrefer to all 10 rotatedfactors. See HarryHarman, op. cit.,p. 198. 22Banksand Textordata are specifiedby alphabeticalcategories, A, B, etc., whereA is the characteristicmentioned in the variablename. For example, variable13, Freedom of thePress, was codedby Banksand Textorso thatA= completefreedom, B=less freedom,and so on. In my coding,A's wererecord- ed as 0, B's as 1, etc.,so thatthe signs on Banksand Textorvariables indicate the reverseof the qualityindicated. This explainswhy oppositesigns are foundon similarvariables, for instance, in factor1, variable2, Freedomof the Press (B & T) loads - .94 and variable20, No Press Censorship,(DON), has a loadingof .76. These qualitiesare in fact positively,not negatively, related. "8Thisrepresents a view of developmentsimilar to one proposedby Harold D. Lasswellin "The Policy Sciencesof Development,"World Politics (Janu- ary,1965), pp. 293ff.

This content downloaded from 130.58.65.13 on Mon, 24 Aug 2015 18:34:02 UTC All use subject to JSTOR Terms and Conditions 1969] THE STUDY OF POLITICAL DEVELOPMENT 83 builders.24By recognizingthis possiblevalue bias, we indicatethat the factoris not necessarilythe most importantcriterion for mea- suringdevelopment based on empiricalinter-nation differences. More- over, not all indicatorsof democracycontributed fully to this dimension.Some of the highloading variables on this factor(Cut- right'svariables in particular)have uniquesecondary characteristics as will be illustratedby factorseven. It is clear, however,that this polyarchicor power sharing quality is a major dimenslon by whichdata collectorshave measuredpolitics across nations. The seconddimension is labeled "ExecutiveStability." Although it accounts for far less of the variance (8.2%) than the first factor,it is importantbecause each of the threemajor data collectors(YPDP, B & T, DON) representedin this studyexpressed interestin similar though operationallydistinct measures of ex- ecutive tenure,and each of these is highlycorrelated with this singlefactor.25 These two factsenhance the validityof the "Execu- tive Stability" factor. Huntington'scriteria of developmentare directlyrelevant to factor II. A lengthy and peaceful tenure amongpolitical executives serves to indicateboth autonomy and co-

24A numberof the variables included in this dimensionare fromBanks and Textor and of these several were influencedby Almond's analysis of political functionsin Almond and Coleman, op. cit., pp. 1-64. In anotherfactor analysis, using only Banks and Textor data, Philip M. Gregg and Arthur S. Banks, "Dimensions of Political Systems: Factor Analysis of A Cross-Polty Survey,' The American Political Science Review (September, 1965), pp. 602-614, the firstfactor which emerged,labeled "Access," closely parallels the firstfactor reportedhere. There are several importantdifferences in the way the variables from the Banks and Textor survey were employed in the Gregg and Banks factor analysis as compared with this one. The full ranges of the variables reported by Banks and Textor were employed in our factor analysis, while Gregg and Banks collapsed their variables so that each was dichotomous or trichotomous. The second largest factor emergingin the Gregg and Banks analysis (labeled "Differentiationof Political Institutionswithin Former Colonial Dependencies") was based largely on nominal dichotomousvariables excluded from our analysis. These excluded variables include "Date of Independence," "Ex-French Dependency," "Ex-Spanish Dependency," "African Areal Group- ing," "Post-colonial Bureaucracy," and "Underdeveloped Tutelary Political Modernization." In excluding these variables our view was that many cross- national variables are simply not related to the developmentsyndrome per se, althoughthey might conceivably be relevantto the comparison of polities for other purposes. 2"See Donald T. Campbell and Donald W. Fiske, "Convergentand Discrimi- nant Validation by the Multitrait-MultimethodMatrix," PsychologicalBulletin, 56(1959), pp. 81-105 for a discussionof validation of a single trait using several measuresor authors.

This content downloaded from 130.58.65.13 on Mon, 24 Aug 2015 18:34:02 UTC All use subject to JSTOR Terms and Conditions 84 THE JOURNAL OF POLITICS [Vol. 31 herence."26But, althoughlongevity is also a positiveindicator of "adaptability,"this is with referenceto groups and not individual office-holders.In fact, accordingto Huntington,turnover of leaders,especially i'nter-generational changes, indicates greater adaptability.Hence, this factormay reflectmixed qualities of institutionalization (development)such as negativeadaptability and positive autonomyand coherence.If one accepts Huntington'sview of development,then, certainly more discriminating and refinedmeasures of stabilityand change in leadershipthan underliethis factorare needed. It is difficultto see any directrelation between this factor and the developmentcriteria of the Almondor SSRC schemata. Stability,however, may be relevantto some of theirsubcategories. For instance,the capacityof a politicalsystem and its abilityto make effectivepolicy changes (an SSRC dimension)might be associated with stabilityof executivetenure as the French Fourth Republic illustrates.The second factorthen, like the first,is both composed of data fromseveral sources and relevantto at least one theoreticalconstruct. Factor III (variance explained, 7.5 percent) I have called "Domestic Violence." The high loading variables which suggest this title are from YPDP and DON. Revolutions,domestic killed, death by violence,and acts of guerrillawar combine to yield a measure of the stabilityand regulatorycapacity of a government.Although this dimensiondoes not speci'ficallycorre- spond to any of the qualities mentionedby Huntington,clearly the variablesunderlying this factorare indicatorsof development in Huntington'sview. He states while discussingautonomy of politicalorganizations, "a highlydeveloped political system has procedures to minimizeif not eliminatethe role of violence in the system."27He later adds that "coups d'etat and militaryinter- ventionsin politicsare one index of low levels of institutionalization: they occur where political institutionslack autonomyand coherence."28The otherauthors also discussthis qualityin relation to development.Almond and Powell, for example,discuss state buildingwhich involves the capacityto penetrate,regulate behavior, "6SeeHuntington, op. cit.,pp. 405, 407. It is interestingto note,however, that legalityof governmentchange, variable 25, is not at all correlated withthis factor. "7Huntington,op. cit., pp. 401-402. 28lbid.,p. 407.

This content downloaded from 130.58.65.13 on Mon, 24 Aug 2015 18:34:02 UTC All use subject to JSTOR Terms and Conditions 1969] THE STUDYOF POLITICALDEVELOPMENT 85 and overcomerevolutionary threats as a "challenge" of political development.29Similarly, the SSRC authorsdescribe crises such as penetrationand participationwhich relate to the controlof domestic violence,and include in theircriteria for the capacity dimension qualitieswhich could certainlybe measuredby, amongother things, the amountof internalpolitical violence in a system.30Thus while not explicitlysuggested as one of the broad criteriaof development (in the schematawe have examined),domestic violence generally could be related to the qualities mentionedin the development syndromesof all the theory-builders,as an operationalif not analy- ticalindex.3' Since thisfactor is composedof variablesclosely linked to violentunrest, it is an indicatorof the securityprovided by a politicalsystem, and in contrastto factorfive, discussed later, pro- vides a moredistinct barometer of a polity'sdecay. Factors three and fiveare similarto the firsttwo factorsfound in severalother factoranalyses of cross-nationaldata on conflict.32 The fourthfactor in the analysis (variance explained7.1 percent) is called "ParticipantPolitical Socialization." The highloading variableswhich lead to this label are Percentof PopulationVot- ing (YPDP), Political Enculturation(B & T), and Stability of the Party System(B & T). These variablesgive us an indica- tion of the degreeto whichmembers of the politicalsystem have learned to participateand to demand stable and familiarpolitical practices.As such, they reflectboth the range and coherenceof political socialization.Theoretically derived dimensionsto which this factormay be related are Huntington'scoherence continuum and the SSRC equality criteria(which includesparticipation). In addition,other studies have also stressedparticipation as an index of development.33If othervariables relevant to socializationwere includedin a futurefactor analysis it would be possible to con- 29Almondand Powell,op. cit.,p. 35. "0SeePye, Aspects,pp. 64-65. "1Inaddition, Pennock, op. cit.,p. 427,specifically includes security (assessed by domesticviolence) among his criteriaof development. "2Seethe firsttwo factorsreported in Tanter,op. ctt.,p. 50; Rummel, "Dimensionsof Conflict,"op. cit.,p. 69; and Ivo K. and RosalindL. Feira- bend, "AggressiveBehaviors Within Polities, 1948-1962: A Cross-National Study." Journalof ConflictResolution (September, 1966), p. 255. "3SeeDaniel Lerner,The Passingof TraditionaSociety (New York: The Free Press, 1958), pp. 43-75. Lerneruses percentageof populationvoting as the indexof politicalparticipation. The causal modelof developmentsug- gestedby Lernerhas been statisticallyanalyzed by HaywardR. Alker,Jr., in "Causal Inferenceand PoliticalAnalysis," Mathematical Applcations in Poltical Science,II (Dallas: The ArnoldFoundation, 1966), pp. 19-29. Alker

This content downloaded from 130.58.65.13 on Mon, 24 Aug 2015 18:34:02 UTC All use subject to JSTOR Terms and Conditions 86 THE JOURNAL OF POLITICS [Vol. 31 firmwhether socialization and participationrepresent an empirically unique dimensionfor assessing development. Ri'ots,anti-government demonstrations, and generalstrikes load highlyon the fifthfactor. This dimension,consequently, is labeled "Internal Turmoil."All three identifyingvariables are fromthe DON project.Although Huntington is concernedwith destabililzing and disintegrativeeffects such as unrestand turmoilwhich are measuredby the variablessignificantly contributing to this factor, his criteriaof institutionalizationdo not specificallyinclude indices of this order.Moreover, Huntington is primarily1fnterested in phe- nomenalike coups,revolutions and militaryinterventions as indicators of the failureof i'nstitutionalization.34Aside fromits lack of theoreticalsignificance, another possible criticismis the fact that only data fromDON contributesto this factor.This suggeststhat it may be an idiosyncrati'cindex of development. It mustbe noted,however, that this qualityhas been replicated in a factoranalysis by RaymondTanter using similardata over a differenttime period.If confli'ctdata compiledby Tanter,Eck- stein or the Feierabendswere includedin this study the criticism of this factor as a single author product would certainlybe vitiated.35 The sixth factor,like the fifth,is a dimensionreflecting data fromonly one study. It is labeled "GovernmentMilitary Activity" since the two high loaders fromYPDP are indi'cativeof central governmentmilitary expenditure and resources.It is possible to relate this dimensionto what Huntingtondiscusses as complexity, or the otherauthors mean by differentiation,since highgovernment expensesand a large army are oftenassociated with a complex governmentalorganization having a varietyof agenciesand offices. uses thisfactor as an indexof participationfor the analysis.See also Gabriel A. Almondand SidneyVerba, The Civic Culture(Princeton: Princeton Uni- versityPress, 1965), pp. 19-40,501-505. Almond and Verbadevelop the ideal type of a participantpolitical culture which is associatedwith modernand democraticpolities. 34Ibid.,pp. 402, 407. "5SeeRaymond Tanter, "Dimensions of ConflictBehavior Within and Be- tweenNations, 1958-1960," Journal of ConflictResolution (March, 1966), p. 50. Ecksteinand Feierabenddata are reportedin R. J. Rummel,"Dimen- sionsof ConflictBehavior Within Nations, 1946-59," Journal of ConflictRes- olution(March, 1966), pp. 65-73,and Ivo K. and RosalindL. Feierabend, op. cit.,pp. 249-271.

This content downloaded from 130.58.65.13 on Mon, 24 Aug 2015 18:34:02 UTC All use subject to JSTOR Terms and Conditions 1969] THE STUDY OF POLITICAL DEVELOPMENT 87

It is unlikely,however, that any of the theorybuilders would be satisfiedwith the two variablesloading high on this factoras empiricalindicators for these dimensions of development.More relevant and immediateoperational indices of differentiationand com- plexityare conceivable.It remainsfor data collectionsto broaden theirvariable lists to include completemeasures, for example,on the size and specializationof governmentpersonnel.36 Factor seven, althoughhaving moderate loadings (between.40 and .54) on variables 18, 34, and 42, is largely,in our view, a CutrightFactor (based on variables 24 and 43) . There are no signal loadings.Cutright, in creatinghis index of development, reliedlargely on measuringpolitical representativeness.38 Cutright's viewof politicalrepresentativeness shares some characteristics among nations with the amountof revenuecoming from taxation (variable 42). This relationshipmay be due to the fact that in com- munistand less developedstates, revenuefrom taxation tends to be low, while in the "developed" Westernstates, tax revenues, especiallyfrom incomes, tend to be much higher. Theoretically, the firstfactor, based on democraticcharacteristics, should have accountedfor a greatershare of the variancein Cutright'smeas- ureswhich specifically emphasize "democratic" political development. Perhaps the reason these variables load on a separate factoris due to some unique quality in Cutright'sconstruction of indices, the coding of his data, or the choice of time periods for this study.39 The last interestingfactor in the analysisI have labeled "Ter- ritorialIntegration." The two highestloaders were Sectionalism (B & T) and Federal vs. UnitaryStructure (DON). This factor seems to emphasizeterritorial integration, that is, "the progressive "6Somedata collectorshave gathereddata of thisorder. In Russett,et al., op. cit.,p. 71, the percentof populationemployed by the governmentis includedas a variable,but measuresfor only21 countrieswere collated. "7In an earlierfactor analysis based on 45 variables,Cutright's Political RepresentativenessIndex (variable43) emergedas the singlehigh loading on a factor. "8Inhis article"Political Structure," op. cit., Cutrightrenames his earlier index-afteralterations-an index of politicalrepresentativeness. "9DeaneNeubauer has also criticizedthe sensitivityof Cutright'smeasure, notingthat among the 23 nationswhich *he discriminated among in terms of "polyarchy,""many" were tied forthe highestscore possible on Cutright's scale. See Deane E. Neubauer,"Some Conditionsof Democracy,"American PoliticalScience Review, LXI, No. 4 (December,1967), p. 1007.

This content downloaded from 130.58.65.13 on Mon, 24 Aug 2015 18:34:02 UTC All use subject to JSTOR Terms and Conditions 88 THE JOURNALOF POLITICS [Vol. 31 reductionof culturaland regionaltensions and discontinuities"as opposed to political integrationwhich refersto "developinginte- gratedprocess in a participantpolitical community."40While the authorsof all threeof the outlinedschemata recognize integration as a problemor crisis in states undergoingpolitical development, only the SSRC specificallymention integration (as a sub-category of differentiation)and theiruse relates more to political than to territorialintegration. It is plausible,however, to considerHunting- ton's coherencetrait, which is based on consensusof norms,as a measureof integration.And had Huntingtondiscussed "scope" as a criterionof development,this quality mightalso be linkedto the reductionof territorialdiscontinuities. The SSRC committee's criterionof capacity (which includes the scope of "political and governmentalperformance") also can be related to territorialintegration.Territorial integration, nevertheless, is only peripherally tied to these theoreticalschemata. Moreover, it may be associated with negativedevelopment if it resultsin overburdeninga central government,thereby reducing system effectiveness.4'Only two variables load highlyon this factorand its contributionto ex- plainingvariance among the variables is unsurprisinglylow (4.0 percent). These eightfactors represent what shouldbe, as a consequence of the factoranalytic model, independent traits of politicaldevelopment.The fact that severalof these dimensions(factors 5, 6, and 7) have high loading variables drawn froma single data source suggest these dimensionsmay be more or less the product of individualperspectives. In other words,these dimensionsmay be methods factorsrather than underlyingor basic traits.42If we leave out these three factorsfrom our analysis, we have a list of fivedimensions of politicaldevelopment, Power Sharing,Execu- tive Stability,Domestic Violence,Political Socializationand Ter- ritorialIntegration. "0Foran elaborationof theseconcepts see JamesS. Colemanand Carl G. Rosberg,Jr., editors, Political Parties and National Integrationin Tropical Africa(Berkeley: University of CaliforniaPress, 1964), p. 9. "See CharlesE. Lindblom,The Intelligenceof Democracy(New York: Free Press,1965) fora discussionof the potentialnegative effects of central- ization. "See Campbelland Fiske,op. cit. Of course,if moreauthors' measures on thesevariables had been includedin the analysis,these factors might not be subjectto thiscriticism.

This content downloaded from 130.58.65.13 on Mon, 24 Aug 2015 18:34:02 UTC All use subject to JSTOR Terms and Conditions 1969] THE STUDY OF POLITICAL DEVELOPMENT 89

Cd :3 >4 u >4 U u u >4 pq z z Z by z ocn.-C 0 H z z 94 P4 P4 Z 8 PL4 0 W w w w z 9 M 0 u u uc)U

z 0 u z 94 P--4 cfs 1W Cld z 0 h 0 cn 0a ce $4.14 Cd 44-4 44-4 z 0 o 0

44-4 CC,3: 0 ...4 0 Cd C-d >4 o Z >4 0 > .2 u X4 W 0 02 0 0 'C's ce 44-4 ) V U u 0 w eq 0 a.04 Cd N Cd w

u E-4

z 0

0 >4 z z 0 0 w cn z 1-4 P4 2 0 0 u

>4 u

u Z z pq w u "

u u

2 P P4

This content downloaded from 130.58.65.13 on Mon, 24 Aug 2015 18:34:02 UTC All use subject to JSTOR Terms and Conditions 90 THE JOURNALOF POLITICS [Vol. 31

Table 2 summarizesthe relationshipsamong politicaldevelopment dimensions.Blanks indicate complete lack of equivalence among the lists. Where only one aspect of a theoreticaldimension iIs relatable,this aspect is specificallymentioned. Among the rela- tionshipswhich the Table does suggest,many are partial or ten- uous, as I have indicated in the earlier discussion.Moreover, the factor analysis dimensionsare rather specific compared to the broad rubricsemployed in the theorists'schemata. This perhaps reflectstheir closeness to actual data. While these factoranalysis resultsare certainlynot freeof subjectiveinfluences, they nevertheless representstructurally independent dimensi'ons underlying some of the data collated by empiricallyoriented researchers and relatablein some fashionto dimensionsof developmentsuggested in theoreticalliterature.

III

This examinationby factoranalysis of selectedpolitical indi- catorsunderlines problems of findingcommonly acceptable measures or scales of development.Our studyindicates that thereis a common thoughnot universalinterest in a measure of development related to democraticpractices. The firstand dominantfactor foundin the data reflectsthis. Anothercharacteristic suggested by the factoranalysis is a tendencyfor data studies to be idiosyn- cratic, utilizingmeasures which are not convincinglyrelated to measuresfrom other studies but whichwere supposedby the analysis to be related (e.g. democracy).This fact, therefore,calls iIntoquestion the generalizabilityof researchresults based on these idiosyncraticmeasures.43 Comparingthe factoranalysis resultswith the theoreticaldi- mensionsoutlined in Table 2 reveals that the data generallydo not cluster along dimensionsidentifiable as similar or common with those suggestedby theorybuilders. Not even the individual variablessubmitted to the factoranalysis coincidewith suggested measuresof the theorybuilders such as secularizationor differentiation. This tenuousrelationship between the theoretically derived lists '3See,for example, Donald J. McCroneand CharlesF. Cnude,"Toward a CommunicationsTheory of DemocraticPolitical Development:A Causal Model," The AmericanPolitical Science Review (March, 1967), pp. 72-79, whichutilizes Cutright's Political Development Index (1963).

This content downloaded from 130.58.65.13 on Mon, 24 Aug 2015 18:34:02 UTC All use subject to JSTOR Terms and Conditions 1969] THE STUDY OF POLITICAL DEVELOPMENT 91 and the multipledata sourcesfactors may be partiallyexplained by the fact that the theoreticallists largelycontain broad rubricsem- bracing,often with difficulty,empirically distinct phenomena. An- otherreason for the absenceof some of the dimensionsproposed by theory-buildersfrom the factor analytic solution may be due to the deletion fromour study of some variables,such as per- centageemployed by the centralgovernment (YPDP), because of missingdata. Finally the discontinuitiesbetween theory and data based developmentcriteria may be attributedto the concernof most data collectorsfor a more general comparisonof nations ratherthan an explicitmeasurement of "development." These circumstances,nevertheless, do not obviatethe clear need forthorough measures which do capturevariations along dimensions of structuraldifferentiation, role specificity,and secularvalue processes, none of which was obtained in this factor analysis of currentdata. Betterdata for variablesalready measuredare likewise required." In addition to improvedmeasures on these variables, we are going to need historicaltime series and trenddata. Measuresof this genreare requiredfor calculating rates of change and makingprojections about probability,timi'ng and directionof futurepolitical changes. Also we must begin to distinguishbetween system level and lower level measures,such as group or individualcharacteristics. For example,although equality and autonomymay both be related to democracy,inequality among individualsmay or may not be reflectedin low autonomyat the systemlevel (which is where Almondand Powell cast theiranalysis). By recognizingdifferences in the unit being measured some of the ambiguitywe have foundin comparingtheoreti'cal dimensions may be alleviated. In establishinga scale for a particulardevelopment dimension or for any explanatoryvariable which mightbe associated with development,two strategiesmay proveserviceable. First, combining several similarindicators may increasethe overall validityof the finalvariable. For example,combining measures of radio audience, newspaperreadi'ng and telephones,either by factoranalysis or some otherscaling technique,may yield a more comprehensivevariable

'"The originalDON conffictvariables for 1955-57,for example,contain up to one thirdtied scores for zero instancesof conflictfulevents. More dis- criminatingand sensitivetechniques of measurementseem to be required.

This content downloaded from 130.58.65.13 on Mon, 24 Aug 2015 18:34:02 UTC All use subject to JSTOR Terms and Conditions 92 THE JOURNAL OF POLITICS [Vol. 31 for communicationlevels among nations.45For otherpurposes, of course,such variablesmay be more usefullyemployed separately, for example, in order to examine the additive and interesting political effectswhich exposureto radio, readinga newspaper,or attendinga cinemamay have on a traditionalpeasant. The second strategywould be to developintermethod and inter-authorindices. For example,factors one, two and four representdimensions of developmentwhich are based on the proceduresand results of several data collectionsand combine more than one operational technique.As data collectionsimprove the validity,breadth, and span of theirdeposits, methodological advances such as these can be expected.46 Data collectorswill be aided in these effortsif theorybuilders explicitlyspecify operational indicators of the criticaldimen- sions and variables of developmentwhich they sketch.47This task may lead theoreticiansto considermore carefullythe phenomena which compriseboth their values and the mechanisms whichmediate valued outcomes.Complicated theoretical machinery may be less elegant but more useful for empiricalwork and policy makingif operationalterms in plain languageare attached to it. Authorsof theoreticallyconstructed development dimensions may also wish to reconsiderthe meaningfulnessof terms such as differentiation,secularization or rationalizationwhich failed to find an empiricalcounterpart among the factors.Secularization, in particular,may either be highlycorrelated (perhaps forming a single empiricalpattern) with otherdevelopment qualities such as power-sharing(equality, autonomy) or it may be composedof empiricallyhetereogeneous traits which are difficultto assess with "See Alker,op. cit.,pp. 20-23. Two othercommunication variables, tele- vision sets and movie exhibitions,werg discarded in creatingthis combined measurebecause of missingdata and theirfailure to clusterwith the other variables. "The secondedition of the Yale PoliticalData Program'sWorld Hand- book of Politicaland Social Indicators(forthcoming), for example,moves in thesedirections. Some of thesesuggestions for improvingthe gatheringand use of aggregatedata are consideredin moredetail in HaywardR. Alker,Jr., "ResearchPossibilities Using AggregatePolitical and Social Data," in Stein Rokkan,editor, Comparative Research Across Cultures and Nations (Paris: Mouton,1967). "Huntington,op. cit., and Pennock,op. cit., have made attemptsto do just this. Almond,whose theoretical dimensions least correspondto the ones derivedby factoranalysis, is a clear exampleof a theoristwho has so far failedin thistask.

This content downloaded from 130.58.65.13 on Mon, 24 Aug 2015 18:34:02 UTC All use subject to JSTOR Terms and Conditions 1969] THE STUDY OF POLITICAL DEVELOPMENT 93 any precision. In either case some revisionor furtherexplica- ti'onmay be indicatedby the absence of data relatableto these ratheramorphous theoretical dimensions.48 Perhaps the most ambiguousquality of developmentrevealed in this studyis stability.From the discussionof factors2, 3, and 5 it is clear that there are no clean, unambiguousrelationships betweenmeasures of stabilityand theoreticaldimensions related to this quality. Both data collectorsand theorybuilders may be blamed for the fuzzinessin this quality. Huntingtonis the prime theoristinterested in stability,but he attemptsto measuredevelop- ment in termsof traits such as coherence,which explain rather than describe stability.Ambiguity occurs because a few of the operational indices for his four continua, such as coups or civil disordersare commonlyaccepted as signs of instability.This makes it difficultto operationalizedistinctly his particulardevelop- mentcontinua from more generalized notions of stability.Equally open to criticismare data collectorswho tend to count events as equal whichare quite different,for example,peaceful and vio- lent changes of executives(as in factor2) or who classi'fyto- getheras instancesof conflictboth demonstrationsand revolutions. Several analyses have revealed the differentqualities in these conflictevents,49 although one mighthave suspectedthat countries which suppressstrikes and demonstrations(and thus have none) are more likely to experienceserious conflictsuch as revolution. In seekingto reconcilethe varyingconcepts and data related to instabilityit may be usefulto conceptualizethi's quality in terms of manifestand latent characteristics.Kinetic instabilitycould refer to observed disrupti'veactivity, while potential instability could be a measureof the likelihoodand intensityof futurekinetic instability.Whether this suggestionproves useful or not, it is cer- tain that theorybuilders and data collectorsneed to combine their effortsto develop more subtle and satisfactorymeasures, not only forstability, but formost all the developmentdimensions. '8Ofcourse, although two conceptsare empiricallyhighly correlated (or load on the same factor),it may still be profitableto maintaintheir theoretical distinction,provided they are operationallydistinct. Moreover, if oblique factorrotation were employed, it mightbe possibleto detectdifferent, but correlated,factors. "See Tanter,op. cit.,and Rummel,"Dimensions of Conflict,"op. cit.

This content downloaded from 130.58.65.13 on Mon, 24 Aug 2015 18:34:02 UTC All use subject to JSTOR Terms and Conditions 94 THE JOURNALOF POLITICS [Vol. 31

The eight factorsolution to the list of political development variablesbrought together by this study representsneither a fin- ished nor a satisfactorytypology for empirical explorationof political development.The differentinterests that influencedthe various data collectorsto assemble the particularvariables which theychose are compoundedby this study'sown criteriafor select- ing variables. These are just two biases whichoperate to affectthe outcomeof the factoranalysis approach. In addition,weaknesses in the data itselfmay also have had a biasingeffect, for example, due to missingdata, ill-informedjudgments by observers,or injudi- cious selectionof cuttingpoints. Relativityand subjectiveness,therefore,weaken this inductive-empirical approach to establishingdimen- sions of politicaldevelopment. The results,nevertheless, are impor- tantas provisionalstandards for comparing and assessingtheoretical dimensions.They provideinsight about lacuna in both data collec- tions and theoreticalwritings, and underscoreclear weaknessesin presenttheories. The differencesbetween and amongtheory-oriented writers and data collectorswhich I have pointedout are partiallyproduced by theirvarying concepts about the basic characteristicsof development.Such differingvisions and values hinderuseful dialogue and cloud untestedand unrefinedpropositions.50 Thus Lasswell's re- minderthat the study of political developmentdemands explicit statementsof preferenceas to what constitutesdeveloped charac- teristicsis particularlyincisive.5' Increased candidness about value preferenceswill be helpfulin improvingstandards of clarityin the studyof politicaldevelopment. This may facilitatebroader agreement on the dimensionsunderlying political developmentand on the data which are relevant to its measurement.To this end greaterdialogue and idea comparisonsamong theoristsand data collectorsare needed. 50Forexample, the assumptionthat the diffusionof "worldculture" is an inevitableprocess which will accountfor a greatdeal of politicalchange may be correctin the long run,say 25 or 50 years,but is of littleor no value in attemptsto explainor predictshort run changes. 51Lasswell,op. cit.,pp. 290-91.

This content downloaded from 130.58.65.13 on Mon, 24 Aug 2015 18:34:02 UTC All use subject to JSTOR Terms and Conditions