REVIEW OF RURAL AFFAIRS Changing Characteristics of Villages in Tamil Nadu A Vaidyanathan, R Srinivasan Illustrating the imaginative use of the Primary Census he spatial aspects of development have been the focus Abstract for Tamil Nadu from the 1991 to 2011 censuses, of wide-ranging research the world over. In India, the emphasis has been more on tracking regional disparities this paper separates villages that are chronically T in overall development indicators. Studies of these disparities backward from those that are more developed in terms across towns and villages are relatively rare. Village studies, of demographic and economic characteristics. It also though numerous, have been mostly isolated case studies at makes use of the data to describe changes in spatial the disaggregated level or studies using a sample of villages spread over a vast area. Neither of these two strands has used distributions over time. the demographic data and infrastructure data at the village level/urban ward level provided once in a decade since 1961 by the Director General of Census Operations. The Primary Census Abstract (PCA), published by the Directorate of Census Opera- tions, gives demographic details of every revenue village and urban ward.1 Identifi cation of villages by district and taluk is enabled through serially allotted codes. The Village Directory, published by the Census Directorate, has information on the availability of infrastructure facilities and public utilities in each village and a Town Directory has similar details for every town and city.2 These two databases enable not only disaggregated analysis up to the village/urban ward level, but also identifi ca- tion of the outliers—the laggard v illages and those that have been consistently at the higher end of the development ladder. The main aim of this paper is to illustrate the imaginative use of one of the two databases for Tamil Nadu, the PCA, to separate villages that are chronically backward from those that are more developed in terms of demographic and economic characteris- tics and to describe the change in spatial distributions over time. One of the authors earlier carried out a similar exercise for the erstwhile North Arcot District in Tamil Nadu using 1971 and 1981 census data (Vaidyanathan 2013). He has also illus- trated the richness of this database for the state as a whole in another study (Vaidyanathan 2014). In this paper, we use the PCA for all villages from the three censuses between 1991 and 2011. This paper is divided into three parts. Section 1 paints a broad picture of the average Tamil Nadu village and variations in its characteristics over the three censuses. The population size, sex ratio (SR), literacy rate (LR), work participation rate (WPR), the proportions of agriculture and non-agriculture workers and cultivators to total workers (AW/TW and Non- agri/TW) and proportions of agricultural labourers to total ag- ricultural workers (AL/aW) are calculated for each village in every c ensus. The decadal changes in a representative average Tamil Nadu village are discussed in this context. A Vaidyanathan ([email protected]) is a long-standing Section 2 focuses on the comparison between the averages c ontributor to EPW and an Honorary Fellow of the Centre for in the bottom decile (BD) and top decile (TD) of the villages by Development Studies, Thiruvananthapuram; R Srinivasan (seenu242@ the values of each of the above features in 1991 and 2011.3 It gmail.com) teaches econometrics at the University of Madras. highlights differences between the bottom and the top deciles Economic & Political Weekly EPW december 26, 2015 vol l no 52 65 REVIEW OF RURAL AFFAIRS in the mean value of each of the selected socio-economic indi- The proportion of Scheduled Castes (SCs) and Scheduled cators and their association with other characteristics in dif- Tribes (STs) to total population (%SCST) increased by 0.9 per- ferent years, as well as the quantum of such changes from 1991 centage points in the 1990s and this almost doubled to 1.8 per- to 2011. The focus is on describing distinct patterns of variation, centage points in the 2000s. Correspondingly, the dispersion which might trigger further analysis on exploring the underlying of SCs/STs declined over these two decades. The concentration causes for these changes. of the SC and ST population is increasing in rural Tamil Nadu Section 3 describes the spatial distribution of villages in the and their dispersion has been declining over time. bottom and top deciles of each characteristic across districts The WPR measured as the percentage of total workers to total and taluks in 1991 and 2011. Mapping the spatial concentration population (TW/T Pop) increased from 45.1 to 50.3 and then to of villages in the top and bottom deciles provides the basis for 50.7 in the three censuses. Changes in age, gender and educa- assessing whether and to what extent they are geographically tion need to be examined to understand this trend. The inter- concentrated and whether such geographical concentrations village variability (CV) of the WPR has progressively declined. change over time. The share of agricultural workers in the total workforce (AW/TW) in the state increased marginally during the 1990s 1 Overall Trend through 1991, 2001 and 2011 and declined sharply in the 2000s. The percentage of non- The population of rural Tamil Nadu declined by 1.9 million in agricultural workers (NAW/TW), which gives an idea of the extent 2001 and increased slightly by 2.3 million in 2011. The annual of diversifi cation of employment, declined during the 1990s but population growth rate in rural Tamil Nadu is far less than the increased sharply in the next decade. Because of this, in abso- overall population growth rate of the state because of rapid lute terms, between 1991 and 2001, the number of agricultural urbanisation. A progressive decline in the number of villages workers increased while non-agricultural employment did not over this period could be due to tiny villages being abandoned, show any change. In the subsequent decade, the number of contiguous villages being amalgamated, and, more likely, the agricultural workers declined marginally and that of non-agri- larger villages being reclassifi ed as towns because of urbanisa- cultural workers increased by 1.4 million, which was more tion. The growth in total rural population and a reduction in than the increase in the total number of workers. Unlike the the number of villages are refl ected in a progressive increase WPR, the inter-village variability of NAW/TW is much greater in the average size of villages and a decline in the extent of than that of AW/TW, and both seem to have increased. d iversity in size (Table 1). On an average, about one-third of the agricultural workers are Table 1: Salient Socio-economic Characteristics of Tamil Nadu Villages, cultivators and two-thirds are labourers. The share of l abourers 1991–2011 fell in 2001, but in 2011 increased to more than what it had been Village Characteristics 1991 2001 2011 Mean CV (%) Mean CV (%) Mean CV (%) in 1991. We fi nd contrasting trends in the dispersions of AW/TW Total population (crore) 3.68 – 3.49 – 3.72 – and AL/TW—the former is increasing, while the latter is declining. Number of villages 15,902 – 15,400 – 15,049 – Table 2 shows the simple correlation coeffi cients between Population/village 2,315 112.5 2,268 96.4 2,474 94.9 different characteristics across villages in 1991 and 2011. Villages Literacy rate 47.1 27.7 58.2 18.3 65.8 13.4 with a higher SR tend, in both periods, to be smaller in size, Sex ratio 981 8.7 992 9.8 993 12.6 Table 2: Correlation between Characteristics of Villages in 1991 and 2011 SR_06 945 26.7 933 24.6 936 25.8 Variable Total Pop Sex Ratio %SCST Literacy TW/T Pop Non-agri/ AL/AW %SCST 24.3 95.2 25.4 92.4 27.2 87.9 Rate TW (TW/Total pop) 45.1 24.2 50.3 20.6 50.7 18.4 Total pop (Non-agri/TW) 32.5 55.0 29.9 71.5 34.8 63.9 1991 1 (AW/TW) 77.5 23.9 70.1 31.2 65.2 34.6 2011 1 (AL/AW) 57.7 43.6 61.2 37.7 68.4 34.1 Sex ratio (AL/TW) 44.7 47.1 42.9 50.4 44.6 50.5 1991 -0.057# 1 (CUL/TW) 32.7 74.7 27.2 78.1 20.6 95.3 2011 -0.016* 1 The LR substantially increased from 47.1 in 1991 to 58.2 in %SCST 1991 -0.176# -0.024# 1 2001 and further to 65.8 in 2011. This was accompanied by a 2011 -0.163# 0.011 1 progressive reduction in inter-village variation (measured by Literacy rate its coeffi cient of variation or CV). Both refl ect the success of 1991 0.207# 0.141# -0.204# 1 public investment in school education and other non-formal 2011 0.082# 0.080# -0.152# 1 literacy programmes. TW/Total pop 1991 -0.101# -0.039# 0.060# -0.231# 1 The average SR shows a progressive though uneven increase 2011 -0.142# -0.073# 0.016* -0.326# 1 over the period. Child sex ratio (SR_06) declined from 946 to Non-Agri/TW 933 during the 1990s and then marginally increased to 936 in 1991 0.331# 0.017* -0.146# 0.390# -0.284# 1 the subsequent decade.