Regional Disparities in Employment and Income in Rural Andhra Pradesh A.

Amarender Reddy1 Special Project Scientist (Economics) International Crops Research Institute for Semi-Arid Tropics (ICRISAT), Hyderabad a.amarenderreddy@cgiar.org

The study was a part of the updated version of the Ph.D. thesis of the author at Indian Agricultural Research Institute (IARI), New Delhi under the guidance of Prof. Praduman Kumar

Electronic copy available at: http://ssrn.com/abstract=1595195

Disparities in Employment and Income in Rural Andhra Pradesh, India Abstract Andhra Pradesh is one of the largest states in India with more than 60% living on agricultural sector. The paper studies regional difference in employment, unemployment and income in Andhra Pradesh, in the backdrop of demand for separate state of Telangana. The empirical results are based on unit survey data of rural households of Andhra Pradesh based on employment and unemployment surveys conducted by National Sample Survey Organisation (NSSO) in 61st round pertaining to the period 2004-05. Descriptive statistics, multinomial logit models and simple regression models were used to attain the objectives of the study. Work Participation Rate (WPR) was low and unemployment was high in Coastal for both men and women whereas WPR was high and unemployment was low in both Telangana and Rayalaseema. Among males 43 per cent of total workforce engaged in self-employment in agriculture followed by casual labourer in agriculture (27%) and self-employment in industry (11%) in Andhra Pradesh. Among females 40 per cent belongs to self-employed in agriculture and another 40 to percent to casual labourer in agriculture. About 9 per cent were self-employed in industry. Share of agricultural workers was less in Coastal region compared to Telangana and Rayalaseema. As with the development of agricultural sector, there was simultaneous increase in demand for non-agricultural sector goods and services, which increased share of non-agricultural sector workers in total work force in Coastal region. Share of self-employed is higher in both Telangana and Rayalaseema regions compare to Coastal region. Some studies point out that, higher share of self-employed in agriculture in low agricultural productivity regions like Telangana and Rayalaseema is a sign of disguised unemployment. Weekly wage earnings were higher in males, higher educated, upper caste workers compared to their counterparts. Coastal region reported higher wages compared to Telangana and Rayalaseema regions. Introduction Andhra Pradesh is ranked 4th largest in India in terms of area, with its projected population of 8.4 crores in 2011; it is the 5th most populous State in India. The total geographical area of Andhra Pradesh is 275.04 lakh hectares, only 39.8 percent of this is used for agricultural sector. About 72% of population lives in rural areas. Even though about 62.2% of workers dependent on agriculture (out of which 22.5% are cultivators and remaining 9.6% are agricultural labourer) its share in GSDP decreased from about 40% in 1980 to about 17% in 2009 (Andhra Pradesh Statistical Abstract, 2011). Andhra Pradesh agriculture is primarily smallholder agriculture. Approximately 84 percent of the land holdings are of less than or equal to 2 hectares (ha), with a mean holding size of 0.7 ha. Evidence suggests occupational diversification, from lower- to higher-value activities, possesses substantial potential to increase opportunities of income and employment in rural areas (Reddy, 2010). Rural households have a larger endowment of labor in relation to land, and therefore, diversification towards non-farm activities can be an important opportunity to utilize their surplus labor. On the demand side too, there are significant opportunities to diversify towards non-farm activities. Demand for non-farm activities is witnessing exponential growth and the factors such as rising per capita incomes and a fast-growing urban population is responsible. And, in the recent past, the factors underlying demand for nonfarm activities had been quite robust and are unlikely to subside in near future, resulting in a further increase in their demand. In light of the growing disparities in Andhra Pradesh (A.P.) and demand for separate state of Telanagana, the study looks into regional differences in employment, unemployment and income across Coastal, Telangana and Rayalaseema for the period 2004-05 by using NSSO unit data. The specific objectives of the study are as follows. (i)To examine the occupational structure of households in different regions of rural Andhra Pradesh, (ii) To assess the level of unemployment, underemployment and poverty in different categories of households, (iii) To find out various socio-economic factors affecting employment pattern in different regions and among
2

Electronic copy available at: http://ssrn.com/abstract=1595195

households, and (iv)To suggest policy measures to reduce unemployment and underemployment across regions. Data and Methodology The study used unit data based on quinquennial survey of employment and unemployment conducted by the National Sample Survey Organisation, (NSSO) in its 61st round survey covering 432 villages and 5550 households. The state was divided in to four NSSO regions by grouping contiguous districts, which are similar with respect to population density and cropping pattern. They are Coastal consisting of 9 agriculturally prosperous districts, inland northern consisting 10 districts of Telangana region, which is relatively backward, but rich in natural resources, south western region comprising only two districts of Rayalaseema region and inland southern consisting the remaining two districts of Rayalaseema. But for all analytical purposes, the last two regions of Rayalaseema are pooled together to overcome obstacles in interpreting small sample size and also due to historical, socioeconomic and geopolitical reasons. So our final list of regions comprises Coastal (9 districts), Telangana (10 districts) and Rayalaseema (4 districts). For easy interpretation, landholding and monthly per capita expenditure groups were delineated as follows. Monthly per capita income groups are divided into five equal quintiles on the basis of expenditure; For convenience of grouping we call bottom 20% as very poor, next 20% as poor, middle 20 % as medium, next 20% as non-poor lower and top 20 % as non-poor higher. Total land holding was given by (land owned including) + (land leased-in) - (land leased out) + (any land possessed by the household which is neither owned nor leased-in). Landholdings were grouped in to landless (0 ha), marginal cultivated class (0.01 to 0.4 ha), small (0.41 to 1 ha), medium (1.01 to 2 ha), Semi-large cultivated class (2.01 to 4 ha) and large land holding (> 4 ha). Age groups are classified according to standard grouping i.e., children (<15 years), young (15-29), aged (30-59) and old aged (> 59 years). Education level of persons was grouped in to six categories (1) illiterate (2) below primary (3) primary educated (4) middle and secondary educated (5) higher secondary educated (6) graduate and above. Regional disparities in socio-economic conditions Telangana region occupying the largest area of 114.8 lakh hectares accounting for 41.8 per cent of total geographical area of the state, followed by Coastal region with 92.5 lakh hectares (33.7%) (Table 1). Rayalaseema covers the remaining 24.6 per cent of the total geographical area of Andhra Pradesh. Population density is higher in Coastal (309/sq.km) followed by Telangana (227/sq.km) and Rayalaseema (174/sq.km). Per cent of rural population is higher in Rayalaseema followed by Coastal and Telangana regions. For historical reasons, the Telugu-speaking people had been divided into two separate entities, namely, Telangana of Hyderabad state and Andhra area (consisting of Coastal and Rayalaseema) of composite Madras State. For centuries together, the farmer was under feudal (Nizam of Hyderabad) rule and the latter under the colonial (British) rule, which has had an effect on the development of these regions. While the Andhra area, under the colonial rule for centuries and there after ruled by popularly elected governments, has a better socio-economic features at the time of state formations in 1956. On the other hand, 10 districts of Telangana, under the rule of feudal Nizams for centuries, had very low base on socio-economic indicators. Within 11 districts of Andhra there were two distinct regions on the basis of Agro-economic grounds, i.e., Coastal, which was more prosperous agriculturally and called as granary of south India and Rayalaseema, which was most dry region in South India. Average normal rainfall in the state was 925 mm, but in the year 1999-2000 only 817 mm was received. Average rainfall received by different regions varies significantly i.e., highest recorded in Coastal (876) followed by Telangana (782) and the least was in Rayalaseema (772). Coastal and Telangana received below their normal rainfall, but Rayalaseema received above their normal. Of the total area of the state 22.6%

was under forests. Land put to non-agricultural uses accounted for 7.9%. Net sown area was 37.8% of the total land area. Net sown area was highest in Rayalaseema (41.4%) followed by Coastal (40.3%) and the least was in Telangana (32.6%).Inter-regional inequalities in respect of size of operational holdings are found to be significant. In Coastal, the average size of the operational holding declined from 1.60 ha in 1976-77 to 1.15 ha 1998-99, in the Telangana from 2.70 ha to 1.74 ha and in Rayalaseema from 3.00 ha to 2.10 ha in the same period. In 1999-2000, 37 per cent of net sown area was irrigated in Andhra Pradesh. Per cent of net irrigated area was highest in Coastal (56 per cent) followed by Telangana (31 per cent) and the least was in Rayalaseema (21 per cent). Of the total area irrigated more than once, Coastal accounted for 57 per cent followed by Telangana (30 per cent) and Rayalaseema (13 per cent). Number of pump sets were higher in Telangana (as there surface water resources are limited) followed by Rayalaseema than Coastal. Within agricultural sector of the state, food crops predominate. In 1999-2000, they account for 64 per cent of total net sown area of the state. But, per cent of total food crops varies considerably across regions (76 per cent in Coastal, 71 per cent in Telangana and 34 per cent in Rayalaseema). Average yield of most of the crops is higher in Coastal compared to Telangana and Rayalaseema. Overall, the above figures indicate that, Coastal region is prosperous agriculturally followed Telangana and Rayalaseema regions. While in educationally and other human development indicators Coastal region stood first followed by Rayalaseema and Telangana, which will have definite bearing on employment and unemployment situation. Of the total units of factories, about 54 per cent of total units were located in Telangana, followed by Coastal (29 per cent) and Rayalaseema (16 per cent). In terms of capital investment, Coastal ranks first (68.7 per cent) followed by Telangana (24.5 per cent) and Rayalaseema (6.6 per cent). In terms of employment Telangana ranks first (66 per cent), followed by Coastal (26 per cent) and the least was in Rayalaseema (7 per cent). Status of Occupation and Income of Households There was a significant difference in share of different occupational groups among different regions. Share of agricultural labour and self-employed in non-agriculture and otherhouseholds in total households were high in Coastal compared to other two regions while share of self-employed in agriculture households was high in Telangana. While in Rayalaseema, share of non-agricultural labour households was high compared to other two regions (Table 2). The above figures indicate that the proportion of agricultural labour, selfemployed in non-agriculture and service sector households were high in agriculturally developed region (Coastal) compared to agriculturally less developed regions (Telangana and Rayalaseema). It may be due to the fact that, developed agricultural sector in Coastal has simultaneously increased absorption of labour in non-agricultural sector as indicated from higher share of self-employed in non-agriculture households in total households. The share of agricultural labourer households was also high in Coastal, as predominance of landless and marginal landholders coupled with higher mechanization in agriculture induced most of the marginal farmers to work as agricultural labourer. On the other hand, low productive and subsistence agriculture in Telangana and Rayalaseema regions not raised demand for agricultural labour, hence most of the households are working on their own farm as self-employed in agriculture, most of the times it is disguised unemployment.

Table 1: A statistical account of different regions (2004-05) Variables Area (000 km) Per cent to total Population (lakhs) Per cent to total Population density/Km 2 Rural population (lakhs) Urban population (lakhs) Per cent of rural population Per cent of literates to total population (Males) Per cent of literates to total population (Females) Per cent of literates to total population (Total) Total number of towns Number of village with population 1000 and above Number of beds available % of total Per cent of net sown area to total area Per cent of net area irrigated to net sown area Area irrigated more than once (% of total) Per cent of irrigated pumps Per cent of irrigated area to total area Average size of operational holdings (ha) Per cent of distribution of holdings < 1.0 Per cent of distribution of holdings > 10 ha Average yield (Paddy) Average yield (Groundnut) Number of factories (%) Number of factory worker (%) Total capital (%) Capital/labour ratio Large and medium factories Number (%) Large and medium factories Investment (%) Large and medium factories Employment (%) Region Coastal Telangana Rayalaseema 92.8 114.8 67.4 33.7 41.8 24.6 28.7 26.1 11.7 43.1 39.2 17.5 309 227 174 21.4 18.2 9 7.3 7.9 2.7 74.6 69.8 76.8 55.7 36.6 46.2 124 181 38.3 40.3 56.3 56.8 18.8 54.7 1.15 60.5 0.5 2952 1106 29.7 26.1 68.7 6.53 40 42 37 52.9 29.2 41.3 88 84 44.7 33.6 31.4 30.0 59.8 30.9 2.7 50.6 1.5 2381 918 54.2 66.6 24.5 0.88 47.3 53 55.4 58.6 30.7 44.9 52 48 16.8 41.4 20.7 13.0 21.3 14.4 2.1 42.3 2.6 2622 1119 16 7.28 6.6 1.68 12.6 4.8 7.5 A.P. 275 100 66.5 100 242 48.6 17.9 73.1 55.1 32.7 44.1 264 313 100 37.8 37.5 100 100 100 1.56 56.1 1.3 2759 1082 100 100 100 2.37 100 100 100

Source: Statistical Abstracts of Andhra Pradesh (2005) Table 2: Per cent distribution of households by household occupation Household Occupation Self-employed in non-agriculture Self-employed in agriculture Agricultural labourer Non-agricultural labourer Others All Number of sample H/Hs Coastal 18.2 25.4 39.3 4.0 13.1 100 2598 Region Telangana Rayalaseema 15.1 15.4 36.3 28.6 32.8 34.4 5.8 10.3 9.9 11.3 100 100 1976 976 A.P. 16.6 29.9 36.1 5.8 11.6 100 5550

Another dimension of regional differences can be observed by studying structure of income groups by region which was depicted in table 3. Share of very poor H/Hs in total H/Hs was high in Rayalaseema (30 per cent) followed by Telangana (24 per cent) and Coastal region (14 per cent). On the other hand, share of non-poor-higher H/Hs in total H/Hs was high in Coastal (24 per cent) followed by Telangana (17 per cent) and Rayalaseema (15 per cent). As already mentioned these figures clearly show that in agriculturally developed Coastal region poverty was less, whereas in agriculturally less developed regions (Rayalaseema and Telangana) poverty was high. Table 3: Per cent distribution of households by income group Income Group Very poor Poor Medium poor Non-poor lower Non-poor higher All Number of sample H/Hs Coastal 14.4 18.3 19.6 23.8 23.9 100 2598 Region Telangana Rayalaseema 23.8 30.2 20.7 21.3 20.7 18.4 18.3 14.9 16.6 15.1 100 100 1976 976 A.P. 20.5 19.7 19.8 20.3 19.8 100 5550

Activity Status of Workers Distribution of workers by employment activity status was depicted in table 4, which indicates that share of self-employed were highest in Telangana (62 per cent among men and 53 per cent among females) followed by in Rayalaseema (57 per cent in males and 51 per cent in females) and the least was in Coastal (56 per cent in males to 52 per cent in females). Share of casual labourer was highest in Coastal (36 per cent in males and 44 per cent in females) followed by Rayalaseema (35 per cent in males and 45 per cent in females) and the least was in Telangana in both sexes (30 per cent in males to 41 per cent in females). Regular employed were high in Telangana (8.3 per cent in males and 6 per cent in females) and low in Coastal (about 7.9 per cent in males and 3.3 per cent in females). Table 4: Per cent distribution of workers by activity status Activity Status of Employment Male Self-employed Regular employed Casual labourer Total Total number of sample workers Female Self-employed Regular employed Casual labourer Total Total number of sample workers Sector of Employment Share of agricultural workers was more in Telangana (76%) followed by Coastal (70%) and Rayalaseema(66%) among men, among women also share agricultural workers is higher in
6

Coastal 56.8 8.0 35.3 100 2946 52.8 3.5 43.7 100 1608

Region Telangana Rayalaseema 62.1 8.2 29.7 100 2510 53.1 6.0 40.9 100 2030

AP

57.6 58.9 8.0 8.0 34.4 33 100 100 1297 6753 52.7 52.9 3.9 4.7 43.4 42.4 100 100 799 4437

Telangana (85%), but the least was reported in Coastal (76%). As with the development of agricultural sector, there was simultaneous increase in demand for non-agricultural sector goods and services, which increases share of non-agricultural sector workers in total work force. This process was further strengthening by farm mechanization and commercialization that reduces demand for agricultural workers in total work force in Coastal region (Table 5). Among females share of non-agricultural workers is the highest in Coastal region (24.2%) and the least in Telangana region (16.4%). Employment Sector/ Activity Status Share of self-employed in agriculture was high in Telangana (48 %) followed by Rayalaseema (42 %) and the least in case of Coastal (39 %) region. While the share of casual labourer in agriculture was quite high among Coastal (30 %) followed by Telangana (25 %) and closely followed by Rayalaseema (23 %). Share of selfemployed in services was quite high in Coastal (7 %) compared to other two regions. While regular employed in agriculture was high in Telangana due to interlinked credit-labour markets in this region. Both share of self-employed and casual labourer in industry was quite high in Rayalaseema region compared to other two regions. Among women also same trend was observed but slightly higher share of casual labourer in agriculture compared to self-employed in agriculture were reported. Most of the casual labourers were working in agricultural sector and very few were working in industrial and service sectors. Share of regular employed were higher in Telangana (8.3 % in males and 6 % in females) and lower in Coastal (about 7.9 % in males and 3.3 % in females) with higher concentration in service sector in all regions (Table 6). Table 5: Per cent distribution of workers by sector of employment Sector of Employment Male Agriculture Industry Services Total Total number of sample workers Female Agriculture Industry Services Total Total number of sample workers Coastal 69.9 17.3 12.8 100 2946 75.8 14.9 9.3 100 1608 Region Telangana Rayalaseema 75.7 15.2 9.2 100 2510 84.6 12.6 2.8 100 2030 65.8 24.3 9.9 100 1297 79.9 13.3 6.8 100 799 AP 71.3 17.8 10.9 100 6753 80.6 13.6 5.9 100 4437

Table 6: Per cent distribution of workers by sector/activity status of employment Sector/activity status of workers Male Self-employed in agriculture Self-employed in industry Self-employed in services Regular employed in agriculture Regular employed in industry Regular employed in services Casual labourer in agriculture Casual labourer in industry Casual labourer in services Total Total number of sample workers Female Self-employed in agriculture Self-employed in industry Self-employed in services Regular employed in agriculture Regular employed in industry Regular employed in services Casual labourer in agriculture Casual labourer in industry Casual labourer in services Total Total number of sample workers Coastal 39.0 11.0 6.7 0.8 1.9 5.3 30.1 4.4 0.8 100 2946 32.7 13.5 6.6 0.8 0.1 2.6 42.2 1.3 0.1 100. 1608 Region Telangana Rayalaseema 48.6 9.5 4.0 2.2 1.9 4.1 24.9 3.8 1.0 100 2510 45.5 6.0 1.6 0.5 4.5 1.1 38.6 2.2 0.2 100 2030 42.0 11.8 3.8 0.8 2.4 4.8 23.0 10.1 1.3 100 1297 40.6 9.4 2.8 0.5 0.1 3.2 38.8 3.7 0.8 100 799 AP 43.1 10.6 5.1 1.3 2.0 4.8 26.8 5.2 1.0 100 6753 40.0 9.3 3.6 0.6 2.1 2.0 40.0 2.1 0.3 100 4437

Employment Status WPR was low and unemployment was high in Coastal for both men and women whereas WPR was high and unemployment was low in both Telangana and Rayalaseema. This indicates even though Coastal region was well developed agriculturally and industrially, most of the persons suffer from unemployment and underemployment due to high population density and high capital intensive nature of agricultural and non-agricultural sectors in this region (Table 7).

Table 7: Per cent distribution of persons by employment status Employment status Male Fully employed Underemployed Work Participation Rate Unemployed Student Others Total Total number of sample persons Female Fully employed Underemployed Work Participation Rate Unemployed Student Others Total Total number of sample persons Coastal 82.0 0.4 82.4 3.8 4.9 8.9 100 3574 43.3 2.1 45.4 2.7 2.0 49.9 100 3544 Region Telangana Rayalaseema 85.5 0.3 85.8 1.6 5.2 7.6 100 2929 66.8 0.6 67.4 1.3 2.5 28.8 100 3012 84.9 0.6 85.5 2.3 3.1 9.1 100 1517 51.3 1.0 52.3 1.3 2.1 44.3 100 1528 AP 83.8 0.4 84.2 2.7 4.7 8.4 100 8020 53.6 1.3 54.9 1.9 2.2 41.0 100 8084

Results of multinomial regression analysis of household occupation The multinomial regression analysis is widely used in employment choice models. It determines which variables influence the choice of occupation (type of sector/status of employment) of workers. The multinomial model requires that a particular occupational category be designated as the numerate against which all results should be compared. The study considered self-employed in agriculture as the comparison group as these households were predominant in rural India. Choosing this group for comparison purposes allows us to ask whether, other occupational groups can be regarded as systematically different in any way. This implies that parameter estimates for the categories which were included should be interpreted not as correlates of employment in a given occupational category, but as indicators of the strength of association of a particular explanatory variable with respective occupational group relative to the same explanatory variable with self-employed in agriculture households. The parameter estimates along with t-statistics of variables included has been given in table 8. The parameter estimate for household income is significantly negative for agricultural labourer and significantly positive for other occupations, showing high-income households less likely to be engaged as agricultural labourer and more likely to be engaged in other occupations such as government service etc., compared to selfemployed in agriculture. All occupational groups were having significant negative association with size of landholding compared to self-employed in agriculture, which indicate having land necessarily increase self-employment options in agriculture compared to all other occupations. Scheduled tribes were less likely to work as self-employed in nonagriculture compared to self-employed in agriculture, while scheduled castes were more likely to be engaged as agricultural labour, non-agricultural labourer and other occupations compared to self-employed in agriculture. Other backward classes were more likely to be engaged as self-employed in non-agriculture, agricultural labourer and non-agricultural labourer compared to upper castes. Households with large family size were less likely to be in other occupations and more likely to be engaged as self-employed in non-agriculture, agricultural labour occupations compared to self-employed in agriculture. Regression
9

model indicates education of head of household plays crucial role in determining household principal occupation. Below primary up to secondary education level having positive and significant association with both self-employed in non-agriculture and other occupations, but negative association with agricultural labour occupation. Higher secondary did have significant positive association with self-employed in non-agriculture and other occupations, but insignificant in case of all other occupations. In same lines, graduate and above do have strong negative association with agricultural labourer but significant positive association with self-employed in non-agriculture and other occupations. All educational statuses i.e., below primary up to graduate, labour hhs were having negative coefficient and the magnitude of the coefficient and significance becoming stronger as educational status increases. Other occupations were having strong significant positive association only at above middle education of the head of household. A little education of head of household will help the household chances to be in self-employed in non-agriculture compared with self-employed in agriculture households. Age of head of household was having significant positive association with other occupations compared to self-employed in agriculture. Households belongs to Coastal region are having more probability to earn major share of income from agricultural labour, non-agricultural labour and other occupations, while less likely to earn from self-employed in non-agriculture compared to self-employed in agriculture. While households belongs to Telangana region are less likely to earn livelihood from self-employed in non-agriculture and agricultural labourer, while more likely to earn from non-agricultural labourer and other-occupations compared to households in Rayalaseema. Cox pseudo R2 has been explained about 59% of variation in the categorical dependent variable. So the model explains considerable variation in choice of household major source of income from the household and regional factors. In short household characteristics play an important role in deciding household major source of income. Household choice of major occupation to considerable extent determined by a host of household characteristics such as, landholding, family size, educational status of household head, age of head of household. The physical capital and human capital indicators i.e., size of land holding, education of head of household and age of head of household does have significant negative association with inferior (agricultural labour) occupation, levels of education of head of household make significant difference between agricultural labour households and non-agricultural labour households as in non-agricultural labour hhs primary education and medium education turned out to be insignificant. Secondary and above education of head of household and number of literate in household does have positive association only with other households. On the whole higher education, asset holding such as size of land holding were having significant positive association with high remunerative occupations (other occupations). However social and traditional variables such as scheduled caste and backward caste households were having significant positive association with agricultural labour in agriculture. At last household size and composition of the household also played a very important role in choice of household occupation, for example family size was having significant positive association self-employed in agriculture and agricultural labour occupations. It is interesting to note that after controlling for all the above household characteristics, influence of regional dummies are significant, hence will act as prime determining factor in choice of occupation of households.

10

Table 8: Parameter estimates of household principal occupation (against self-employed in agriculture as the base category) choice model Explanatory Variables Self-employed in Agricultural non-agriculture labour Coefficient t Coefficient t Others Non-agricultural Labour households Coefficient t Coefficient t

Intercept -2.09** -4.4 0.18 0.4 -1.90 -3.2 -2.78 -5.2 Log (land) -2.53** -28.0 -2.25** -27.3 -2.80** -23.8 -1.94** -20.6 Family size 0.18** 4.1 0.11** 2.9 -0.07 -1.0 -0.91** -11.7 Educational level dummies Below primary 0.41* 2.5 -0.56** -3.8 0.29 1.3 0.33 1.6 Primary 0.68** 3.7 -0.74** -4.1 0.23 0.9 0.75** 3.3 Middle & secondary 0.65** 3.7 -1.07** -6.2 -0.01 -0.2 1.13** 5.9 Higher secondary 0.78* 2.2 -0.99* -2.5 -0.86 -1.1 2.31** 7.4 Graduate and above 1.45** 4.1 -1.87** -2.9 -0.93 -0.9 3.25** 10.6 Age (in years) 0.03 0.3 -0.32** -3.5 -0.17 -1.2 0.56** 4.7 Social group dummies Schedule tribe -0.78** -3.1 0.30 1.7 -0.88 -1.8 -0.03 -0.1 Schedule caste 0.27 1.3 1.92** 11.6 1.86* 7.3 1.09** 5.1 Backward caste 0.35** 2.6 .50** 4.0 0.90* 4.3 0.23 1.4 Sex of head (male=0; female=1) -0.42 -1.9 -0.02 -0.1 -0.25 -0.9 0.43* 2.0 Regional dummies Coastal -0.71** -4.4 0.78** 5.7 2.11* 10.2 0.67* 3.7 Telangana -0.45** -2.8 -.73** -5.3 1.18* 5.9 0.37* 1.9 Number of sample observations 804 1796 273 534 Model Fitting Information -2 LLR 7356.9 Chi-Square 4224.5 Df 56 Sig. 0 Pseudo R-Square 0.609 Total sample households 4948 ** and * indicate significant at 1per cent and 5 per cent level respectively Note: Number of sample observations in self-employed in agriculture was 1547. Outlier observations have been removed from the regression by using filtering method. Factors influence sector of employment of workers In the sectoral analysis, three sectors namely agriculture, industry and services were taken for analysis. The study adopted agricultural sector as comparison groups as any shift from this sector to non-agricultural sector is positive for workers and to the economy in rural India. As expected probability of working in industry and services was less compared to working in agricultural sector for large landholders. As in activity status here also, we introduced five education level dummies to represent the persons educational level which includes literate but below primary, primary, middle and secondary, higher secondary and graduate & above. Below primary educated up to graduate were having significant positive association with both industry and service workers compared to agricultural workers. The significance and magnitude was increased as educational level increases. The coefficients were having high magnitude and significance among higher educated for service workers compared to industrial workers (Table 9). It indicates that probability of service workers was more among higher educated workers compared to even
11

industrial workers. Age as a proxy for experience was having significant positive association with service sector employment compared to agricultural workers, but in case of industrial workers, it turned to be insignificant. Scheduled tribes and scheduled caste workers were less likely to be employed in industrial and service sectors compared to agricultural sector, however other backward caste workers did not show any significant association. In other words, social discrimination was still there in rural areas especially for scheduled tribe and caste workers, it was not the case with other backward caste workers after controlling for educational level and size of land holdings. Probability of female workers was less in service sector compared to agricultural sector employment. While workers in costal region probability to be engaged in industry is higher, while probability of workers in Telangana region to be engaged in both services and industry is less than Rayalaseema region. Hence we can conclude that, Being in Telangana , the probability to be engaged in non-agricultural activities is less than Coastal and Rayalaseema regions after controlling for household characteristics. Table: 9: Parameter estimates of sector of employment against agricultural workers as the base category among workers of above 15 years of age Explanatory variable Sector of employment Service Industry Coefficient t-statistics Coefficient t-statistics -2.83 -8.6 -0.76 -3.0 -0.42* -23.6 -0.54* -34.6

Intercept Ln (land) Educational level dummies Below primary 0.33 2.0 0.84** 7.8 Primary 0.72 4.8 1.22** 11.4 Middle &secondary 1.74* 14.7 1.54** 17.3 Higher secondary 3.02** 16.4 1.84* 9.4 Graduate and above 3.73** 20.6 1.62* 7.3 Age group 0.42 5.9 0.23 3.4 Social group dummies Schedule tribe -0.82** -3.3 -0.92** -6.5 Schedule caste -0.22 -1.6 -0.91** -9.3 Back ward caste 0.72 6.6 -0.14 -1.7 Sex of person (female =1; male=0) -0.33 -3.7 -0.12 -1.7 Regional dummies Coastal 0.11 0.5 0.64* 6.8 Telangana -0.34* -2.7 -0.34* -3.6 Sample size 896 1654 8222 Model Fitting Information -2 LLR ratio 6535.0 Chisquire 2965.4 Df 26 Sig. 0 Psuedo R squire 0.31 Total no. of sample workers 10772 ** significant at 1 per cent and * indicates significant at 5 per cent level Note: Number of observations in base category (agricultural workers) was 8222. Outlier observations have been removed from the regression by using filtering method.

12

The selected socio-economic variables viz., household characteristics i.e., household, assets (land cultivated), social group, and personal characteristics such as educational level, age (experience), sex and regional dummies together contribute about 30 per cent of variation in sector of employment. Factors influencing weekly wages Weekly wages was Rs 250 for males and Rs 134 for females in 2004-05 at constant prices of 1999-2000 in rural Andhra Pradesh. Weekly wage was high in higher-educated as compared to lower-educated workers. Workers of scheduled tribe and scheduled caste were reported low wage in both years compared to upper castes. Workers belong to Coastal region reported higher wages followed by Rayalaseema and the least was reported in Telangana region (Table 10). Average wages of non-Hindus were high compared to Hindus as non-Hindus were mostly engaged in non-agricultural sector. Child workers reported low wages i.e., only Rs 113.7 per week.

Table 10: Wage earnings per week among different socio-economic groups in rural Andhra Pradesh in 1999-2000 and 2004-05 at constant 1999-2000 prices Weekly wage income (Rs) Weekly wage income (Rs) Variable 1999-2000 2003-04 Change (%) Variable 1999-2000 2003-04 Change (%) Education level Age group Illiterate 121.8 152.9 25.5 Child 88.8 113.7 28 Below primary 149.2 179.8 20.5 Young 141.2 179.0 26.8 Primary 24.6 Aged 16.9 174.7 217.7 189.3 221.3 Middle &secondary 0.9 Old 240.0 242.2 109.8 126.3 15 Higher secondary 538.2 603.9 12.2 Region Graduate and above 685.2 737.3 7.6 Coastal 180.3 224.5 24.5 Telangana 154.0 181.6 17.9 Social group Scheduled tribe 15.4 Rayalaseema 18.1 140.5 162.1 162.3 191.7 Scheduled caste 143.4 195.9 36.6 Income group Upper caste 182.0 214.6 17.9 Very poor 113.5 135.4 19.3 Poor 125.5 150.1 19.6 Religion Hindu 22.6 Medium 164.9 202.2 148.5 200.5 35 Non-Hindu 184.0 193.8 5.3 Non-poor lower 163.0 181.6 11.4 Total 166.5 201.0 20.7 Non-poor higher 339.2 361.2 6.5

To estimate the differential impact of human capital and other physical capital influence on earnings of casual labourer in agriculture the study run a regression equation with weekly earnings of casual labourer in agriculture as dependent variable (Table 11). Among all educational dummies only middle educational level was having significant positive association with wage rates. This was also in confirmation with earlier findings that wage s of casual labourer were not associated with level of education. Age as proxy for experience is also having positive association with weekly earnings. Male workers earned higher wages than female counter parts. Workers belong to both Coastal and Telangana regions are earned higher wages than Rayalaseema region workers after controlling for household characteristics.

13

Table 11: Factors influencing weekly wage earnings of casual labourer in agricultural workers (weekly wage income as dependent variable) Independent variables (Constant) Ln(land) Educational level dummies Below primary Primary Middle secondary Higher secondary Graduate and above Age Group Sex (male=0; female=1) Regional dummies Coastal Telangana Adjusted R square Number of sample workers Casual labourer in agriculture Coefficient t-statistics Standardized coefficient 5.454** 22.9 -0.020** -3.8 -0.060 -0.023 0.087 0.039 0.024 0.309 0.054** -0.422** 0.0 1.5 1.8 0.3 1.7 3.6 -23.6 0.000 0.025 0.029 0.005 0.026 0.057 -0.376

0.190** 0.156** 0.21 3481

6.6 9.0

0.147 0.200

* Indicates significant at 5 per cent level and ** indicates significant at 1 per cent level. Outlier observations have been removed from the regression by using filtering method. Conclusion The paper studies regional difference in employment, unemployment and income in Andhra Pradesh, in the backdrop of demand for separate state of Telangana. Further the study also assessed what are the factors, which influence choice between agriculture and non-agricultural occupations/sectors and also between self-employed, regular employed and casual labourer? The empirical results are based on unit survey data of rural households of Andhra Pradesh based on employment and unemployment surveys conducted by National Sample Survey Organisation (NSSO) in 61st round pertaining to the period 200405. Descriptive statistics, multinomial logit models and simple regression models were used to attain the objectives of the study. Work Participation Rate (WPR) was low and unemployment was high in Coastal for both men and women whereas WPR was high and unemployment was low in both Telangana and Rayalaseema. This indicates even though Coastal region was well developed, most of the persons suffer from unemployment and underemployment due to high capital intensive nature of agricultural sector in this region. Share of agricultural workers was less in Coastal region compared to Telangana and Rayalaseema. As with the development of agricultural sector, there was simultaneous increase in demand for non-agricultural sector goods and services, which increased share of non-agricultural sector workers in total work force in Coastal region. Share of selfemployed is higher in both Telangana and Rayalaseema regions compare to Coastal region. Some studies point out that, higher share of self-employed in agriculture in low agricultural productivity regions like Telangana and Rayalaseema is a sign of disguised unemployment. Average weekly earnings in rural Andhra Pradesh were Rs. 250/- for males and Rs. 134/for females in 2004-05 at constant prices of 1999-2000. Weekly wage earnings were higher in males, higher educated, upper caste workers compared to their counterparts. Coastal
14

region reported higher wages compared to Telangana and Rayalaseema regions. Regular employed in services recorded the highest wages while casual labourer in agriculture reported the lowest weekly earnings. Overall there was a 21 per cent increase in real wage earnings during 1999-2000 to 2004-05 with 19 per cent increase in males and 25 per cent increase in females. This shows that even though there were high gender differences in weekly wage earnings, there is a tendency in decrease in gender differences in wage income. References 1. Andhra Pradesh State Statistical Abstracts (2011) Directorate of Economics and Statistics, Government of Andhra Pradesh, India various issues 2. NSSO (National Sample Survey Organization) (2007). Household Consumer Expenditure Survey Data, 61st National Sample Survey Round 2004-05. CSO, India. 3. Reddy AA and Praduman Kumar (2006) Occupational Diversification of Households in Rural Andhra Pradesh, Journal of Indian School of Political Economy, Jan-June, 2006, Pp. 77-91 4. Reddy, AA (2010) Disparities in Agricultural Productivity Growth in Andhra Pradesh, Indian Economic Journal, Vol. 58, No. 1, April-June 2010. Pp. 134-154 5. Reddy AA (2010) Regional Disparities in Food Habits and Nutritional Intake in Andhra Pradesh, India, Regional and Sectoral Economic Studies, Vol. 10-2 (2010) 125-134

15