Abstract
This article examines agricultural growth and rural inequality in West Bengal during the post-reform period. Two qualitatively different phases of growth and inequality are identified: from 1993–1994 to 2004–2005, when agricultural growth in West Bengal was moderately higher compared to all of India, but the state witnessed increasing rural inequality; and from 2004–2005 to 2009–2010, when agrarian growth decelerated and rural inequality declined. To unravel the underlying changes in rural inequality, this study uses a class framework and suggests certain plausible explanations. It shows that the benefits of good agricultural growth in the first period were reaped by the middle and small farmers; moreover, it is possible that a shift in population occurred from these classes toward the absentee landlord class who are also self-employed in the nonfarm sector. During the next period, the deceleration of agrarian growth was associated with a sharp decline in the proportion of population of the farming classes and increase in the population share of the poorer classes, that is, agricultural labor and nonagricultural worker. Hence, declining inequality does not necessarily signify development, since the poorer classes evidently did not experience any real improvement; rather, a process of proletarization has begun in rural Bengal.
Introduction
The state of West Bengal was ruled by the Left Front government for 34 years which can broadly be categorized as the pre-reform and post-reform periods. There is a plethora of literature on the West Bengal economy which focuses on the unprecedented agricultural growth mainly during the pre-reform period, that is, the decade of the 1980s. For example, Saha and Swaminathan (1994) and Rawal and Swaminathan (1998) have studied the turnaround in agricultural growth during the 1980s. Bardhan and Mookherjee (2003, 2006) conducted field-based analysis of the farm economy and evaluated the impact of policies such as land reform and input subsidies. Some of the literature also focused on the distributional aspects; Chatterjee (1998) showed that the 1980s witnessed declining overall inequality. The pattern of inequality within the farm sector was such that a marginal farmer gained more than a rich farmer owing to policies such as land reform, distribution of subsidized input, and panchayat reform (Chattopadhyay, 2005). It is evident that much of the scholarship focused on this pre-reform period to establish that the policies pursued by the Left Front in the decade of the 1980s made a difference in the agriculture sector in terms of high agricultural growth and declining rural inequality. However, these similar aspects, that is, agricultural growth and the corresponding pattern of rural inequality, were not adequately studied during the second phase of Left Front rule, in the post-reform period. This motivates the present article to investigate agricultural growth and its impact on the income distribution in rural West Bengal during this second phase of Left Front rule.
A careful analysis of the rural economy during the post-reform period is required as the agricultural sector experienced massive changes in terms of various policy measures. During this period, subsidies for fertilizers and high-yielding variety seeds, as well as for credit to small and marginal farmers were withdrawn, unlike in the decade of the 1980s (Bhattacharya & Bhattacharya, 2007). These changes are likely to have impacted the rural economy which was dominated by small and marginal farmers in the post-land reform scenario.
The objectives of this article are to examine the agricultural growth between these two periods, 1993–1994 to 2004–2005 and 2004–2005 to 2009–2010, and how rural inequality changed. This study provides a class-based inequality analysis by decomposing inequality into inter and intra-class components in order to understand which component led to the changes in inequality. Any policy or set of policies are likely to generate benefit for certain classes. Policies such as land redistribution, operation Barga, and panchayat reform were beneficial for the small and marginal farmers. Similarly, it can be expected that certain classes will reap the benefits in the new policy regime. Thus, the class analysis of inequality would be an important and relevant measure to find out which class or classes benefitted in the era of economic reform and, in turn, shaped the changes in overall rural inequality.
The rest of the article is organized as follows. The following two sections present the structure of the rural economy and the employment trends and relate agricultural growth to rural inequality. Then, subsequent sections define the class framework and discuss the structure of the economy and population share, as well as the changes in the latter across classes. A final section analyzes the changes in inequality by looking at the relative position of classes, before providing a conclusion.
Structure of the Rural Economy
This section focuses on the structure of the rural economy in terms of the changing composition of rural employment. The 2011 census showed that 68.13 percent of the population lived in rural West Bengal, and this state occupied the third position in terms of contributing the highest share to the rural population, preceded by Uttar Pradesh and Bihar. Thus, West Bengal had a high rural base as compared to most other states.
Table 1 shows that the share of employment in the agriculture sector had declined for the female category, whereas for the male category, it remained fairly constant. However, within the manufacturing sector, the proportion of male employment had declined, whereas the proportion of female employment had increased marginally. However, the decline in male employment was much higher than the increase in female employment. The decline was offset by the rise in the share of employment in the service sector for both the male and female categories. The female category especially witnessed a drastic rise in the proportion of employment in this sector. This means that the rural service sector was providing employment opportunities during this period. But the later period exhibits a different trend. The proportion of employment in agriculture declined for both the male and female categories. On the other hand, the proportion of employment in the manufacturing sector increased drastically for both the male and female categories. Similarly, the proportion of employment had also increased in the service sector for both the male and female categories between 2004–2005 and 2009–2010 (Table 1).
Composition of Rural Employment (%)
Even though the proportion of employment was rising for the manufacturing and service sectors, the employment share of agriculture was still high: 59 percent of males and 42 percent of females were employed in the agricultural sector in the rural areas during 2009–2010. This indicates that agriculture remained one of the most significant sources of employment in rural Bengal.
The dependence on agriculture in terms of employment did not decrease significantly, even though the contribution of this sector to the overall Net State Domestic Product (NSDP) had decreased not only for the state, but also for the Indian economy. During the post-reform period, structural changes were observed where the tertiary sector contributed the lion share in the overall NSDP for the state as well as GDP for the country but failed to be a major source in terms of providing employment. Thus, the agricultural sector continued to remain a major provider of employment. This motivates the study of agricultural growth and patterns of rural inequality in the post-reform period, which is explored in the following section.
Agricultural Growth and Rural Inequality
This section presents the results of examining agricultural growth and the trend of rural inequality for West Bengal and all of India. The entire period is divided into two sub-periods: from 1993–1994 to 2004–2005 and from 2004–2005 to 2009–2010. To compute the growth rates of agriculture and allied activities, this study has used the NSDP data which are primarily published by the Bureau of Applied Economics and Statistics. The NSDP series for agriculture and allied activities are available are different base prices. For example, the NSDP for agriculture and allied activities between 1993–1994 and 2009–2010 are given in terms of the 1993–1994, 1999–2000, and 2004–2005 base prices. Thus, to make the entire series of NSDP compatible, the NSDP of different base prices were converted into a single base price by applying the simple splicing technique. In the present study, the entire NSDP series have been converted into 2004–2005 prices. Since the number of data points is less, this study has adopted the dummy variable-based regression analysis to calculate the agricultural growth rate by phases. The following equation has been used to calculate the compound annual growth rates of agriculture and allied activities:
where Yt is the value of output from agriculture and allied activities, D1 is the dummy variable, ut is the error term, a1 is the intercept term, and a2, a3, and a4 are the slope coefficients. In the first period, D1 takes the value 0.
Table 2 presents the growth rates for agriculture for West Bengal and India between 1993–1994 and 2009–2010. The growth rates show that West Bengal and India exhibit different trends. During the first period, West Bengal grew at a rate of 2.78 percent and India at 1.93 percent. However, in the later period, Indian agriculture witnessed an improvement; it grew by 3.42 percent, while West Bengal observed a slight deceleration, at a rate of 2.31 percent. This shows that West Bengal’s performance was better than all of India until 2004–2005. Bhattacharya and Bhattacharya (2007) and Khasnobish (2008) confirmed this declining agricultural growth in the post-reform period. The pattern of agricultural growth suggests that the rice economy was moving more and more toward non food grains. Even the yield growth of rice started tapering off in the post-reform period (Khasnobish, 2008).
Given this pattern of agricultural growth, the next focus here is to understand the behavior of rural inequality. Most of the activities in the rural sector can be categorized into farm and nonfarm activities. Thus, rural inequality will be studied in terms of the farm and nonfarm axes. 1 The most popular measure of inequality, the Gini index, has been used here. The data have been taken from the Household Consumer Expenditure Survey by the National Sample Survey Organization (NSSO) for the 50th (1993–1994), 61st (2004–2005), and 66th rounds (2009–2010). 2
The following observations can be made from Table 2. The rural Gini coefficient increased from 0.254 to 0.274 in the period of 1993/1994–2004/2005 and then decreased to 0.245 in 2009–2010. This implies that right after the economic reforms, rural Bengal became relatively unequal, and the inequality index further decreased in 2009–2010. A similar pattern was observed all over India. However, a further disaggregation of rural inequality into the farm and rural nonfarm axis exhibits different trends for West Bengal and India. In the case of West Bengal, the Gini index of the rural farm sector went down from 0.248 to 0.240 during 1993/1994–2004/2005, and in the later period, i.e. between 2004–2005 and 2009–2010, it decreased further to 0.235. This trend suggests that the agricultural sector was witnessing declining inequality throughout the period. However, the Gini index for all India agriculture had registered an increase in the first period and then declined. On the other hand, the Gini index of the rural nonfarm sector had increased from 0.258 to 0.310 in 1993/1994–2004/2005 but decreased to 0.253 in 2009–2010. The same trend was followed by the nonfarm sector all over India.
Agrarian Growth and Rural Inequality in West Bengal, 1993–1994 to 2009–2010
Thus, the general pattern in West Bengal shows that inequality in the nonfarm sector is always higher than that in the farm sector, and the changes in inequality in the nonfarm sector seem to play an important role in shaping overall rural inequality. However, although West Bengal broadly follows a similar trend to the rest of India, the main differences come from the farm sector. The agricultural sector of West Bengal seems to have witnessed a decline in its Gini coefficient throughout the period, whereas India witnessed an increase and then a decline in its Gini coefficient for this sector.
As discussed earlier, agricultural growth in West Bengal has shown an opposite trend as compared to all of India. During the first period, when West Bengal’s agriculture was growing at a faster rate than the rest of India, the state observed increasing rural inequality, along with declining farm and increasing nonfarm inequality. India’s rural inequality also exhibited the same trend, but both the farm and nonfarm sectors witnessed increasing inequality. On the other hand, in the later period, West Bengal witnessed relatively low agricultural growth and declining inequality in both the farm and nonfarm sectors, whereas the rest of India witnessed a relatively higher agricultural growth with a reduction in rural inequality in both the farm and nonfarm sectors. Now, the question that arises is: How does one make sense of the declining farm and increasing nonfarm inequality, given the pattern of relatively higher agrarian growth during the first period. Similarly, the reduction in inequality in both the farm and nonfarm sectors during the later period when agricultural growth declined further also requires further explanation. These are addressed by introducing a class framework, which is discussed in the following section.
Class Framework and Decomposition of Inequality
To unravel the underlying process of change in rural inequality, this study has adopted the class framework proposed by Vakulabharanam (2010). It was used to analyze the increasing inequality during the high growth phase of the Indian economy in the neoliberal era. While studies focusing on inequality have shown that income inequality was exacerbated after the introduction of neoliberal policies (Dev & Ravi, 2007; Himanshu, 2007), Vakulabharanam (2010) illuminated the dynamics between classes, along with their gradual movement toward ‘relative improvement’ and ‘relative pauperization’ during this high growth phase, and suggested that the distance between professional, rural rentier classes, and the unskilled, agricultural worker was on the rise. Thus, this study indicates that the high growth process has been beneficial only for certain classes. The same class framework has been deployed by Sreeraj and Vakulabharanam (2015) to understand the high growth rate and increasing inequality trend in the state of Kerala. By examining the NSSO data, this section attempts to illustrate how the classes are constructed.
The rural class framework 3 is constructed with the available information in the unit-level data provided by the NSSO. The rural sector comprises two sectors: farm and nonfarm. By using the National Classification of Occupation (NCO) categorization for households that identify agriculture as their primary occupation, they are categorized into five different agrarian classes. They are the rich farmer, middle farmer, small farmer, marginal farmer/tenant, and agricultural worker. The landholding of a rich farmer class is above 10 acres, those who own between 5 and 10 acres are classified as middle farmers, those with 2–5 acres are small farmers, while those with less than 2 acres are the marginal farmers/tenants. 4 Those who do not own any land but identify themselves as workers and agriculture as their primary occupation are denoted as agricultural workers. There is an implicit assumption of labor power buying and selling for these agrarian classes. Rich and middle farmers are net buyers of labor power; small farmers are balanced in net selling and buying of labor power, while marginal farmers/tenants are net sellers of labor power (Vakulabharanam, 2010).
Although land ownership broadly reflects an idea of concentration of means of production, it does not fully capture the relations of production in terms of organizing production within the rural farm sector. There might be various types of holdings; for example, if a household owns a large size of land but the household size is small, then the household needs to be dependent on outside labor to cultivate its land. On the other hand, if the household size is also large, then farming might be done by the household members or partly by members and partly by hired laborers. Alternatively, it can lease out (some) land to others. Apart from that, if the household size is also large relative to its land ownership, then the household might have to work for others or opt for leased-in land in order to sustain a livelihood. Given these complexities with the land size, another important variable was introduced to categorize the classes in the farm sector, labor exploitation (Patnaik, 1976). However, Patnaik (1987) applied these criteria to a field-based analysis. The present class framework is based on a large survey which mainly focuses on various aspects of household consumption expenditure and does not capture in detail the complexities involved in labor usage by households. Thus, this present class framework has limitations in capturing those above-mentioned nuances related to land ownership and labor usage. Nonetheless, one can broadly assume that the rich and middle farmers hire net labor power, whereas the small farmers are balanced in net labor power selling and buying, and marginal farmers/tenants are net seller of labor power.
On the other hand, the remaining population in the rural areas whose primary occupation is not agriculture is categorized into five different nonfarm classes. Among them, using the occupational description and code (provided by the NCO), the rural professional class and the nonagricultural worker class has been identified. Of the remaining, based on the survey questions, the self-employed category has also been identified. Within the self-employed category, those who own more than half an acre of land and work in the nonfarm sector are categorized as absentee landlords, while those who own below half an acre are defined as belonging to the nonagricultural self-employed class. Of the remaining categories, those who own more than half an acre and fall under the other working groups are considered as ‘absentee landlord plus others’. The remaining population falls under unclassified categories. Thus, as a whole, we get the following nonagricultural classes: absentee landlord + nonagricultural self-employed, nonagricultural self-employed, nonagricultural worker, absentee landlord + others, rural professional, and rural unclassified.
This class framework is taken into consideration to decompose the overall inequality into inter and intra-class components by following the decomposition technique proposed by Yitzhaki (1994). This technique helps to find the amount of inequality that can be explained by the between component and within component. Additionally, this technique provides insights into overlapping, which is whether the range of consumption of any particular class or group overlaps with the range of consumption of the overall distribution. The overlapping index helps to understand the process of stratification in a society. 5
Relative Position and Proportion of Population of Various Classes
By using the class framework described above, this section now presents the results of the relative position of various rural classes and the changes among them. These changes in relative positions enable us to see how the classes have fared over time. This section also presents the proportion of population of the various classes and the changes in this proportion. As a whole, this section will be helpful to understand the changes in the relative position of the various classes and their impact on the composition of inequality, which is presented in the following section.
The relative position of each class is defined as the ratio of two mean monthly per capita consumption expenditures (henceforth, MPCE). The numerator is the mean of the MPCE of the respective class, whereas the denominator is the mean of the MPCE of the whole rural population. Now, if the mean ratio is greater than the unity, it implies that the specific class in question is better than the average. A mean ratio value less than the unity implies that the mean MPCE of the specific class is below average. Since this study is focused on looking at the changes in the relative positions of each class over time, the growth of the mean ratios from one period to the next has been calculated. If the growth ratio attains a negative sign for any particular class, it means that the mean MPCE of that class has increased at a slower rate than the mean of the population MPCE. One can infer that the relative position of that specific class has worsened over the period. Other classes might have witnessed an increase in their mean MPCE. Table 3 presents the mean ratio, the proportion of population, and the growth ratio for five farm and six nonfarm classes for 1993–1994 and 2004–2005. The following observations can be made from Table 3. In the rural income ladder, the agricultural worker is relatively poorer in the rural farm sector, as the mean ratio is less than 1 for this particular class as compared to the other classes within the farm sector. On the other hand, relatively poorer classes in the nonfarm sector are the nonagricultural self-employed and the nonagricultural worker class, as the mean ratio for these classes in 1993–1994 was less than 1. However, even if the mean ratio has recently become greater than the unity for these two classes, they remain at the bottom rung as compared to other nonagricultural classes. Thus, it is evident that, in rural society, the agricultural worker, the nonagricultural worker, and the nonagricultural self-employed classes remained relatively poorer throughout the period.
Rural Classes: Proportion of Population and Relative Position, 1993/1994–2004/2005
A careful analysis of the relative position of the aforementioned rural classes and changes in them will direct us to an understanding of the changes in the inequality pattern. The changes in the relative position of each class are captured through the growth ratio presented in Column 5 of Table 3. It is observed that in the farm sector, during the period 1993/1994–2004/2005, the relatively poorer class, that is, the agricultural workers, witnessed a negative growth ratio, along with the marginal and rich farmers. Considering the relative position of the rich farmers and marginal farmers in 1993–1994, it is evident from the table that they lost out on growth during this period. However, the middle and small-farmer classes witnessed a positive growth ratio. This indicates that during the first period, the upper and lower segments had lost out on growth, whereas the middle picked up.
Similarly, the relatively poorer classes in the nonfarm sector, such as the nonagricultural self-employed class, witnessed a negative growth, ratio while the remaining classes witnessed a positive growth ratio. The data further reveal that the relatively better-off classes, that is, the absentee landlord categories and the rural professional class, experienced gains at a higher rate. On the other hand, among the relatively poorer classes, the nonagricultural workers recorded a gain, while the nonagricultural self-employed class witnessed a relative loss. However, the gain for the nonagricultural class was not as much as what the better-off classes achieved. Thus, in the first period, it is evident that there is a clear disparity between the better-off and the worse-off classes in the nonfarm sector, which indicates a worsening condition regarding inequality.
These changes in relative positions also need to be observed with respect to the changes in the proportion of population of various classes. The classes that witnessed a major decline in their population shares were those of the rich, middle, and small farmers, whereas the marginal farmers/tenants and the agricultural workers witnessed an increase in their population share. This indicates a movement of population across classes. Most of the population began to concentrate in the marginal farmer/tenant class and the agricultural worker class, which together accounted for 50 percent of the population. At the end of 2004–2005, the farm sector in Bengal was beginning to witness the dominance of the marginal farmer and agricultural worker classes in terms of their population share. On the other hand, in 1993–1994, the nonfarm sector accounted for 38.57 percent of the rural population. Out of this, the shares of the nonagricultural self-employed, the nonagricultural workers, and the absentee landlords who are self-employed in the nonfarm sector accounted for 14.59 percent, 11.3 percent, and 5.84 percent of the population, respectively. Of the remaining classes, the absentee landlord and the other classes represented 1.53 percent and 3.59 percent of the population, respectively. In 2004–2005, a significant increase in the population share is observed for one of the classes in the absentee category, the absentee landlords who are also self-employed in the nonagricultural sector, and a significant decline is observed for the rural professional classes. However, of the remaining classes, the rural nonfarm sectors witnessed very small changes in their population share. For example, the nonagricultural workers and the nonagricultural self-employed witnessed a small increase in their population shares, whereas the absentee landlords witnessed a small decline in their population share. Thus, at the end of 2004–2005, the rural nonfarm sectors were fairly stagnant in terms of changes with respect to a few classes.
The following period, from 2004–2005 to 2009–2010, exhibited a different trend. Almost all the classes witnessed a positive growth ratio, except for the middle farmers. As noted earlier, the agricultural sector witnessed a reduction in its inequality throughout the period. However, an improvement in the relative position of the rich as well as the relatively poorer class in the farm sector leaves us with the question of inequality reduction in the farm sector (see Table 4).
Among the relatively poorer classes in the nonfarm sector during this period, the nonagricultural self-employed class registered a positive growth ratio, while the nonagricultural worker class registered a negative growth ratio. This is exactly the opposite of what had happened in the previous period with respect to these two classes. Nevertheless, among the better-off classes, the class of absentee landlords who are also self-employed in the nonfarm sector witnessed a negative ratio growth along with the rural professional classes. The biggest gain in this phase was recorded by the absentee landlord class.
This trend of changing relative positions needs to be observed along with the changes in the proportion of population for each class. From 2004/2005 to 2009/2010, almost all the classes in the farm sector witnessed a significant decline in their population share, while the agricultural worker class witnessed an increase in its population share. This agricultural worker class held the major share of around 39 percent of the population. It is evident that the farm sector was mainly dominated by the agricultural worker class in this period. Similarly, the rural nonfarm sector shows significant changes with respect to the three classes: nonagricultural worker, nonagricultural self-employed, and both the classes in the absentee landlord category. A significant increase in the population share is observed in the case of the nonagricultural workers and the nonagricultural self-employed, whereas a significant decline can be seen for both the classes of the absentee categories (see Table 4). Thus, it is evident that the rural nonfarm sector was dominated by the nonagricultural worker and the nonagricultural self-employed classes in 2009–2010.
Rural Classes: Proportion of Population and Relative Position, 2004/2005–2009/2010
Thus, by using the NSSO data, this section observes certain characteristics of the different classes in terms of their relative positions. During the first period, it is the middle and small farmers who experienced improvement within the farm sector, while within the nonfarm sector, both the classes belonging to the absentee categories along with the nonagricultural worker class witnessed improvement in their respective relative positions. The next period seems to pose a different trend, where the middle farmer in the farm sector witnessed a worsening relative position, along with the absentee landlords who are also self-employed in the nonagricultural sector and the nonagricultural workers. These three classes experienced gains in the first period but witnessed an alteration in their situation during the second period. The results from observing the relative position of each class and the changes in them along with an inspection of the movement of population across classes over time will indicate the way in which the composition of inequality might have changed. This will be taken up in the following section, where the decomposition of inequality is discussed.
Decomposition of Inequality
The decomposition exercise enables us to identify the inter- and intra-class components of inequality and also helps to understand which component influences the changes in the overall inequality. Table 5 shows that from 1993–1994 to 2004–2005, the inter-class component marginally declined from 0.047 to 0.045, whereas the intra-class component increased from 0.207 to 0.228. Thus, it is the intra-class component which primarily influenced the increasing inequality in this phase.
The marginal decline in the inter-class component can be understood from the earlier discussion on changes in the relative position of various classes. It was observed that the middle and small farmers witnessed relative gains, whereas the rich farmer witnessed a negative mean ratio growth rate, implying a relative loss of position for this particular class. On the other hand, within the nonfarm sector, almost all the classes witnessed some improvement in their relative position. This could be the reason why the inter-class component did not register much change.
On the other hand, the intra-class component is the sum total of the class-wise contribution of inequality which, in turn, depends on the consumption share, Gini coefficient of the class, and the overlapping index. Columns 4 and 9 in Table 5 show that the contribution of marginal farmers/tenants, agricultural workers, nonagricultural self-employed, nonagricultural worker, and absentee landlords who are identified as self-employed classes had largely increased. The increasing contribution of these classes was a result of their consumption share which had increased during this time. In addition to this, most of these classes also witnessed a rising Gini coefficient except for the marginal farmer/tenant class, implying that the distribution within these classes had become unequal during this period. It is important to note that most of these classes also witnessed an increasing proportion of population as described in the earlier section.
Decomposition of Rural Inequality Between 1993/1994 and 2004/2005
Thus, as a whole, the marginal farmers/tenants, agricultural workers, nonagricultural self-employed, nonagricultural workers, and the absentee landlord who are self-employed in the nonagricultural sector had influenced the intra-class component to rise which, in turn, also led the overall Gini index to increase.
During the next period, 2004/2005 to 2009/2010, rural Bengal witnessed declining inequality. The decomposition exercise suggests that both the inter- and intra-class components decreased during this period. The declining inter-class component implies that the distance between the classes might have decreased. As discussed in the earlier section, during this phase, the relatively poorer classes, that is, the marginal farmers/tenants, the agricultural workers, and the small farmers in the farm sector witnessed improvements in their relative positions, whereas within the nonfarm sector, the nonagricultural self-employed class also witnessed a relative improvement. Even though the rich farmers and the absentee landlords who are identified as others also witnessed improvement, their proportion of population was too small to influence the overall distribution.
On the other hand, the intra-class component declined because of the declining contribution of the rich, middle, small, and marginal farmers, both the classes of the absentee categories, etc. The consumption shares of these classes had declined, which largely influenced their contribution to the decline over this period. This declining consumption share is also associated with the declining proportion of population of these classes. The overlapping index presented in Columns 3 and 8 of Table 6 shows that the relatively richer classes, that is, the absentee landlord and others, witnessed a declining value, indicating that this class had formed a separate segment in the overall distribution. However, this class also represented a very negligible proportion of the population.
Decomposition of Rural Inequality Between 2004/2005 and 2009/2010
As a whole, the inequality exercise suggests that the relatively poorer people were dominating the rural income distribution and, in turn, caused the overall inequality to decline during this period. In addition to this, the marginal farmers/tenants and the agricultural workers also witnessed an improvement in their relative positions within the farm sector. However, within the nonfarm sector, the nonagricultural workers witnessed a negative mean ratio growth which implies the declining relative position of this class.
Conclusion
The above analyses have identified two qualitatively different phases of agrarian growth and inequality, that is, increasing inequality with a relatively moderate level of agrarian growth and declining inequality with marginally declining agrarian growth. This article makes an attempt to make sense of the trend of increasing inequality in the given agricultural growth by providing a class analysis.
The first phase, 1993/1994 to 2004/2005, is that of increasing inequality, when agriculture grew at a moderately high rate. In this phase of increasing inequality, the farm sector witnessed declining inequality. An improvement in the relative position of the middle farmers followed by the small farmers was observed. The rest of the classes—the rich farmers, the marginal farmers, and tenants and the agricultural workers in the farm sector—witnessed a worsening condition of their relative position during this period. The proportion of population for the rich, middle, and small farmer classes declined, whereas the proportion of population for the marginal farmers and tenants and the agricultural workers increased. These two classes together accounted for 50 percent of the rural population in 2004–2005. It appears that the benefits of agricultural growth accrued mostly to the middle and small farmers. The rising cost of cultivation, resulting from the increasing cost of fertilizers and pesticides, affected the marginal farmer class, which might be a reason behind its worsening relative position in this period. On the other hand, the worsening relative position of the agricultural workers could be because of the negative growth of the agricultural wage observed between 1993–1994 and 1999–2000 (Srivastava & Singh, 2005). Apart from this, another tendency was observed in the farm sector, that is, the increasing share of the marginal farmers and agricultural workers as compared to the other farming classes. There might be various processes which generated this trend of the rising share of the marginal farmers or tenants and agricultural workers. The implementation of such policies as land redistribution was a continuous process which might have benefitted the landless poor people who could now obtain some land. In this regard, Bakshi (2008) showed that 445,403 hectares of land were redistributed, and the number of beneficiaries was 2817,719 until 2004–2005. Bakshi (2008) also found evidence of the redistribution of land among the Dalits and Muslims in her survey villages. Apart from this, the tendency of large landowners to dispose land might have also worked. Rawal (2001) found evidence of the buying of land by small landowners in his surveyed villages. Moreover, Bakshi (2008) also argued that the land ceiling law worked as one of the deterrent factors for large landowners which prevented them from buying more land. Bardhan and Mookherjee (2011) argued that operation Barga signaled to the large farmers about their political commitment to the landless people which might have influenced divisions among the large farmer households. Moreover, the ‘Patta’ was another incentive among the small and marginal farmers to become further sub divided so that they could obtain a land title.
In the rural nonfarm sector, the classes which saw improvement in their relative positions both belong to the absentee landlord category. Both of these classes had income which was already above the average mean ratios. Among them, the absentee landlords who are self-employed in the nonagricultural sector witnessed an increase in their population share. This indicates a possibility of people from the farm sector diversifying and joining this class. It was also observed in the second section of this article that the proportion of employment in the service sector had increased during this period. This could be because services such as wholesale, trading, and transportation began to emerge. Harriss-White (2008) also showed that, during this period, investment in the husking mills of rural Bengal continued to rise and, as a result, the number of husking mills increased. According to her field observations, rich farmers took interest in investing in these husking mills. In fact, there was also a massive increase in small firms making puffed rice and rice from raw paddy. This partly confirms the trend of the increasing share of the self-employed class in the rural nonfarm sector. Moreover, even though the number of the rice mills started shrinking during this period, it started increasingly to become technologically upgraded and capital-intensive (Harriss-White, 2008). This probably explains the relative gain of the absentee landlords who are identified as others. On the other hand, of the remaining classes within the relatively poorer class, the nonagricultural self-employed class witnessed a negative mean ratio growth. This class witnessed the highest population share among all the classes in the nonfarm sector, contributing 14 percent to the rural population in 2004–2005. Thus, the strengthening of the relatively better-off absentee landlord class with nonagricultural self-employed enterprises and the near stagnation or decline of the other classes were perhaps the underlying reasons behind the increasing inequality in the nonfarm sector during this period.
Moving on to the next period, the farm sector witnessed a process of proletarization: agricultural workers witnessed a sharp increase in their population share. This class also witnessed positive growths in its mean ratios. This is accompanied by a sharp decline in the population share for the peasantry as a whole. Raju et al. (2015) have confirmed the trend of increasing labor absorption for paddy cultivation during the post-reform period for the state. Even when the entire peasantry witnessed a positive mean ratio growth, it was not enough to counteract the sharp decline in their population ratio. Thus, this phenomenon of the strengthening of the poorest strata in the farm sector and the weakening of the remaining classes explain the declining farm inequality. During this same phase when Indian agriculture was witnessing a revival, West Bengal witnessed a deceleration in its agricultural growth. The tendency of division among the households and land had been observed in the earlier period itself, and this trend continued in the period that followed. Bardhan and Mookherjee (2011) also confirm that the increasing trend of divisions among the households happened rapidly. In addition to this, since agricultural growth started declining further, it will not be an exaggeration to surmise that farming had become nonviable for most of the farmers, which might have also influenced the farmers to leave farming and work as laborers in the agricultural fields or join the nonfarm sector.
This process of proletarization is not just confined to the farm sector; it spills over to the nonfarm sector as well. The nonagricultural working class witnessed an increase in its population share from 11.35 percent to 14.90 percent. Apart from this, the other class which gained in population share during this period is the nonagricultural self-employed class, whose share increased from 14.94 to 19.55 percent. As noted earlier, these classes are the poorest classes in the nonfarm sector and during this period, they experienced increases in their population share. Chowdhury and Chakraborty (2014) argued that the growth of nonagricultural employment was the highest in West Bengal from 2004–2005 to 2009–2010. In fact, the second section of this article also showed that the proportion of rural employment had increased in the manufacturing and service sectors during this time. In this regard, Chowdhury and Chakraborty (2014) confirmed this trend by using the NSSO data, concluding that most of the employment was generated in construction, retail trade and tobacco manufacturing industries. They also argued that such jobs were of low quality in nature, with no employment contract, and social security. The overall industrialization in the state had not taken place properly when the agriculture sector witnessed a deceleration in its growth rate. It appears that some employment opportunities had been created in sectors like construction, retail trade, and tobacco industries which provide a very poor quality of employment. Thus, it appears that this declining inequality does not signify development; rather, it points to a process of proletarization that has begun in rural Bengal.
Footnotes
Acknowledgements
I thank Prof. Nagraj for his sincere comments on an earlier draft. I also thank Prof. R. Vijay, Professor Vathsala Narasimhan and Prof. Vamsi Vakulabharanam for their helpful comments and suggestions. I also thank my co-researchers for their suggestions.
Declaration of Conflicting Interests
The author declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.
Funding
The author received no financial support for the research, authorship and/or publication of this article.