74847 CASE STUDY WORLD BANK | AGRICULTURE AND DEVELOPMENT DECEMBER 11, 2012 Women’s Inheritance Rights and Intergenerational Transmission of Resources in India KLAUS DEININGER, APARAJITA GOYAL, AND HARI NAGARAJAN ABSTRACT across reform and non-reform states, based —————–——————————————————————–-———————— on whether the father died before or after the While developing countries have made considerable progress in equalizing HSAA. In reform states, whether or not a women’s economic opportunities, inheritance legislation often remains strongly woman is entitled to inherit land depends on gender-biased. We use inheritance patterns over three generations of individuals the timing of her father’s death. For women to assess the impact of changes made to the Hindu Succession Act that grant whose father died before the amendment daughters equal coparcenary birth rights in joint family property. We show that came into force in her state, the 1956 HSA although the amendment significantly increased daughters’ likelihood to inherit land, substantial bias persists. Our results also indicate a robust increase in applied; for those whose father died educational attainment of daughters, suggesting an alternative channel of wealth thereafter, the HSAA governed any transfer. inheritance-related wealth transfers. Non- Hindu households as well as households from BACKGROUND birth. These are commonly referred to as the non-reform states are not affected by the Hindu Succession Act Amendment (HSAA), HSAA, and thus serve as potential control Inheritance regulations play an important and were passed in the following states: groups. role in determining how wealth is Andhra Pradesh (1986), Tamil Nadu (1989), transferred across generations. Although Maharashtra (1994), and Karnataka (1994). First, we compare the difference in likelihood there is substantial evidence of inheritance The HSAA thus constitutes an interesting of inheriting land between generation II males legislation being gender biased in many natural experiment that allows us to explore and females belonging to the same countries, the impacts of such legislation whether changes in inheritance legislation can household, before and after the HSAA came have received little attention in the literature improve women’s access to physical and into force. We do this by observing the (Behrman and Rosenzweig 2004). To human capital. amounts of land inherited by siblings within a address this gap, we examine the impact of household, and the legal change is tested to reforms (implemented 1986-1994) that DESIGN affect the probability of females’ receiving land strengthened women’s inheritance rights in inheritance. India. We use data from the 2006 Rural Economic and Demographic Survey (REDS), conducted The Hindu Succession Act (HSA) of 1956 by India’s National Council for Applied followed the main (mitakshara) Hindu law Economic Research on a nationally tradition and distinguished individual representative sample of the country’s 16 property from joint ancestral assets, main states. Aside from standard household including land (Agarwal 1994). While the survey information, REDS covers key data on former could be bequeathed at will, rights to the parents of the current household head the latter were limited to a group referred to (generation I), the household head’s siblings as the coparcenary, which included only (generation II), and the household head’s male members of a dynasty. As a result, children (generation III). In particular, we women were rarely able to inherit land. In focus on a sample of 8,190 rural households order to eliminate gender inequality, a (a total of 72,655 individuals across the three number of states im p le m e n t e d generations). substantively similar amendments to the HSA, stipulating that the daughter of a To estimate the impact of the HSAA, we coparcener acquires coparcenary rights by compare female and male land inheritance Figure 1: Woman harvesting vegetables WWW.WORLDBANK.ORG/ARD 2 WOMEN’S INHERITANCE RIGHTS AND INTERGENERATIONAL TRANSMISSION OF RESOURCES “Married Post HSAA� is included as an Table 1: Household Characteristics indicator variable because based on Type of state Reform states only traditional customs, inter-generational asset Total G 1 male dead transfers to females may occur at the time Non- G 1 male died sample Reform died after reform alive before of marriage rather than the father’s death. reform reform Panel A: Generation I Second, we compare various outcome Land ownership (share) 0.74 0.73 0.77*** 0.73 0.72 0.74** variables for females depending on whether Area owned (acres) 6.47 5.24 7.06*** 4.21 6.68 4.65*** they live in a reform state and whether their Male schooling (years) 2.18 2.23 2.18* 2.24 2.25 2.22* father died before or after the HSAA took Female schooling (years) 1.35 1.43 1.25* 1.45 1.43 1.41 effect. In so doing, we can see if the HSAA Hindu (share) 0.93 0.96 0.92*** 0.96 0.96 0.97 had an overall positive impact. This is No of observations 8190 2574 5616 662 901 1011 important because fathers who are required Panel B: Generation II to transfer land to their female offspring Males post-HSA may cut back on other transfers Years of schooling 4.49 4.64 4.16* 4.68 4.78 4.26* in anticipation of such bequests. The Inherited any land (share) 0.61 0.58 0.62** 0.04 0.68 0.71* dependent variables we examine are 1) the Area inherited (acres) 2.46 2.56 2.31* 0.01 2.38 2.14* share of household land inherited by the Gifts given values 1,975 2,267 1,624*** 2,675 1,612 2,930*** female, 2) the value of total transfers (gifts Total transfer values 445,896 317,702 511,813*** 299,624 299,509 349,725*** and land) received, and 3) the amount of No of observations 15,586 5,417 10,169 1,335 1,828 2,254 land owned by the female in 2006 at the Females time of the interview. Years of schooling 3.24 3.51 3.19* 3.54 3.52 3.54* Inherited any land (share) 0.08 0.11 0.06*** 0.03 0.08 0.17*** Although we expect the HSAA to most Area inherited (acres) 0.25 0.41 0.16*** 0 0.32 0.56** directly affect land inheritance, exploring Value of Gifts given 6,581 8,284 5,852*** 8,513 6,307 9,543*** some of the downstream effects postulated No of observations 14,902 4,346 10,556 1,146 1,555 1,645 in the literature can provide additional Panel C: Generation III insight. HSAA’s impacts on generation III’s Males human capital are of particular interest; Years of schooling 5.33 5.43 5.29* 5.67 5.06 5.23** information on generation III’s education No of observations 13,905 3,682 10,223 822 1338 1522 level allows us to make inferences on possible longer-term effects of the HSAA. Females Greater access to inherited assets may Years of schooling 4.15 4.46 3.76*** 4.83 3.97 4.61** strengthen the position of generation II No of observations 11,882 3164 8718 695 1103 1366 females in intra-household bargaining, thus Notes: Authors’ computations using the NCAER ARIS -REDS 2006 Survey. Stars indicate significance of mean differences across reform and non-reform states, as well as across the increasing human capital by the next timing of generation I male’s death. * significant at 10%; ** significant at 5%; *** significant at 1% generation. Or, if greater land transfers to females are accompanied by reduced interest. In the total sample, only 8 percent of RESULTS access to education, HSAA’s positive females inherited land (compared to 61 effects may be weakened or cancelled out, percent for males). The likelihood of males Table 2 shows the results of our empirical possibly resulting in an overall negative inheriting land was virtually unaffected by the analysis. We use data on 6,891 individuals in impact. Thus we explore how the HSAA HSAA (68 percent for males whose father 1,805 land owning Hindu households, in may affect female bargaining power and its died before HSAA versus 71 percent for which the generation I male had died at the effect on seemingly unrelated matters. those whose father died after the HSAA). In point of survey. The basic result in column 1 contrast, for females the corresponding suggests that females are significantly less DESCRIPTIVE EVIDENCE figures are 8 percent and 17 percent. This likely (by 72 percent) to inherit land than suggests that the HSAA had a significant males. However, females whose father died Basic attributes for the three generations in impact on female inheritance of land. after the HSAA had come into effect are our sample are summarized in Table 1. roughly 15 percentage points more likely to Panel A shows data for generation I; a total Panel C presents data on generation III’s inherit land than those whose father died of 5,616 households are in non-reform educational attainment. The average level of before the reform. While this indicates states and 2,574 households are in reform education for the total sample of generation HSAA’s relatively large impact on women’s states. It is not surprising that there are III females is 4.15 years (6 years of likelihood to inherit land, it also suggests that significant differences between reform and education corresponds to completion of legal change alone cannot completely non-reform states across almost all elementary school). For generation II compensate for the underlying gender bias. observable characteristics. females, this figure was 3.24 years, and for generation I females, 1.35 years. This In column 2, we find that females whose Panel B, which provides data on 30,488 indicates that educational attainment among father died in the five years immediately after generation II individuals, is of particular women has been increasing over time. t he HSAA c am e i nt o eff ect are 10. 1 3 WOMEN’S INHERITANCE RIGHTS AND INTERGENERATIONAL TRANSMISSION OF RESOURCES percentage points more likely to inherit land. Table2: Effect of the Hindu Succession Act Amendment on Women’s Inheritance of Any Land For females whose father died six or more Hindus Non-Hindus years after the reform, this effect increases to (1) (2) (3) (4) (5) 23.5 percent. This may be due to increased Female -0.724 -0.71 -0.723 -0.714 -0.541 awareness and knowledge of the HSAA over [0.175]*** [0.164]*** [0.205]*** [0.215]*** [0.172]*** time. Female*Father's Death Post 0.145 0.152 0.147 0.081 HSAA [0.036]*** [0.034]*** [0.042]*** [0.063] Columns 3 and 4 show interactions with several variables, such as father’s land Female* Death Pre 1-6 Years 0.009 endowment and the caste of the generation I [0.040] male. Interestingly, HSAA’s effects are Female* Death Post0-5 Years 0.101 estimated to be slightly more pronounced for [0.047]*** SC/ST (Scheduled Castes or Scheduled Tribes) households and for households with Female* Death Post6+ Years 0.235 less land endowment. This may be due to [0.042]*** lower caste households possibly being more Female* Death*Father's Land- -0.036 -0.034 -0.07 holding receptive to new opportunities or having [0.011]*** [0.013]*** [0.054] stronger incentives to move away from Female* Death*SC/ST 0.176 0.182 traditional norms (Luke and Munshi 2011). [0.054]*** [0.061]** Column 4 results also indicate that the HSAA Female*Death*Married Post affects land transfers primarily through 0.043 0.051 HSAA inheritance (rather than through marriage). [0.056] [0.115] Finally, column 5 results show that the HSAA Observations 6,891 6,891 6,891 6,891 345 did not affect non-Hindus. R-squared 0.73 0.74 0.74 0.74 0.88 Notes: The sample consists of generation II males and females in the four reform states. As HSAA’s positive impacts on female land Father’s Death Post HSAA is an indicator variable for whether the father of the individual died after the reform. Father’s Death Pre 1-6 Years is an indicator variable for whether the father’s inheritance may be counteracted by negative death occurred in the six years leading up to the reform. Father’s Death Post0 -5 Years is an impacts in other areas, it is important to indicator variable for whether the father died in the first five years after the reform, and Father’s assess whether there were any substitution Death Post6+ Years is an indicator variable for whether the father died in the sixth year after the reform and beyond. Father's landholding is a continuous variable. SC/ST refers to scheduled effects. Table 3 shows that the share of land castes or scheduled tribes. Married Post HSAA is an indicator variable for whether the individual inherited by females (column 1) and total land married after the reform. All regressions include gender specific year of birth fixed effects and owned (column 2) increased significantly for household fixed effects. All pair wise interactions are included. Robust standard errors in those in reform states whose father died after brackets are clustered by village. *significant at 10%; ** significant at 5%; *** significant at 1% the HSAA took effect in their state. In addition, results in column 3 suggest that the value of gifts in reform states is slightly higher for females who married post HSAA. Such results support the notion that an increase in female land inheritance was not compensated for by a reduction in other transfers. Thus, HSAA can be considered to have had a positive net effect. To explore one of the longer-term effects of the HSAA, we include Table 4 which summarizes the effect on educational attainment. We find that girls whose educational decisions were made after the HSAA cane into effect had 0.37 more years of elementary schooling than their older cohorts (column 1). Father’s land holding and caste do not appear to have a significant effect, although parents’ education is estimated to have a greater impact on girls’ schooling (column 2). In order to determine whether our estimates are picking up broader trends in educational attainment which are independent of HSAA’s Figure 2: Children helping with farm work 4 WOMEN’S INHERITANCE RIGHTS AND INTERGENERATIONAL TRANSMISSION OF RESOURCES impact, we present results for the Table 3: Effect of the Hindu Succession Act Amendment on Women’s Land Inheritance and Other Transfers educational attainment of non-Hindu girls in Share of land Land owned Total gifts column 3. From this, we are able to rule out inherited (log) given (log) the existence of a common trend in female (1) (2) (3) education across religions. In column 4, we Reform State*Father's Death Post HSAA 0.021 0.117 0.137 examine the impact of HSAA on boys’ [0.008]** [0.022]*** [0.160] education in order to see if there are any Observations 9,877 9,877 9,877 R-squared 0.62 0.62 0.58 spillover effects. Remarkably, we find no RS*Father's Death Post HSAA 0.023 0.109 0.124 evidence that improving inheritance rights [0.009]** [0.0417]*** [0.157] for daughters had any significant effect on RS *Married Post HSAA 0.012 0.077 0.356 the education levels of their brothers (i.e. [0.013] [0.13] [0.180]* no negative spillover). RS *Death Post HSAA*Married Post HSAA -0.008 -0.041 0.317 [0.006] [0.032] [0.242] CONCLUSION Observations 9,877 9,877 9,877 R-squared 0.62 0.63 0.58 Our analysis of HSAA’s impact provides Notes: The sample consists of all ever married generation II Hindu females in the reform and non potentially important lessons for India -reform states. Reform State (RS) is an indicator variable that takes the value 1 if the state is Andhra Pradesh, Tamil Nadu, Maharashtra or Karnataka and zero otherwise. All regressions (where similar changes were made in 2005 include state specific year of birth fixed effects and household fixed effects. All pair wise on a national scale) and other countries interactions are included but not shown. Robust standard errors in brackets are clustered by where inheritance rights remain strongly village * significant at 10%; ** significant at 5%; *** significant at 1%. biased against women. We find that although the HSAA significantly increased Table 4: Effect of the Hindu Succession Act Amendment on Educational Attainment women’s likelihood to inherit land, it did not Hindu Girls Non-Hindu Girls Hindu Boys fully compensate for the underlying gender (1) (2) (3) (4) inequality. The finding of a significant Reform State*Young 0.373 0.243 -0.043 -0.06 increase in girls’ educational attainment [0.059]*** [0.101]** [0.026] [0.114] after the HSAA suggests that it led to a Reform State*Young*Father's Land Holding 0.005 0.011 0.024 genuine improvement in women’s [0.011] [0.093] [0.018] socio-economic status; it did not lead to Reform State*Young*Father's Education 0.024 0.019 0.007 substitution from human capital to physical [0.013]* [0.023] [0.003]* capital for women. Reform State*Young*Mother's Education 0.009 0.021 0.008 [0.004]* [0.016] [0.012] While we have robust evidence in favor of Reform State*Young*SC/ST 0.083 0.085 legal changes like HSAA, there are [0.131] [0.056] several areas where further research could improve our understanding of the dynamics Observations 7762 7762 679 9247 of legal and behavioral change. For R-squared 0.13 0.16 0.32 0.17 example, since legal changes fall short of Notes: The sample consists of generation III males and females across reform and non-reform states. All regressions include state specific year of birth fixed effects and household fixed effects. achieving the objective of fully equalizing All pair wise interactions are included but not shown. Robust standard errors in brackets are women’s inheritance status to that of men, clustered by village. * significant at 10%; ** significant at 5%; *** significant at 1%. efforts to identify factors that could help to close this gap are necessary. In addition, it REFERENCES This case study was prepared by Klaus is possible that legal reforms like HSAA Deininger from the Development Economics prompt changes in households’ Agarwal, B.1994. A Field of One's Own: Research Group, Aparajita Goyal from the reproductive behavior; these dynamics also Gender and Land Rights in South Asia South Economics and Policy Group of the need to be examined. Finally, in order to Asian Studies. Cambridge; New York and Agriculture and Environmental Services circumvent such legal reforms, parents Melbourne: Cambridge University Press. Department of the World Bank, and Hari could become more strategic by Nagarajan, Senior Fellow at the National converting joint assets into individual Behrman, J. R. and M. R. Rosenzweig. Council of Applied Economic Research. The property so that they may be bequeathed 2004. "Parental allocations to children: New World Bank’s Gender Action Plan Trust Fund at will to their sons. This could potentially evidence on bequest differences among and the Global Land Tools Network provided lead to an adverse effect on women’s asset siblings." Review of Economics and generous financial support. ownership. Examining such issues would Statistics 86 (2): 637-40. The findings, interpretations, and conclusions allow for a better appreciation of the expressed are entirely those of the authors. magnitude and incidence of associated Luke, N. and K. Munshi. 2011. "Women as They do not necessarily represent the views of welfare effects. agents of change: Female income and the Government of India, and of the World mobility in India." Journal of Development Bank and its affiliated organizations, or those Economics 94 (1): 1-17. of the Executive Directors of the World Bank or the governments they represent.