This is the sixth in this year’s series of posts by PhD students on the job market
Social transfer programs have become increasingly common as a form of social protection across the developing world. As social transfers expand in developing countries, it is crucial to understand their labor market effects, which could have important implications for their effectiveness and distributional impacts. In theory, social transfers could reduce labor supply through an income effect or increase it through a health-productivity effect, but empirical evidence on either effect is scarce. Furthermore, there is little evidence on how possible changes in labor supply from social transfers affect wages and welfare. For instance, it is possible that a decrease in labor supply will drive up wages, improving the welfare of the poor who are typically net labor suppliers. This pro-poor effect could be particularly beneficial in years with negative economic shocks, when poor households rely on wage labor as a coping mechanism. It is thus possible that social transfers have large indirect poverty-reducing effects in addition to the direct effects of the transfers.
In my job market paper – with Kathy Baylis and Ben Crost – we examine the labor supply and wage effects of one of the world’s largest social transfer programs - India’s Public Distribution System (PDS). The PDS provides in-kind transfers of staple food to the poor at a highly subsidized price. The program is India’s most important social program, providing assistance to over 800 million people and accounting for 60% of the social assistance budget. More broadly, in-kind transfers - particularly of food - are an important part of social transfers around the world. Approximately 1.5 billion people worldwide receive in-kind food transfers (World Bank, 2018), and about 44% of individuals covered by social programs receive in-kind transfers (World Bank, 2015).
Identification Strategy : Mandated targets from the National Food Security Act (NFSA)
Before the NFSA of 2013, states in India had substantial discretion in setting the prices and quantities of PDS rations provided to program beneficiaries. This changed after the passage of the NFSA, which imposed national targets on all states. For instance, states were mandated to provide 5 kg per capita of staple grains to eligible households every month at prices no higher than 3 Rs/kg for rice and 2 Rs/kg for wheat. States whose pre-NFSA prices or quantities fell short of those targets had to expand their subsidies while states who were already in compliance did not.
The above figure shows a time-series of the PDS rice price offered to beneficiary households by the eight states in our data. As shown in the left panel, Pre-NFSA, states had substantial discretion over the prices offered. Post-NFSA, states that were out of compliance were forced to bring down prices to comply with the NFSA mandate. For instance, the states of Bihar and Maharashtra reduced their prices from 7 Rs/kg and 6 Rs/kg to the mandated price of 3 Rs/kg. To isolate the variation generated by the NFSA from discretionary state-level changes to PDS entitlements, we implement an instrumental variables approach based on counterfactual entitlements that would have existed if states had expanded PDS by the bare minimum needed to comply with the NFSA mandate. As shown in the right panel of Figure 1, we assume that states in compliance with the price mandate, such as Jharkhand and Andhra Pradesh, made no changes to PDS prices. We further assume that states that voluntarily lowered their prices beyond NFSA targets, such as Madhya Pradesh and Karnataka, only did the bare minimum to comply with the mandate. Finally, we assume that all states complied with the mandate in June 2013, when NFSA was officially enacted, ignoring state-level variation in the timing of the reform’s implementation.
In addition, the NFSA mandated that states calculate PDS rations on a per-capita basis, allocating 5 kg of subsidized grain per eligible household member. States that calculated ration on a per-household level were forced to switch to a per-individual allocation, leading to more generous transfers for large households relative to small ones. As a result of these mandated changes, the NFSA generated substantial variation in PDS entitlements across state, time, household-size and PDS eligibility status. This variation lends itself well to estimating the causal effect of the transfer, since it was generated by a national rule and is therefore not likely correlated with changes in local policies or economic conditions. We combine this policy variation with individual-level data from ICRISAT’s “Village Dynamics in South Asia” panel between 2010 and 2015.
Transfers reduce labor supply and increase wages
We find that increases in transfer value, equivalent to the PDS expansion in the state of Bihar, causes labor supply to decrease by 3.29% and increase wages by 8.59%. We show that the program’s effect on wage leads to a redistribution of income from richer households, who are net labor buyers, to poorer households, who are net labor suppliers. The indirect benefits to the poor from increased wages are large relative to the direct effect of the transfer. For the poorest quintile the indirect welfare gains from the wage increase is about 5.4% of household consumption.
Transfers reduce poor’s reliance on labor income in bad years
We find that the program effects are strongest for men in poor households working outside the household in the non-farm market, consistent with the notion that poor households use non-farm labor as a coping mechanism. In addition, the program effects are larger in years with a bad productivity shock, particularly in the lean season. These results suggest that social transfers reduce the labor that is supplied by the poor to cope with risk. Our results also suggest that social transfers can mitigate the excess sensitivity of labor supply and wages to productivity shocks, as shown in Jayachandran (2006), by reducing the dependence of the poor on labor income during bad years, and hence can stabilize wages against negative economic shocks.
Policy implications
Policy-makers have historically been concerned that social transfers decrease labor supply and that this may make them less effective at reducing poverty. While our results provide evidence that social transfers cause a modest reduction in labor supply, they show that the resulting increase in wages has substantial positive implications for poverty-reduction. Our results also suggest that social transfers can mitigate the vicious cycle of high labor supply and low wages that afflicts poor households in bad years by reducing their dependence on labor income as a coping mechanism. Overall, our results imply that the labor market effects of social transfers beget important additional benefits for the poor.
Aditya Shrinivas is a post-doc at Cornell University.
Join the Conversation