Crowding-Out or Crowding-In? Public and Private Investment in India with Mehdi Raissi and Volodymyr Tulin, World Development, 2018, 109: 323-333.

This paper contributes to the debate on the relationship between public-capital accumulation and private investment in India along the following dimensions. First, acknowledging major structural changes that the Indian economy has undergone in the past three decades, we study whether public investment in recent years has become more or less complementary to private investment in comparison to the period before 1980. Second, we construct a novel data-set of quarterly aggregate public and private investment in India over the period 1996Q2-2015Q1 using investment-project data from the CapEx-CMIE database. Third, embedding a theory-driven long-run relationship on the model, we estimate a range of Structural Vector Error Correction Models (SVECMs) to re-examine the public and private investment relationship in India. Identification is achieved by decomposing shocks into those with transitory and permanent effects. Our results suggest that while public-capital accumulation crowds out private investment in India over 1950-2012, the opposite is true when we restrict the sample post 1980 or conduct a quarterly analysis since 1996Q2. This change can most likely be attributed to the policy reforms which started during the early 1980s and gained momentum after the 1991 crises.

(Under review)

Welfare programs are implemented for a variety of reasons. However, of key importance is to understand the economic impact of such government transfers at the local level. This paper estimates the multiplier effects of transfers spending using a novel data-set of state-level expenditure under rural welfare programs in India between 1980-2010. We use narrative analysis to motivate state- level heterogeneity in program expenditure during the implementation of new nation-wide programs as being largely exogenous to local output fluctuations. We estimate local multipliers using this "narrative shock series" and find government transfers to be quite consequential for the local economic activity.

This paper underscores two key aspects in the impact evaluation of policy interventions. First, while defining "treatment" and "control" based on exposure to policy, it’s important to acknowledge older policies. A new policy can be misinterpreted to have had no impact if the existence of a comparable, older policy is ignored. Second, it’s important to measure both "spot" and "stock" effects of the policy. Comparing the impact of public works on private wages, while the paper does not find any measurable impact on wages via spot-effects, the build-up of program capital, or stock-effect, does explain significant increases in private wages.

Recent evidence suggests that large employment guarantee schemes in India have increased private wages. Juxtaposed with this body of work are studies that show how the lack of administrative capacity, political will, and other supply factors cause program provision to be rather limited and highly variable across districts and over time. This paper attempts to understand the cost of variability in program provision in terms of the labor market outcomes. We find that in the presence of downward wage rigidity, forward-looking employers compress wage increases today because of the uncertainty regarding the level of program provision in the future. Our theory generates two key empirically verified predictions: (i) greater variability in program provision results in a larger compression of wage increases; and (ii) that compression of wage increases is more severe in districts where inflation is low relative to where inflation is high. This has important policy implications as we show that by simply reducing the variability in program provision, without increasing the average expenditure, can be welfare enhancing.