Welcome

Energy subsidies are one of the most common forms of in-kind transfer worldwide, yet they are inefficient and often regressive. However, when governments have limited data with which to identify poor households, in-kind transfers can outperform cash as a means of progressive redistribution. We formulate a model of redistribution with limited information to evaluate fuel subsidy and cash policies in Indonesia. In our model, a planner with progressive redistributional goals and limited information chooses among cash transfers, nonlinear fuel subsidies, and combinations. The optimality of a nonlinear fuel subsidy versus a targeted cash transfer is determined by the joint distribution of household income, observable characteristics, and household fuel demand. We estimate the primitives of this model for the Indonesian population using rich administrative survey data and variation generated by large fuel policy reforms and quantify optimal interventions. Combinations of cash and in-kind programs, involving self-financing fuel pricing policies, generate double the welfare gain of pure targeted cash or fuel policy approaches.

Anti-price gouging laws, present in most US states, penalize retailers if they make large price increases to disaster supplies during states of emergency. Price caps during periods of high and inelastic demand may worsen or alleviate shortages of essential supplies — they can decrease the incentive to restock supplies by reducing the resale price at which the new inventory sells, or they can increase the incentive to restock supplies by generating more unmet demand at initial inventory levels. We use retailer scanner data and novel trucking data to estimate the effects of US natural disasters on quantities transacted and prices. We estimate that disasters increase demand for a set of 20 essential goods, with increases in mean quantities transacted but also in the probability of having a stockout. Prices increase, with the largest changes concentrated in a minority of retailers, but we see no effects of anti-price gouging laws on the probability of price hikes. On the supply-side, we see limited evidence of marginal cost increases and we estimate a shift in restocking from the disaster period to the week proceeding it. Motivated by this evidence, we specify a structural model that will evaluate how stringent enforcement of price caps would affect shortages and consumer surplus during disasters. We construct a non-parametric identification argument that combines an instrument for restocking costs with the observed joint distribution of restocking and quantities transacted to recover latent demand and initial inventory levels.

The US is one of few countries worldwide without a national mandate for paid maternity leave, and as such, we know little about the effects of paid leave in this context. The first cases of statewide maternity leave in the U.S. came about somewhat unintentionally through changes in state Temporary Disability Insurance (TDI) in the 1960s and 1970s. Six to seven decades later, we have the opportunity to explore the long run effects of access to paid maternity leave on children, later in their lives. We do so employing a differences-in-discontinuities design around changes in state TDI policies which made disability insurance available for pregnancy. In "first stage" results, we provide new estimates of the effects of paid maternity leave availability on infant birth weights, improving upon existing methodologies estimating the same effect in the literature, and finding much larger positive effects. In a "second stage" exercise, we will link measures of the education and earnings of children with their birthdate records using Census and Numident data to provide the first estimates of the long run impacts of paid maternity leave on children for the US.