Knowledge that Transforms

We investigate the causal impact of public discourse on socially responsible market behavior. Across three laboratory experiments, having market participants engage in public discourse generally increases market social responsibility. These positive impacts are robust to variation in several characteristics of the discourse. We provide evidence that discourse strengthens beliefs that others support socially responsible exchange. However, relaxing requirements to engage in discourse sharply reduces its effectiveness. Our findings suggest that campaigns encouraging discussion of appropriate market behavior can have sizable impacts on addressing inefficiencies due to market failures but that policies encouraging broad public engagement may be important. (JEL C92, D62, D83, D91, P36, M14, Z13)

This study examines the link between credit supply and hospital health outcomes. We use bank stress tests as exogenous shocks to credit access for hospitals that have lending relationships with tested banks. We find that affected hospitals shift their operations to increase resource utilization following a negative credit shock but reduce the quality of their care to patients across a variety of measures, including a significant increase in risk-adjusted readmission and mortality rates. The results indicate that access to credit can affect the quality of health care hospitals deliver, pointing to important spillover effects of credit market frictions on health outcomes. (JEL G21, G32, I11, I18)

We study regressions with multiple treatments and a set of controls that is flexible enough to purge omitted variable bias. We show these regressions generally fail to estimate convex averages of heterogeneous treatment effects—instead, estimates of each treatment’s effect are contaminated by nonconvex averages of the effects of other treatments. We discuss three estimation approaches that avoid such contamination bias, including the targeting of easiest-to-estimate weighted average effects. A reanalysis of nine empirical applications finds economically and statistically meaningful contamination bias in observational studies; contamination bias in experimental studies is more limited due to smaller variability in propensity scores. (JEL C21, C31, C51, H75, I21, I28)

How do radical reforms shape economic development over time? In 1790, the French Constituent Assembly overhauled the kingdom’s organization to establish new local capitals. In some departments, the choice of local capitals over rival candidate cities was plausibly exogenous. We study how changes in administrative presence affect state capacity and development in the ensuing decades. In the short run, administrative proximity increases taxation and investments in law enforcement. In the long run, capitals obtain more public goods and grow faster. Our results shed light on the dynamic impacts of state building following one of history’s most ambitious administrative reforms. (JEL D70, H41, H71, O18, O43)

This paper identifies how bank branching benefited local economies during the Great Depression. Using archival data and narrative evidence, I show how Bank of America’s branch network in 1930s California created an internal capital market that diversified away local liquidity shortfalls, allowing the bank to maintain 49 percent higher credit growth from 1929 to 1933 than competing banks. The bank’s presence mitigated cites’ property value contractions and strengthened their recovery through 1940. Linked individual data show that the bank’s proximity to workers hastened the transition from agricultural employment to human-capital–intensive sectors in the 1930s, generating structural change and higher wages. (JEL E32, G01, G21, N12, N22, N92, R23)

We propose a theory of gradualism in the implementation of good policies, suitable for environments featuring time consistency. We downplay the role of the initial period by allowing agents both to wait for future agents to start equilibrium play and to restart the equilibrium by ignoring past history. The allocation gradually transits toward one that weighs both short- and long-term concerns, stopping short of the Ramsey outcome but greatly improving upon Markovian equilibria. We use the theory to account for the slow emergence of both climate policies and the reduction of global tariff rates. (JEL C73, E21, E61, F13, H30)

Once established, government spending programs tend to continue. A commonly held view is that spending inertia leads to unsustainable debt, ultimately requiring fiscal adjustments such as “sequestration.” We show that by insuring against political turnover, inertia may reduce politicians’ incentives to accumulate debt. However, large preexisting commitments and the prospect of future stabilization can lead to overspending to dilute past administrations’ commitments. Finally, we show that political polarization amplifies incentives to prioritize inertial programs, potentially explaining the increased share of mandatory spending in the US budget. (JEL D72, E62, H61, H63)

This paper provides empirically implementable sufficient statistics formulas for optimal nonlinear tax systems in the presence of across-income heterogeneity in preferences, inheritances, income-shifting capabilities, and other sources. We characterize optimal smooth tax systems on income and savings (or other commodities), as well as simpler tax systems. We use familiar elasticity concepts and a novel sufficient statistic for heterogeneity correlated with earnings ability: the difference between across-income variation in savings and the causal effect of income on savings. We apply these formulas to the United States and find that the optimal savings tax is mostly positive and progressive. (JEL E21, G51, H21, H24)

Politicians may target public goods to benefit their constituents, at the expense of others. I study corruption in the context of Indian electricity and estimate the welfare consequences. Using new administrative billing data and close-election regression discontinuities, I show that billed electricity consumption is lower for constituencies of the winning party by almost 40 percent, while actual consumption, measured by nighttime lights, is higher. I document the covert way in which politicians subsidize constituents by manipulating bills. These actions have substantial welfare implications, with an efficiency loss of US$0.9 billion, leading to unreliable electricity supply and significant negative consequences for development. (JEL D72, D73, L94, L98, O13, O17)

We study the design of negotiation strategies when a principal and agent must decide how to split a pie while the agent's outside option changes over time. The principal's optimal strategy under commitment demonstrates a new, but intuitive, set of negotiation dynamics. When the agent is tempted to leave, the principal gradually promises a larger share (decreasing demands) and more time to explore the outside option (decreasing pressure), illustrating a complementarity between these two tools. Although the principal's expected utility is decreasing in the outside option, his expected utility and demands are increasing in the outside option's drift and volatility. (JEL D63, D82, D86)