Knowledge that Transforms

Using data from a two-year pricing experiment, we study the impact of subsidy policies on weather insurance take-up. Results show that subsidies increase future insurance take-up through their influence on payout experiences. Exploring mechanisms of the payout effect, we find that for households that randomly benefited from financial education, receiving a payout provides a one-time learning experience that improves take-up permanently. In contrast, households with poor insurance knowledge continuously update take-up decisions based on recent experiences with disasters and payouts. Combining subsidy policies with financial education can thus be effective in promoting long-run insurance adoption.

We examine the consequences of a sudden increase in household debt burdens by exploiting variation in exposure to household foreign currency debt during Hungary’s late-2008 currency crisis. The revaluation of debt burdens causes higher default rates and a collapse in spending. These responses lead to a worse local recession, driven by a decline in local demand, and negative spillover effects on nearby borrowers without foreign currency debt. The estimates translate into an output multiplier on higher debt service of 1.67. The impact of debt revaluation is particularly severe when foreign currency debt is concentrated on household, rather than firm, balance sheets.

Rural economies in many developing countries are characterized by a lean season in the months preceding harvest, when farmers have depleted their cash and grain savings from the previous year. To identify the impacts of liquidity during the lean season, we offered subsidized loans in randomly selected villages in rural Zambia. Ninety-eight percent of households took up the loan. Loan eligibility led to increases in on-farm labor and agricultural output, driving up wages in local labor markets. Larger effects for poorer households suggest that liquidity constraints contribute to inequality in rural economies.

Behaviors and information often spread via person-to-person diffusion. This paper highlights how diffusion processes can facilitate coordination. I study contagion in a discrete network with Bayesian players. In addition to characterizing the extent and rate of adoption, we uncover a new effect: when large cascades are possible in equilibrium, exposure conveys information about a player’s network position. This effect underscores a novel trade-off in the design of marketing campaigns, suggesting conditions under which word-of-mouth is relatively more effective. A generalization of the model to multi-type networks suggests a new approach to targeted seeding.

We study the long-run effects of forced migration on investment in education. After World War II, millions of Poles were forcibly uprooted from the Kresy territories of eastern Poland and resettled (primarily) in the newly acquired Western Territories, from which the Germans were expelled. We combine historical censuses with newly collected survey data to show that, while there were no pre-WWII differences in educational attainment, Poles with a family history of forced migration are significantly more educated today than other Poles. These results are driven by a shift in preferences away from material possessions toward investment in human capital.

We formalize the argument that political disagreements can be traced to a “clash of narratives.” Drawing on the “Bayesian Networks” literature, we represent a narrative by a causal model that maps actions into consequences, weaving a selection of other random variables into the story. Narratives generate beliefs by interpreting long-run correlations between these variables. An equilibrium is defined as a probability distribution over narrative-policy pairs that maximize a representative agent's anticipatory utility, capturing the idea that people are drawn to hopeful narratives. Our equilibrium analysis sheds light on the structure of prevailing narratives, the variables they involve, the policies they sustain, and their contribution to political polarization.

Abhijit Banerjee delivered his Prize Lecture on Sunday 8 December 2019, at the Aula Magna, Stockholm University.

We consider bargaining environments in which public opinion provides leverage by making compromises costly. Two parties make initial demands, before knowing the public opinion. If deadlocked, they can bargain again after public opinion forms, but suffer reputation costs if they compromise, i.e., scale back their demands. We show that in equilibrium, parties may choose to make incompatible demands initially and gamble over public opinion even though one or both parties must bear a cost later. We characterize when deadlocks arise, and how this affects the welfare of the public in a representative two-party democracy compared to a direct democracy.

We study a bargaining model in which a buyer makes frequent offers to a privately informed seller, while gradually learning about the seller’s type from “news.” We show that the buyer’s ability to leverage this information to extract more surplus from the seller is remarkably limited. In fact, the buyer gains nothing from the ability to negotiate a better price despite the fact that a negotiation must take place in equilibrium. During the negotiation, the buyer engages in a form of costly “experimentation” by making offers that are sure to earn her negative payoffs if accepted, but speed up learning and improve her continuation payoff if rejected. We investigate the effects of market power by comparing our results to a setting with competitive buyers. Both efficiency and the seller’s payoff can decrease by introducing competition among buyers.