Knowledge that Transforms

Matching theory typically assumes that agents know their values for possible partners and confines attention to settings in which matching is either static, or driven by population dynamics. In many environments of interest, instead, dynamics originate in the agents learning their preferences through interactions with other agents. In this short paper, we illustrate how platforms can use appropriately designed auctions to account for the joint value of experimentation and cross-subsidization in dynamic matching markets. The model is a stylized version of the general one in Fershtman and Pavan (2016).

Maternal and neonatal mortality rates in the slums of Nairobi, Kenya are among the highest in the world. Mounting evidence suggests that delivering in a facility is not enough to ensure mortality reductions: women must deliver in high-quality facilities and arrive early enough for appropriate care if complications arise. We designed an RCT combining labeled cash transfers and pre-commitment incentives to encourage earlier and more effective delivery facility choice and to promote earlier facility arrival. We find that the intervention improves planning, increases delivery at the desired facility, and encourages more timely arrival at delivery facilities.

This paper introduces and analyzes a procedure called Testing-Based Forward Model Selection (TBFMS) in linear regression problems. This procedure inductively selects covariates that add predictive power into a working statistical model before estimating a final regression. The criterion for deciding which covariate to include next and when to stop including covariates is derived from a profile of traditional statistical hypothesis tests. This paper proves probabilistic bounds for prediction error and the number of selected covariates, which depend on the quality of the tests. The bounds are then specialized to a case with heteroskedastic data with tests derived from Huber-Eicker-White standard errors. TBFMS performance is compared to Lasso and Post-Lasso in simulation studies. TBFMS is then analyzed as a component into larger post-model selection estimation problems for structural economic parameters. Finally, TBFMS is used to illustrate an empirical application to estimating determinants of economic growth.

We conducted an experiment marketing microloans to farmers in the USA during Spring 2015 and found a simple direct mail letter increased borrowing from a government program. The subsequent spring, we built on this finding and enriched the design to test for information spillovers. The direct effect result did not replicate in the second year, thus lowering the likelihood that spillovers would be present and detectable. These results add to recent evidence on how (seemingly subtle) differences in context and treatment content affect consumer responses.

The relations between unobserved events and observed outcomes can be characterized by a bipartite graph. We propose an algorithm that explores the structure of the graph to construct the “exact Core Determining Class,” i.e., the set of irredudant inequalities. We prove that in general the exact Core Determining Class does not depend on the probability measure of the outcomes but only on the structure of the graph. For more general linear inequalities selection problems, we propose a statistical procedure similar to the Dantzig Selector to select the truly informative constraints. We demonstrate performances of our procedures in Monte-Carlo experiments.

Structural estimation of matching games with transferable utility, including matching games of trading networks and many-to-many matching, is increasingly popular in empirical work. I explore several modeling decisions that need to be made when specifying a structural model for a matching game. One decision is the choice of a game theoretic solution concept to impose in the structural model. I discuss pairwise stability, competitive equilibrium, and noncooperative games such as auctions. Another decision is whether to work with a continuum of agents or a finite number of agents. I explore other issues as well.

Asymmetric information is a key feature of the marriage market. In HIV-endemic settings, HIV risk is an important partner attribute that may influence marriage timing and partner selection. We use a sample of married women in rural Malawi to validate a model of positive assortative matching under asymmetric information. Several correlations support this framework, suggesting that HIV risk contributes to adverse selection in the marriage market in this setting.

Manmade climate change (CC) has catastrophic consequences. The United States has already experienced wholesale population realignment due to climate as households have relocated to the Sunbelt and West. The irony is that people are moving toward the heat and major storms associated with CC. As CC intensifies, with high rates of internal US factor mobility, firms and households will likely again relocate to areas with higher utility and profits, reducing CC costs. Yet current research typically focuses on CC costs in a given location without considering this realignment. We propose several avenues to overcome such shortcomings in US CC modeling.