Knowledge that Transforms

Decisions are typically made by drawing on multiple sources of information, yet the correlation among these sources is often hard to assess. We study robustly optimal strategies—those that maximize the worst-case payoff guarantee across all correlation structures. With two states and two actions, the robustly optimal strategy is strikingly simple: use only the best information source and ignore the rest. With more actions, a robustly optimal strategy decomposes the problem into local comparisons and assigns each comparison to the source that is best at that comparison. With more states, the same specialization logic survives through a concavification characterization: the decision problem is decomposed into payoff-relevant components, and each component is assigned to its best source. These results offer a new rationale for why ignoring information can be optimal and highlight the benefit of using specialized information sources when correlations are ambiguous.

In this paper I revisit the interpretation of the linear instrumental variables (IV) estimand as a weighted average of conditional local average treatment effects (LATEs). I focus on a situation in which additional covariates are required for identification while the reduced-form and first-stage regressions may be misspecified due to an implicit homogeneity restriction on the effects of the instrument. I show that the weights on some conditional LATEs are negative and the IV estimand is no longer interpretable as a causal effect under a weaker version of monotonicity, i.e. when there are compliers but no defiers at some covariate values and defiers but no compliers elsewhere. The problem of negative weights disappears in the interacted specification of Angrist and Imbens (I1995), which avoids misspecification and seems to be underused in applied work. I illustrate my findings in an application to the causal effects of pretrial detention on case outcomes. In this setting, I reject the stronger version of monotonicity, demonstrate that the interacted instruments are sufficiently strong for consistent estimation using the jackknife methodology, and present several estimates that are economically and statistically different, depending on whether the interacted instruments are used.

We study how online job search advice affects job search behavior and labor market outcomes of unemployed workers. In a large-scale field experiment embedded in the online platform of the Danish public employment service, we provide job seekers with vacancy information and occupational recommendations. Using a two-stage randomized design with regionally varying treatment intensities, we examine both direct effects of advice and treatment spillovers. We find that advice significantly increases employment and earnings when provided to a limited share of job seekers. As treatment intensity rises, these effects diminish and eventually disappear. The decline is driven by negative spillovers among treated job seekers, consistent with increased congestion in the occupations toward which job seekers redirect their search. Spillover effects on untreated job seekers are heterogeneous and depend on how advice alters occupation-specific competition in local labor markets.

Quality regulation seeks to ensure quality and foster competition by reducing vertical differentiation, but it may also have adverse effects on market structure. We study this trade-off in the context of pharmaceutical bioequivalence, which is the primary quality standard for generic drugs. By exploiting the introduction of bioequivalence in Chile, we find that stronger regulation reduced the number of drugs in the market by 18% and increased average paid prices by 13%. We estimate a model of drug entry, certification, and demand to study how drug quality, aversion to generics, and certification costs shape the equilibrium effects of quality regulation. We find that quality regulation increased demand for generic drugs by resolving asymmetric information and reducing aversion to unbranded generics, which induced the entry of high-quality drugs in place of low-quality drugs. Consumer welfare increased despite higher prices and a lower number of firms. We compare minimum quality standards with alternative designs of quality regulation, including quality disclosure.

To study competition, innovation, and industry dynamics that arise as a result of their interaction, we develop a new oligopolistic general-equilibrium Schumpeterian growth model. This model ties together the endogenous growth, oligopolistic competition, and dynamic industrial organization literatures in a single unified framework. Within each industry, there are an endogenously determined number of large firms (“superstars”) that compete á la Cournot and a continuum of small firms which collectively constitute a competitive fringe. Firms dynamically choose their innovation strategies, cognizant of other firms' choices, and their entry and exit are endogenous. The model is consistent with the macroeconomic trends observed in the United States since the 1970s, such as the domination of industries by a small number of superstar firms, the rise of markups, market concentration, profits, and R&D spending, and the decline in business dynamism, productivity growth, and the labor share. It replicates the empirical relationship between innovation and competition within and across industries. As an application, we estimate the model to disentangle the effects of separate mechanisms on the structural transition observed in the United States, which yields striking results: (1) While the increase in the average markup causes a significant static welfare loss, this loss is overshadowed by the dynamic welfare gains from increased innovation in response to higher profit opportunities. (2) The increasing costs of innovation are found to be the primary determinant of lackluster productivity growth, i.e., ideas are getting harder to find.

When financial constraints bind, firms adjust employment in response to cash flow shocks. A 2010 revaluation of business rates, a UK tax levied on business-occupied properties, implied that similar firms, occupying similar properties in narrow geographical locations, experienced different tax changes. I find that, on average, for every ￡1 of additional cash flow triggered by the tax change, 39 pence were spent on employment, with small and leveraged firms responding the most. I use these findings to discipline a general equilibrium model with firm heterogeneity and financial frictions and quantitatively determine the aggregate effects of a fiscal transfer to firms.

Standard New Keynesian (NK) models feature an optimal inflation target well below 2%, limited welfare losses from business cycle fluctuations and long-term monetary neutrality. We develop an NK framework with endogenous productivity and downward nominal wage rigidity (DNWR) which challenges these results. The interaction between endogenous growth and DNWR generates asymmetric hysteresis effects on unemployment and R&D. As a consequence, the model features a nonvertical long-run Phillips curve and a trade-off between price distortions and output hysteresis that changes the welfare-maximizing inflation rate to above 2%. Deviations from the optimal target carry welfare costs multiple times those in traditional NK models. Taylor rules that respond to the unemployment rate address more effectively the asymmetric hysteresis in our model. Results are robust to the inclusion of the effective lower bound on interest rates.

We compare and combine two distinct approaches to estimating market power in the US brewing industry—one based on production data, the other on demand. Both methods produce similar estimates of average markups and reveal a recent upward trend. We then combine the two approaches in two ways. First, we replace conventional instruments in demand estimation with a moment involving a production-based markup estimate, yielding similar results. We then evaluate the common (but controversial) assumption that retail markets operate competitively, finding that demand-based markups recovered with the assumption of competitive retail markets align with production-based estimates, as long as downstream costs are accounted for.

The structure of the Social Security program presents a trade-off between benefit duration and size; individuals who claim early receive lower monthly payments. Surprisingly, even wealthy Americans with high life expectancies frequently claim benefits at the earliest eligibility age, a behavior inconsistent with conventional life-cycle theory predictions. In fact, there is a striking similarity in claiming ages across groups with vastly different life expectancies, wealth levels, and incentives to delay. This paper resolves this puzzle by introducing policy details and subjective beliefs about longevity and program rules into a rich, forward-looking life-cycle model. When included in the model, subjective beliefs substantially narrow the claiming-age gaps across education (college vs. non-college) and marital-status (married vs. single) groups. Specifically, differences in program beliefs alone account for 48.6% of the education claiming age gap and 75.1% of the marital status gap. Counterfactual policy simulations demonstrate that failing to account for these claiming mechanisms dramatically overstates individual responses to an increase in the normal retirement age and understates the resulting fiscal savings.

Even as global oil demand has increased in recent years, it is plausible to envision scenarios in which clean energy technologies or climate policies drive demand to essentially zero by the century’s end. This paper asks what such a demand decline, when anticipated, might mean for global oil supply. One possibility is a “green paradox”: producers accelerate extraction. However, because extraction requires durable capital investments, the opposite may occur: producers reduce their investment rate, decreasing extraction. To evaluate the relative strengths of these opposing mechanisms, I develop a model of global oil supply that incorporates both, among other industry features. For model inputs with the strongest empirical support, both effects are modest, with disinvestment typically outweighing the green paradox. In order for the green paradox to substantially increase cumulative global oil extraction, investments must have short time horizons, and producers’ discount rates must be less than 4% real.