Research Database

We benchmark six global optimization algorithms by comparing their performance on challenging multidimensional test functions as well as on a method of simulated moments estimation of a panel data model of earnings dynamics. Five of the algorithms are from the popular NLopt open-source library: (i) Controlled Random Search with local mutation (CRS), (ii) Improved Stochastic Ranking Evolution Strategy (ISRES), (iii) Multi-Level Single-Linkage (MLSL), (iv) Stochastic Global Optimization (StoGo), and (v) Evolutionary Strategy with Cauchy distribution (ESCH). The sixth algorithm is TikTak, which is a multistart global optimization algorithm used in some recent economic applications. For completeness, we add three popular local algorithms to the comparison—the Nelder-Mead downhill simplex algorithm, the Derivative-Free Nonlinear Least Squares (DFNLS) algorithm, and a popular variant of the Davidon-Fletcher-Powell (DFPMIN) algorithm. To give a detailed comparison of algorithms, we use benchmarking tools recently developed in the optimization literature. We find that the success rate of many optimizers varies dramatically with the characteristics of each problem and the computational budget that is available. Overall, TikTak is the strongest performer both on the test functions and the economic application. The next-best performing optimizers are StoGo for the test functions and MLSL and ISRES for the economic application.

We specify and calibrate a life-cycle model of labor supply and savings incorporating health shocks and medical treatment decisions. Our model features endogenous wage formation via human capital accumulation, employer-sponsored health insurance, and means-tested social insurance. We use the model to study the effects of health shocks on health, labor supply and earnings, and to assess how health shocks contribute to earnings inequality. We also simulate provision of public insurance to agents who lack employer-sponsored insurance. The public insurance program substantially increases medical usage by the uninsured, leading to improved health and life expectancy, which generates higher Social Security costs. But the program also creates positive labor supply incentives, and substantially reduces costs of social insurance, Medicaid and free care. On balance the net program cost is modest, and all agents in the model are ex ante better off in a balanced budget simulation. In contrast, improving access to Medicaid has perverse labor supply effects, does little to improve health, and makes almost all agents worse off in a balanced budget scenario.

This paper studies the implications of household heterogeneity for trade. I develop a model where household heterogeneity is induced via incomplete markets and results in heterogeneous price elasticities. Conditional on exposure to trade, heterogeneous price elasticities imply that different households value price changes differently, and thus rich and poor households experience different gains from trade. I calibrate the model to match bilateral trade flows and micro-facts about household-level expenditure patterns and elasticities. I find gains from trade that are pro-poor and that the average gains from trade are substantially larger than representative agent benchmarks.

This paper develops a New Keynesian model with sovereign default risk. Inflation is set by forward-looking firms, monetary policy is an interest rate rule, and the fiscal government borrows externally, long-term, with an option to default. In this framework, default risk creates inflation pressures through an expectations channel, and tight monetary policy disincentivizes fiscal overborrowing. The model sheds light on temporary inflation events in emerging market data, short-lived spikes in inflation, spreads, and domestic policy rates. As spreads rise, firms increase their prices in expectation of higher future inflation during defaults. Monetary policy tightens, which reduces inflation and helps bring spreads down by disciplining government borrowing. These monetary-fiscal interactions imply that delivering the flexible price allocation may not be optimal for monetary policy.

How generous should social insurance be when quits account for a large share of transitions into non-employment? We address this question using a multi-sector directed search model extended to incorporate endogenous quits both to other jobs and to non-employment. Workers quit too often in the competitive equilibrium, and private markets co-ordinate on excessively high “efficiency” wages. Quantitatively, we find that unemployment insurance is optimally much less generous in an economy with quits than in one without. An extended Baily-Chetty formula is derived to illustrate the source of this difference.

Multinational affiliates are more productive than domestic firms, so how do they affect a host country through the labor market? We use data for Norway to show that the labor market is characterized by a job ladder, with multinationals on the upper rungs. We calibrate a general equilibrium job ladder model with endogenous multinational entry to the Norwegian data. In a counterfactual where multinationals face an infinite entry cost, payments to labor fall and profits of domestic firms rise, but the impact is heterogeneous. Competition for workers increases low down on the job ladder, while it decreases high up.

We develop a two-country macroeconomic model that we fit to a set of aggregate prices and quantities for the U.S. and the rest of the world. In addition to a standard array of shocks, the model includes time variation in agents’ preference for safe bonds. We allow for a component of this time variation to be common across countries and biased toward dollar-denominated safe assets, and refer to this component as global flight to safety (GFS). We find that GFS shocks are the most important shocks driving world business cycles, and are also important drivers of activity in the U.S. and especially abroad. An adverse GFS shock lowers global GDP and inflation, widens global corporate credit spreads, and appreciates the dollar. These effects are very close to those obtained from a structural VAR which uses the excess bond premium (Gilchrist and Zakraj¡sek, 2012) as proxy for global flight to safety.

Heathcote et al. (2010) conducted an empirical analysis of several dimensions of inequality in the United States over the years 1967-2006, using publicly-available survey data. This paper expands the analysis, and extends it to 2021. We find that since the early 2000s, the college wage premium has stopped growing, and the race wage gap has stalled. However, the gender wage gap has kept shrinking. Both individual- and household-level income inequality have continued to rise at the top, while the cyclical component of inequality dominates dynamics below the median. Inequality in consumption expenditures has remained remarkably stable over time. Income pooling within the family and redistribution by the government have enormous impacts on the dynamics of household-level inequality, with the role of the family diminishing and that of the government growing over time. In particular, largely due to generous government transfers, the COVID recession has been the first downturn in fifty years in which inequality in disposable income and consumption actually declined.