The Work of Andres Rodriguez-Clare

Trade theory and the growth of nations

Andres Rodriguez-Clare is a Costa Rican-American economist, and the Edward G. and Nancy S. Jordan Professor of Economics at UC Berkeley. Previously, he was a professor at Penn State University from 2005 to 2011, and an associate professor at the University of Chicago from 1994 to 1999. After his father was elected President of Costa Rica, he worked as Chief of the Council of Economic Advisors from 1998 to 2002. He received a bachelor’s from the Universidad de Costa Rica in 1988, a master’s from Ohio State in 1989, and a PhD in economics from Stanford University in 1993. He was elected a fellow of the Econometric Society in 2024. His work focuses on international development and trade.

With Pete Klenow, he had a substantial influence on development accounting in “The Neoclassical Revival in Growth Economics: Has It Gone Too Far?”. The “neoclassical revival” of their title is typified by Mankiw-Romer-Weil (1992), and seeks to find how much of the differences in income between countries can be explained by variation in capital accumulation. In particular, Mankiw-Romer-Weil consider human capital, or the skills of a people, as a capital which is accumulated much the same as physical capital like machines and railroads are. They then regress the level of human and physical capital on output, and find that capital accumulation explains the vast majority of economic growth.

What Klenow and Rodriguez-Clare do is consider the endogeneity of capital accumulation in response to different policies, as countries which discourage capital accumulation will also tend to discourage things such as the adoption of new technology. They instead first pin down the coefficients by estimating the return to schooling, and then use those returns to estimate the total contribution to economic growth. They find that, if the exponents in the production function are constrained to be in line with microeconomic data, most of the differences in productivity between countries return to being due to differences in productivity.

His best known work in trade theory is “New Trade Models, Same Old Gains?”, with Costas Arkolakis and Arnaud Costinot. The increased availability of firm-level microdata allowed economists to extend standard trade models to account for firms with differing productivities. Increased international trade, as in the Melitz (2003) model, causes more business to go to the more productive firms, allowing the possible gains from trade to be larger than under prior economic models. What Arkolakis, Costinot, and Rodriguez-Clare show is that, so long as entering firms draw from a Pareto distribution of possible productivities, the gains from trade can be entirely summarized as the share of domestic expenditures and the elasticity of trade with respect to trade costs. This makes it equivalent to models like that of Krugman (1980) with homogenous firms. The microeconomic organization of firms does not matter, so long as the Pareto distribution assumption holds, and one can estimate the gains from trade with linear regressions using easily available macro data. Rodriguez-Clare and Costinot would use this method in “Trade Theory With Numbers”, estimating the gains from trade to the US to be between 2 and 8 percent.

With Dave Donaldson, Arkolakis, Costinot, and Rodriguez-Clare would analyze models in which markups are allowed to vary (such as that of Melitz and Ottoviano, 2008) in “The Elusive Pro-Competitive Effects of Competition”. What they show is that, if there is a Pareto distribution of firms and tastes are non-homothetic (meaning that the proportion of goods purchased is not the same as income changes) competition actually decreases the gains from trade relative to models with constant markups. The intuition here is that more of our consumption shifts to the more productive firms, who have higher markups as a consequence. In their work, the pro-competitive effects of trade are “elusive”.

Rodriguez-Clare would tackle the elusive benefits of industrial policy with Dominick Bartelme, Costinot and Donaldson in “The Textbook Case for Industrial Policy”. The classic argument for industrial policy is that if industries get more efficient as they become larger, and these economies of scale are external to the firm, then firms will underinvest. A government subsidy can increase total welfare. What they show is that the external economies of scale can be estimated from data on trade flows, but that the possible gains from an optimal industrial policy are not very large. Given that any industrial policy greatly expands the scope for rent-seeking, as well as being based on correlations which may not persist over time, we might safely dismiss it as a policy choice.

A theme of his work in trade theory is on the diffusion of ideas across national borders. In works like “Trade, Diffusion, and the Gains from Openness”, he argues that the gains from trade consists of both direct trade and goods, and also a general openness to foreign direct investment, and new ideas. This can explain why the gains from trade itself can be small – most economists' notion of trade incorporates openness to the outside world. It also suggests that the Lerner symmetry need not hold when ideas are diffused through contact with the outside world. Thus, import protection like tariffs is equivalent to an export subsidy when considering only what goods are produced, but is not equivalent in its effects on welfare.

In a related work, “Multinationals, Linkages, and Economic Development”, he explores the conditions under which it is better to have a multinational company open a branch in a developing country, or to have it trade for the goods with a domestic firm. He pin points the existence of backward and forward linkages as key. To be precise, linkages follow when there is a love of variety in the use of inputs, inputs must be purchased locally, and there are increasing returns to scale. Backwards linkages are when demand from a final goods producer leads to more specialized inputs being made, benefiting all the other final goods producers; forward linkages are when an input producer increases their output, allowing more variety for final good producers. These are generally measured with input-output tables. Since there are increasing returns, multiple equilibria exist. Imagine two countries, A and B, with A in a high wage equilibrium, and B in a low wage one. He defines a “linkage coefficient” as the ratio of employment generated in upstream industries against the labor hired directly. If this is positive, the multinational headquartered in A and producing in B makes B better off. If it is negative, it makes them worse off.

Lastly, he, Costinot, and Ivan Werning make the somewhat puckish case for import subsidies to the least profitable foreign firms, in “Optimal Trade Policy with Firm Heterogeneity”. The starting point is a Melitz model, where firms select into trade. Since they don’t take into account the effects of their entry on prices, the optimal measure for the government is to pay the fixed costs of the least productive foreign firms, and have them export to your country. This is actually selfishly optimal, and is the reverse of the possible arguments for a tariff being selfishly optimal.

He is rumored to be an excellent table tennis player. I have not had the opportunity to test this myself, but I challenge the professor to a match whenever I should see him.