Working Papers
Structural Change under Limited Growth Pass-Through [paper]
The speed of structural change has slowed over recent decades, while market power and aggregate markups have risen. This paper studies how changes in market structure contributed to this slowdown. I incorporate heterogeneous firms and variable markups into a model of structural change. Market power breaks the proportional link between productivity and prices: productivity improvements are only partly transmitted into lower sectoral prices, and the strength of this growth pass-through depends on the firm distribution. Sectors with a thicker upper tail of the firm distribution exhibit lower growth pass-through. I quantify the mechanism for the United States from 1980 to 2014 and find that growth pass-through declined in every sector as markups have risen. In a counterfactual with full growth pass-through, the expenditure share of services would have been 2.9 percentage points higher by 2014, corresponding to an additional 31.4% of reallocation over the sample period. The effect primarily operates through weaker transmission of productivity growth into prices in the industry sector.
Income Distribution, Welfare, and the Patterns of Trade [update][CESifo WP]
We study how income inequality shapes prices, trade, and welfare in an open economy. We develop a two-country trade model featuring non-homothetic preferences and within-country income inequality. Firms face a trade-off between price and market size, generating an endogenous price schedule in which cheap necessities coexist with expensive luxury goods. Pricing-to-market is driven by firms’ ability to allocate fixed costs across countries. In a North-South setting, the richer country bears a larger share of fixed costs, raising prices particularly for poorer consumers, leading to a Manhattan effect. Within-country inequality also affects trade: a mean-preserving spread shifts demand towards luxuries. In an open economy, pricing-to-market introduces an additional channel absent in autarky: firms shift a larger share of fixed costs toward the more unequal country, causing more consumers there to lose from rising inequality and propagating the effects of increased inequality to its trading partner. Allowing international arbitrage eliminates fixed-cost shifting and the Manhattan effect by imposing uniform pricing across countries.
Parallel Trade Policy in General Equilibrium [paper]
International arbitrage limits firms’ ability to price discriminate across markets. Firms can respond not only by changing prices, but also by withdrawing from low-income markets altogether. We study the trade policy implications of this extensive margin response in a general equilibrium North-South model. Trade policy is a joint decision over trade barriers and parallel trade policy. When income differences are sufficiently large, allowing parallel trade induces some Northern firms to stop exporting to the South to preserve higher prices at home. Restricting parallel trade therefore expands market access in the South, lowers Southern prices, and improves its terms of trade. As a result, the South prefers to prohibit parallel trade, whereas the North prefers to allow it. Parallel trade also overturns conventional incentives over trade barriers: the South prefers positive trade costs because they weaken the arbitrage constraint and induce more Northern firms to export, while the North prefers free trade. When parallel trade is prohibited, both countries prefer free trade. The globally optimal policy combines free trade with a ban on parallel trade.
Work in Progress
The Sources of Labour Market Change
We document labour market change across six advanced economies using harmonised longitudinal household data from the Cross-National Equivalent File and develop an accounting framework that reconciles individual worker transitions with observed employment stocks. Manufacturing employment declines in every country, accompanied by a corresponding rise in services. Sectoral mobility is substantial at annual frequency but strongly directional: services combines high retention with substantial inflows from other sectors, making it the principal destination of reallocation. Workers entering manufacturing typically experience larger wage gains than those already employed there, whereas no comparable pattern emerges for entrants into services. Methodologically, we show that worker flows alone cannot account for changes in employment stocks: survey reweighting and sample entry and exit make quantitatively important contributions. A matched-sample approach that restricts attention to respondents observed in consecutive waves and omits these margins can generate substantial cumulative discrepancies. In the United States and Germany, this approach understates the service employment share by 5.5 and 4.2 percentage points, respectively, by 2019.