research

This paper examines how openness to trade reshapes our understanding of the Food Problem—the persistent specialization of low-income countries (LICs) in low-labor-productivity agriculture. In a closed economy, low agricultural productivity and subsistence needs require a large agricultural workforce. Trade can relax this constraint through imports, while access to foreign markets can expand agricultural production. I develop and calibrate a quantitative trade model with non-homothetic preferences, fixed agricultural land, and input–output linkages. I infer sectoral total factor productivity (TFP) from trade flows, finding higher ratios of agricultural to non-agricultural TFP in LICs than in richer countries. Diminishing returns to labor reconcile this finding with low agricultural labor productivity. I estimate income and substitution elasticities to quantify how domestic expenditure responds to income and relative prices. Counterfactuals show that openness reinforces agricultural specialization while raising welfare. Agricultural TFP increases release labor under autarky but draw workers into agriculture under trade. Industrial TFP increases instead reduce agricultural employment under openness, with little change under autarky. Openness thus changes the forces driving labor out of agriculture and, with them, how we interpret the Food Problem.

We study how monetary policy and exchange rate regimes shape the adjustment to trade shocks in small open economies (SOEs). The analysis highlights different implications for SOEs under different currency regimes, sovereign risk, and trade exposure. We find evidence of incomplete labour market adjustment, with rising unemployment after negative external demand shocks in SOEs under both pegged and floating regimes, with stronger effects in countries facing high-risk premia. We build a dynamic multi-sector trade model with downward nominal wage rigidity (DNWR), intertemporal consumption–savings decisions, and incomplete markets. The model includes two SOE regimes: floating, with endogenous monetary policy, or pegged. Under a peg, a negative external demand shock generates unemployment in the presence of DNWR. Under a float, unemployment can also arise if the monetary policy response limits the nominal depreciation’s ability to restore full employment. Furthermore, a spike in sovereign risk premia can amplify this effect.

This paper examines how demand systems and modelling choices shape quantitative estimates of the gains from trade in gravity-based models. Many applications rely on homothetic preferences, which suppress income effects and sectoral expenditure reallocation. If globalisation raises real income and expenditure shares vary systematically with income, this assumption can distort welfare estimates and mischaracterise sectoral reallocation. Using a multi-sector structural gravity framework with input-output data, I examine how allowing for non-homothetic demand alters estimated gains across alternative production structures and market environments. Four results emerge. First, moving from a one-sector to a multi-sector environment can substantially change estimated gains. Second, tradable intermediates consistently amplify gains through production-network linkages. Third, non-homothetic demand lowers both the average level and the cross-country dispersion of gains, as rising real income shifts expenditure toward less trade-intensive sectors. Fourth, higher cross-sector substitutability attenuates gains under perfect competition but can amplify them under monopolistic competition with firm heterogeneity through entry and selection effects. These findings suggest caution when interpreting gains from trade under homothetic assumptions, especially in samples with large cross-country income differences.