About Me
I am a sixth-year Ph.D. student in the Economics Department at Cornell University. I obtained a BA in mathematics and economics from Cornell in 2021.
I am on the 2026–2027 academic job market.
I work on macroeconomics, environmental economics and firm dynamics. My research focuses on three themes: (1) firms’ technology choices, corporate investment, and their aggregate implications; (2) the micro and macro effects of climate change and emissions regulation; and (3) empirical and computational methods for heterogeneous-agent and spatial economics. I use microdata to answer macroeconomic questions.
My full CV can be found here. You can contact me at zx88@cornell.edu.
Working Papers
The (Mis)Allocation Channel of Climate Change: Evidence from Global Firm-level Microdata
(Job Market Paper)
Abstract: The (Mis)Allocation Channel of Climate Change: Evidence from Global Firm-level Microdata
Macroeconomic estimates imply large GDP losses from climate change, yet estimated firm-level productivity losses are often modest. We show that this gap can be explained by extreme temperatures increasing capital misallocation across firms. Using global firm-level microdata, we find that a single hot day above 30°C increases the dispersion of capital returns by 0.29 log points. This implies an annual aggregate TFP loss of 0.10%, or more than a third of one day of production. The effect is stronger in hotter and more developed regions, where adaptation is expected to be more prevalent. A multi-region firm-dynamics model captures these patterns. Extreme temperatures cause uneven and unpredictable disruptions across firms, generating dispersed capital returns as capital adjusts slowly. Adaptation reduces firms’ average productivity losses but can increase the dispersion of capital returns during extreme heat if firms with more favorable productivity surprises adapt more following investment. The estimated model implies that climate change since 1981 has lowered annual global TFP by 5% through misallocation and points to the importance of differentiated adaptation subsidies.
Endogenous Returns to Scale
Abstract: Endogenous Returns to Scale
We develop a model in which firms choose the returns to scale of their production technologies. More scalable technologies facilitate output expansion but are less effective at small scale. In equilibrium, more productive firms adopt more scalable technologies and grow disproportionately large, thickening the tail of the size distribution. Reallocation of resources toward these firms raises GDP. When aggregate productivity grows, firms adopt more scalable technologies, lowering their customers’ input prices and inducing further adoption downstream. Through this supply-chain amplification, endogenous returns to scale also raise GDP growth. A calibration shows sizable effects, and firm-level data support the model’s predictions.
The Origin of Risk
Accepted at Quarterly Journal of Economics
Abstract: The Origin of Risk
We propose a tractable model in which risk, at both the micro and macro levels, is endogenous and driven by incentives. In the model, each firm chooses the mean and the variance of its productivity process, as well as how it covaries with the productivity of other firms. Aggregate risk arises when firms select productivity processes that are correlated with one another. The theory predicts that larger firms and those with lower markups are less volatile and less correlated with aggregate productivity. We find support for these predictions in the data. Through their impact on risk-taking decisions, distortions such as taxes and markups can make GDP more volatile in equilibrium. In a calibrated version of the model, removing distortions significantly reduces GDP volatility.
The Distributional and Aggregate Effects of Factor-biased Climate Change
Abstract: The Distributional and Aggregate Effects of Factor-biased Climate Change
The distributional and aggregate effects of climate change depend on which factors of production climate change affects and the degree of substitutability between factor inputs. Using firm-level data from 32 countries in Europe and Asia, we find substantial factor bias in climate change. Extreme heat drives firms to shift expenditures from labor to capital, reducing the labor share of income. At aggregate and sectoral levels, this effect is driven by a combination of capital-labor complementarity, and climate change being biased toward damaging capital-augmenting productivity more than it damages labor-augmenting productivity. We use our estimates on the effects on labor- and capital-augmenting productivity to unpack the role of factor-biased climate change in aggregate income and the decline labor share of income over the last forty years. Factor-biased climate change has reduced aggregate global income by 2% since 1981 and explains 10% of the decline in the global labor share. Conventional approaches that assume a unitary elasticity of substitution underestimate aggregate effects by up to 90% and find no effect of climate change on the labor share. Ignoring factor complementarity and biased impacts result in underestimating the negative effects of climate change.
Space-time Jacobian: A Spectral Approach to Dynamic Economic Geography
(Draft Coming Soon)
Abstract: Space-time Jacobian: A Spectral Approach to Dynamic Economic Geography
We propose a general and fast method for solving and interpreting large-scale dynamic economic geography models with forward-looking migration, trade and investment. Our approach exploits the spatial spectral structure of space–time Jacobians, which describe how changes in endogenous variables affect equilibrium conditions across locations and dates. The spatial spectrum links the strength of spatial interactions to the persistence of equilibrium responses, showing how local adjustment can be rapid while differences across distant regions persist. For dynamic economies that share the same spatial patterns, we provide a decomposition that diagonalizes the linearized equilibrium into independent dynamic systems. The resulting decomposition analytically characterizes adjustment, long-run effects, and the conditions for a unique bounded transition. For general gravity-type economies with heterogeneous geography, a generic Fourier spatial spectrum with an iterative method allows for fast solutions. When locations have heterogeneous households, we derive a factorization that allows derivatives over assets and income to be reused across these patterns. We apply our methods to evaluate the dynamic extension of current spatial models of migration, trade, capital accumulation, plant investment, and climate change, with several thousand locations. Lastly, we extend our fast computational approach to second-order responses, nonlinear perfect-foresight transitions, and precautionary responses under risks in a model of climate change. Our method connects fast computation of spatial transitions to their economic interpretation.
Misallocation with Capital Heterogeneity
Abstract: Misallocation with Capital Heterogeneity
We argue that accounting for capital heterogeneity is important for understanding the sources and costs of misallocation. We prove that, conditional on the same observables, further disaggregation of capital types will always lead to a higher measured cost of misallocation. Quantitatively, accounting for capital heterogeneity increases measured costs of misallocation by 7 p.p. (19%) in the U.S. and 6 p.p. (24%) in India. Across countries, structures are consistently more misallocated than equipment. We then estimate a dynamic model to disentangle sources of misallocation that can explain the additional measured misallocation and the efficiency differences between equipment and structures. Results indicate that adjustment costs and imperfect information cannot fully explain the additional misallocation or why structures are more misallocated. Heterogeneous financial constraints and tax policies may contribute to the higher misallocation of structures, while heterogeneous technology and measurement errors play only modest roles.
Publications and Forthcoming Articles
Climate Change and Long-Run Factor Shares
AEA Papers and Proceedings, vol. 116, May 2026, pp. 510–516.
Works in Progress (in presentation mode)
How Emissions Scale: Theory and Evidence from EU ETS
Abstract: How Emissions Scale: Theory and Evidence from EU ETS
How do industrial emissions scale with production under carbon regulation? Using verified emissions from the EU Emissions Trading System linked to firm balance-sheet data, we document an endogenous emission scalability that declines with firm size and carbon prices. A 1 percent increase in real sales raises emissions by 0.52 percent in the smallest firm-size decile and 0.34 percent in the largest, while a one-standard-deviation increase in log carbon prices reduces emission scalability by 0.031. We develop a model of clean technology choice to explain these patterns. Production uses labor and fuel, and firms choose both the number of cleaning methods they install and the intensity with which they operate them. Fixed costs of installing additional methods make adoption more attractive at larger scales, allowing firms to lower emission scalability as they grow. Estimating the model on micro evidence, we find that the endogenous emission scalability mechanism accounts for 98 percent of the model-implied decline in emissions, with little contribution from output contraction. Under projected ETS prices, the EU can achieve near-zero emissions by 2050 at a cost of an 8.8 percent decline in real consumption.
Estimating Aggregate Elasticities in Heterogeneous-Agent Macroeconomics