top of page

Working Papers

Joint work with Junjie Xia

Job market paper

This paper quantifies the impact of input distortions on the labor and capital shares in the aggregate revenue of Chinese industry with heterogeneous productivity, technology, and demand elasticities. The input distortions include anything that prevents firms from using production factors at their market prices or from freely adjusting the usage. Using parameters estimated from firm-level data, we find that removing the input distortions in Chinese industry would raise the industrial labor share by 20 percentage points, and lower the industrial capital share by 1 percentage point. This implies that improving allocation efficiency through removing input distortions can create socioeconomic shocks such as a massive inflow of labor to industry. The distributional impact of input distortions also suggests that improving the allocation efficiency of inputs may not always be the solution to close the gap between developing and developed countries.

The paper is available here and the slides are here.

Joint work with Junjie Xia

Predicted total factor productivity (TFP) gains in China from removing the input distortions under Hsieh and Klenow (2009)'s framework are, in theory, sensitive to the assumption of constant returns to scale and demand elasticity being 3 for all the firms, and that Chinese firms have the same technology as American firms. However, there is little empirical evidence on how the predicted TFP gains would change if these assumptions are relaxed. Using the framework developed by Zhang and Xia (2023), we find that Chinese firms have constant returns to scale on average, that the average demand elasticity is around 8 with large dispersion, and that American firms are more capital intensive than Chinese firms. The large variation in demand elasticities has a minor impact on predicted TFP gains, but using the estimated average demand elasticity implies more than tripled predicted TFP gains. The technological differences between Chinese and American firms do not affect the predicted TFP gains through the estimated input distortions but through aggregation across firms and through the inferred productivity. Our estimates show a 45% TFP gain from removing the input distortions in China in 2005.

Presented at Tilburg Structural Econometrics Group, the GSS seminar at Tilburg University, the IO seminar at KU Leuven, the ENTER Jamboree, China International Conference in Macroeconomics, the IAAE annual conference, the AMES China meeting, and China Center for Economic Research’s Summer Institute, the NSE International Annual Workshop, the EEA-ESEM summer meeting.

The paper is available here.

This paper estimates the impact of notched driving-range-based (DRB) subsidies to consumers on Chinese battery-electric-vehicle (BEV) manufacturers' incentives to reduce their production costs of driving ranges. Chinese consumers received generous subsidies if the driving ranges were above certain thresholds. Using a dynamic structural model to infer unobserved investment decisions on the cost reduction of driving ranges by manufacturers, I find that the discontinuous incentives around the range thresholds of Chinese DRB subsidies increased Chinese BEV manufacturers' probability of investing in reducing production costs of driving ranges in 2019 by up to 50 percentage points. Compared with counterfactual DRB subsidies that are linear in driving ranges and provide the same total amount of subsidy in 2019, the notched scheme induces lower investment probabilities but is a lot cheaper in future periods for the government. This dynamic impact on production costs implies that the environmental benefits and welfare gains of notched DRB subsidies are very likely larger than the estimates of existing literature. It also implies that notched subsidies can be used to induce technological adoption or product upgrading.

Presented at the virtual IO CEPR event.

The paper is available here.

bottom of page