Working Paper: CEPR ID: DP16534
Authors: Jaedo Choi; Andrei Levchenko
Abstract: This paper provides causal evidence on the impact of a large-scale industrial policy -- South Korea's Heavy and Chemical Industry (HCI) Drive -- on firms' long-term performance and quantifies its long-term welfare effects. Using unique historical data on the universe of firm-level subsidies and a natural experiment, we find large and persistent effects of this industrial policy. Subsidized firms grew faster than those never subsidized for 30 years after subsidies ended. Firms upstream from the subsidized firms also benefited from the policy. We build a quantitative heterogeneous firm model that rationalizes these effects through a combination of learning-by-doing and financial frictions. The model is calibrated to firm-level data, and its key parameters are disciplined with the econometric estimates. The HCI Drive generated larger benefits than costs. If it had not been implemented, South Korea's welfare would have been 10-17% lower, depending on how long-lived are the productivity benefits of learning-by-doing. The large majority of the total welfare impact comes from the long-term benefits of learning-by-doing rather than short-term benefits of relaxing financial constraints.
Keywords: Industrial Policy; South Korea; HCI Drive; Learning-by-Doing; Financial Frictions; Welfare
JEL Codes: O14; O25
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| subsidized firms (L11) | faster growth (O49) |
| doubling of subsidies (H23) | 12 percentage point increase in sales growth (L81) |
| doubling of subsidies (H23) | 0.4 percentage point increase in annual growth rate (O49) |
| subsidized firms (L11) | higher total factor productivity (TFP) (O49) |
| upstream firms supplying inputs (L14) | growth benefits (O40) |
| HCI drive (L63) | long-term productivity benefits (J24) |
| HCI drive (L63) | welfare loss of 10.17% if not implemented (D69) |