Portrait of Kai Gu

Kai Gu

PhD Candidate, Department of Economics
University of Wisconsin–Madison

I study how place-based industrial policy reshapes the local economy around it. My job market paper evaluates China’s export processing zones with firm-level manufacturing-survey and customs data. I am on the 2026–27 job market.

Fields
Urban and Regional Economics; International Trade; Public Economics (Fiscal Federalism); Development Economics; Applied Econometrics
Advisors
Christopher Taber, Jesse Gregory
Contact
kgu22@wisc.edu

Job Market Paper

Enclaves without Engines: The Local Gains and Costs of China’s Export Processing Zones

China opened 52 export processing zones in 2001–07, customs-fenced enclaves meant to be engines of local development. With customs and manufacturing-survey records in 0.1-degree cells, 1998–2007, I compare cells at each distance from the first zone with never-exposed cells matched on region, port access and 1998 size. Near the zones, no growth separable from earlier trends and no spillover appears; entry of new domestic establishments falls in a step across productivity, while land prices do not rise.

A model in which parent development zones ration land by coarse signal cells, with the land multiplier as the price, values the excluded entry at 18.6 or 36.8 percent of applicants’ four-year value, depending on how the applicant pool changed. Scarcity accounts for 14 to 24 percent; an auction on information available at application would recover about two fifths. Output rose by 11 log points 40 to 100 km away inside the province.

Paper (PDF)

Working Papers

Who Bears the Cost of a Special Economic Zone? Incumbent Firms across Prefecture Borders in China

Compares incumbent manufacturers at the same distance from an EPZ on either side of the host prefecture’s border, using firm microdata geocoded to the village and a model of a commuting labor market with a home-biased host government. Host-side incumbents are shielded; across the border, incumbents 40–80 km away are 8.9 percentage points less likely to survive, and firms almost never relocate. The cost of a zone is exported down the regional wage gradient to jurisdictions that did not choose it.

Activity or Capability? Export Processing Zones and Portable Export Experience in China

Proposes a firm-level measure of export capability from customs data: whether experience helps a firm enter similar new markets. Processing experience in which the firm owns its imported materials carries about a third of the weight of ordinary exporting; experience processing consigned materials does not transfer. Five years after opening, zone exports equal 39 percent of nearby firms’ pre-opening exports, yet their portable experience changes by only 5 percent (s.e. 4).

Enclaves or Engines? Export Processing Zones and Local Production Linkages in China

Traces the backward-linkage chain with customs and manufacturing microdata, input–output tables, and stacked event studies. The domestic value-added ratio of fenced exports, an upper bound on local sourcing, was 0.27 in 2003; upstream supplier industries did not expand on average, and downstream processors did not respond when a 2005 reform opened sales across the fence. Physical concentration of activity did not become economic agglomeration.

Discovery Lags on Social Media Platforms

with Arseniy Stolyarov

Drafts of the working papers are available on request.

Teaching

Teaching assistant at UW–Madison, 2019–2025, across six courses from undergraduate to PhD level.

Awards