Jingcheng Jiang

Can anti-corruption enforcement improve allocative efficiency when political connections distort industrial policy? Exploiting China's 2013 anti-corruption inspections as a plausibly exogenous shock, I show that enforcement reduced subsidies to hometown-connected firms, with effects concentrated entirely on opaquely described allocations while transparent subsidies remained unaffected. This pattern reveals politicians curtailing corrupt rather than legitimate transfers when monitoring intensifies. Reductions operate primarily through behavioral change among continuing leaders as hometown favoritism transforms from career asset to career liability, disrupting clientelistic equilibria. Despite lower subsidies, regional productivity increases as resources reallocate from politically connected to more productive firms, with gains concentrated among smaller enterprises previously crowded out by larger incumbents in politicians' hometowns. These findings demonstrate a pattern consistent with politicians optimizing under enforcement constraints: when detection costs rise, corrupt transfers decline, career incentives reverse, and efficiency improves as misallocated resources flow toward productive firms, illuminating both how political connections generate substantial misallocation and the institutional mechanisms that restore competitive dynamics.

Geographic proximity makes a large neighbor an attractive trading partner but brings political conflict closer. Since the Cold War, geopolitical fragmentation has concentrated among nearby countries, especially land neighbors, rather than along rival blocs. Using country-level bilateral trade data from 1960 to 2021, I find that a one standard deviation increase in political distance is associated with roughly 10 percent less trade between bordering countries. Political friction follows: divergence is associated with more conflicts, weaker cooperation, and more threats from large neighbors. Trade adjustment is asymmetric: across land borders, smaller countries cut imports from larger neighbors by about a fifth. Reductions occur within existing product categories and are larger among land neighbors sharing an official language, suggesting cultural closeness amplifies political friction. A case study of the Baltic states after Crimea supports this mechanism: they cut Russian resource purchases more sharply than other EU members. In a general-equilibrium model, a West versus China/Russia split yields global welfare losses of 0.02 percent, but 0.6 to 0.9 percent in Taiwan and 0.1 to 0.2 percent in the Baltics. Fragmentation redraws trade at the frontier, where smaller countries bear the costs of retreat.

Competition among subnational governments generates internal non-tariff trade barriers, fragmenting markets within nominally integrated economies. In China's decentralized system, where provincial authorities evaluate city leaders on relative economic performance, politicians face incentives to maximize locally appropriable gains while minimizing spillovers to rivals. This yardstick competition creates systematic distortions in the allocation of public resources. Exploiting exogenous variation in promotion incentives driven by the age of city party secretaries (1996–2018), I show that intensified within-province competition significantly reduces inter-city trade flows, dampens international exports, and restricts firm access to intermediate inputs. Under competitive pressure, politicians reallocate resources toward non-tradable sectors, particularly real estate, while underinvesting in connective infrastructure near jurisdictional borders. Individually rational responses to career incentives thus generate collectively suboptimal market fragmentation, showing that internal trade costs are endogenous to political institutions and that yardstick competition carries dynamic costs.

We study whether transient policy-induced labor supply shocks can permanently redirect technological change. China's zero-COVID mobility restrictions provide a natural experiment: between 2020 and 2022, hundreds of millions of migrants were blocked from urban manufacturing centers, generating a nearly 60% collapse in migrant labor supply. Exploiting cross-county variation in pre-pandemic migrant dependence, we find that more exposed counties experienced sharp post-COVID growth in robotics patenting, with no corresponding change in total patenting activity. We also construct a Bartik IV and find consistent results: uncertainty in origin regions disrupted labor flows to destinations, and counties losing more migrants saw larger increases in automation patents. The response is concentrated in tight local labor markets where native workers could not substitute for absent migrants, consistent with binding and recurrent quantity constraints rather than wage-driven substitution. Displaced migrants bear persistent earnings and employment losses. The results show that temporary restrictions on labor mobility can have lasting effects on the direction of innovation.

Better capital allocation can help economies sustain productivity growth, but correcting misallocation requires local politicians with both the capacity and the incentive to act. Using more than one million manufacturing firm-years matched to politician records in China, I find that misallocation declines as a politician's tenure lengthens, but only in cities that rely more heavily on their own tax base. Where local budgets depend on central transfers, longer tenure brings little improvement. To interpret these patterns, I develop a spatial model with internal trade and costly migration in which tenure builds a politician's capacity to reallocate capital, while fiscal incentives determine whether that capacity is used to reduce or to worsen misallocation. Counterfactuals show that strengthening fiscal incentives raises aggregate manufacturing TFP by 0.21 percent under observed rotation and by 0.46 percent under five-year terms, with trade transmitting the gains across provinces. The results suggest that government intervention supports productivity-led development only when local politicians have both the capacity and the fiscal incentive to improve allocation.

We document a new failure mode of trade-led development. Standard theory predicts that expansions in skill-intensive sectors raise returns to skill and induce human capital investment. Using a Bartik shift-share design that interacts predetermined regional cropping patterns with international commodity price deviations, we test this prediction in rural China after agricultural trade liberalization and find it fails. Cash crop booms, though concentrated in skill-intensive production, generate no rise in the skill premium and no schooling response. The operating mechanism is propagation failure: specialized skills and bundled inputs required for cash crop production block entry, so trade rents are captured entirely by incumbents and never spill into broader labor markets. Staple crop shocks, by contrast, raise returns across a wide range of households and depress schooling through a standard opportunity cost channel. When trade rents bypass rural labor markets, agricultural liberalization widens the human capital gap — not from unusual frictions, but from a structural feature of skill-intensive sectors.