191. Brazil's trade opening and its toll on workers and crime

Episode Summary

Executive Summary: This episode examines how Brazil’s 1990s trade liberalization affected workers, regions, and crime. Economist Rafael Dix-Carneiro explains that tariff cuts hit some places for more than a decade, lowering employment and wages, pushing displaced workers into informality, and raising homicides temporarily through unemployment and local spillovers. The research suggests trade shocks can deepen regional inequality and create broader social costs beyond directly exposed industries.

Main Topics: Brazil’s pre-liberalization economy and shock therapy reforms (Priority: 5/5): The discussion sets Brazil’s historical backdrop: decades of protectionism, hyperinflation, and state-led industrial policy, followed by a sudden 1990 trade-opening package alongside privatization and fiscal reforms. Regional labor-market effects of trade liberalization (Priority: 5/5): Trade liberalization hit regions differently depending on sectoral specialization, producing long-lasting declines in employment and wages in places exposed to large tariff cuts. Why workers did not smoothly reallocate across sectors (Priority: 4/5): The episode explores why standard trade models’ expected movement from import-competing sectors to expanding sectors did not occur, and why workers remained stuck in affected regions. Informality as a fallback labor market (Priority: 4/5): Displaced workers often eventually found jobs only in the informal sector, highlighting informality as a buffer that may have softened unemployment but reduced job quality and protections. Trade shocks and crime (Priority: 5/5): Regions more exposed to import competition saw higher homicide rates for several years, with unemployment identified as the key channel linking economic distress to crime. Policy design and spillovers beyond directly affected workers (Priority: 4/5): The conversation argues that adjustment policies should target not just directly exposed workers, but also broader local communities affected through lower demand, productivity, and crime. Inequality and the distributional consequences of trade (Priority: 3/5): The research finds that trade reduced inequality across Brazilian regions because initially richer regions were harder hit by liberalization, though the overall national welfare effect is not identified here.

Key Arguments: Brazil’s 1990 trade liberalization was abrupt and substantial, with average tariffs around 31% and large cuts in previously protected industries such as autos, apparel, and rubber. Regions specialized in industries facing bigger tariff cuts experienced persistent employment losses lasting 10 to 15 years, rather than a quick adjustment. Wages in hard-hit regions fell in the short run and only stabilized much later at substantially lower levels. Standard explanations for delayed adjustment—slow import response, city decline, or export booms—did not fit the Brazilian evidence. Firms facing import competition adjusted gradually by reducing machinery replacement and shrinking labor demand over time, which depressed productivity and wages. Local agglomeration effects worked in reverse: as employment fell, knowledge spillovers and regional productivity also declined, reinforcing labor-market weakness. Trade-displaced workers often re-entered work only in informal jobs, showing the informal sector served as a fallback employment channel. Import exposure increased homicides for about five to seven years, with unemployment emerging as the main mediator rather than inequality, school quality, or police/public goods. Trade shocks created spillovers beyond directly affected workers, harming nearby service workers and communities. The study’s results are relative, not absolute: they show some places grew more slowly than others, not necessarily that trade reduced Brazil’s national output overall. Trade appears to have reduced regional inequality, accounting for about a quarter of the decline in inter-regional inequality between 2000 and 2010.

Data Points: Population of Brazil: over 200 million - Brazil is described as the largest country in Latin America. Average tariff in 1990: about 31% - Tariff level before the 1990 shock-therapy liberalization. Informal sector share of labor force: about 50% - Estimated size of Brazil’s informal labor market. Homicides in Brazil: about 50,000 per year - Used to illustrate the country’s high violence rates. Most violent cities in the world located in Brazil: 32 of the 50 - Context for the crime discussion. Employment effect duration: 10 to 15 years - Hard-hit regions’ employment remained depressed for a long period after liberalization. Wage effect duration: stabilized only after 15 years - Wages in affected regions stayed lower for many years before leveling off. Crime effect duration: 5 to 7 years - Higher homicide rates appeared after trade shocks but faded later. Inter-regional inequality contribution: 25% - Share of the decline in regional inequality between 2000 and 2010 attributed to trade liberalization. Timeframe of hyperinflation/fiscal instability: 1985 to 1990 with five finance ministers - Shows the unstable pre-reform macroeconomic environment.

Pivotal Quotes: "Brazil in the 80s was a country that was very closed" — Rafael Dix-Carneiro: Describing the protectionist economic environment in which he grew up. "employment in these regions that were hard-hit by improper competition goes down relative to other regions" — Rafael Dix-Carneiro: Summarizing the core employment finding from the regional trade-shock analysis. "One interesting thing that comes out of this research is that input competition not only affects workers directly exposed to the shock but it also affects other people who have nothing to do with it through social developments such as increases in crime" — Rafael Dix-Carneiro: Explaining the broader spillovers from trade shocks beyond directly affected industries.

Implications: Trade policy can have lasting local labor-market, inequality, and public-safety effects. Adjustment aid should target communities, not just directly displaced workers, and policymakers need better evidence on what kinds of labor-market support actually work.