We develop a structural equilibrium model with heterogeneous firms that choose whether to operate in the formal or in the informal sector. The model features a rich institutional setting, where formal firms must comply with minimum wages, and are subject to firing costs as well as payroll and revenue taxes. However, taxes and labor market regulations are imperfectly enforced by the government, giving rise to incentives for some firms to be informal. Finally, the economy consists of tradable and non-tradable sectors that interact. Only formal firms producing tradable goods can export.

We estimate the model using multiple data sources, including matched employer-employee data from formal and informal firms and workers in Brazil, as well as several other sources of firm- and worker-level data such as household surveys, manufacturing and services censuses, and customs data. Then, we conduct a series of counterfactual experiments to better understand the impact of trade shocks on an economy with a large informal sector. While the focus of the present paper is on trade, we note that the framework we develop can be applied to study the effects of several other policies, such as changes in payroll taxes, value added taxes, minimum wages, and unemployment benefits, either individually or jointly.

Brazil, with its excellent sources of data for the formal and the informal sectors, provides an excellent setting for our work. Nearly two thirds of businesses and 40 percent of GDP are informal, and the labor regulations are both substantive and weakly enforced. Moreover, there is a clear definition of what constitutes informality: we define as informal workers those who do not hold a formal labor contract, clearly observable through the worker's booklet "carteira de trabalho." Informal firms are those not registered with the tax authorities, which means that they do not possess the tax identification number required for Brazilian firms "Cadastro Nacional de Pessoa Juridica" which we can also observe.

Our estimated model rationalizes several findings reported in the empirical literature, while yielding new insights. We find that trade openness, induced by a reduction in iceberg trade costs, leads to large declines in informality in the tradable sector, an effect robust to the initial level of trade costs, the magnitude of their decline, and the regulatory environment. On the other hand, we find that the effects of trade openness on informality in the non-tradable sector are more nuanced and context dependent. As a result, the overall effect of trade on informality is ambiguous, and generally small. This result is consistent with the casual observation that the informal sector has not substantially shrunk in middle-income economies despite the large-scale liberalization episodes these experienced in the 1980s and 1990s.

Further, we find that trade openness is associated with substantive increases in productivity and welfare. Importantly, we show that the productivity gains from trade are severely understated in the tradable sector if one leaves out the informal sector (as analyses of trade liberalization episodes typically do). For the non-tradable sector, our results point to a bias in the opposite direction, but in the aggregate, the tradable sector effect dominates. As a result, we conclude that the productivity gains from trade for the economy as a whole are understated in analyses focusing exclusively on the formal sector of the economy.

One of the main rationales for reducing informality is the wish to increase productivity. Our counterfactual analysis shows that indeed, reducing or eliminating informality through stricter monitoring and enforcement raises productivity. However, the productivity gains are achieved at the expense of employment and welfare. In contrast, trade liberalization achieves sizable productivity gains while raising aggregate welfare and seems therefore a superior way to increase productivity.

Our analysis also has implications for wage inequality. We find that the inclusion of the informal sector reverses predictions on the effects of trade on inequality that is driven by firm heterogeneity. Focusing on the formal sector alone (again, as most earlier analyses have done), we observe that trade liberalization contributes to a rise in wage inequality. However, the effect in the informal sector goes in the opposite direction, while the distance between average formal and informal wages decreases. As a result, trade liberalization reduces aggregate wage inequality driven by differences across firms.

Finally, our results lend strong support to the view that the informal sector serves as an "unemployment buffer" during bad times: in the case of negative aggregate shocks, unemployment increases by considerably more if informality is repressed. However, this "unemployment buffer" role of informality does not translate into a "welfare buffer." We find that, in the event of a negative economic shock, welfare declines by less with lower informality. This somewhat counterintuitive result is due to a positive selection effect arising from the exit of inefficient, informal firms in that case---in other words, from a strong "creative destruction" effect.

Last Modified: 01/25/2021
Modified by: Rafael Dix-Carneiro