Government Banks, Household Debt, and Economic Downturns: The Case of Brazil

BIS Working Papers  |  No 876  | 
11 August 2020



Higher debt increases during credit booms tend to be followed by relatively sharper crisis episodes. The literature established this empirical regularity at the country or region level. Using contract level information from the Brazilian Credit Registry matched with information on labor income from the Ministry of Labor, this paper explores the individual borrower level underpinnings of this phenomenon.


The paper shows that a higher increase in debt-to-income ratio at the individual level during the credit expansion period in Brazil predicts a lower consumption during the subsequent crisis. Furthermore, this effect is stronger when borrowers use credit types commonly presenting higher interest rates.


The later phase of the Brazilian household debt boom, occurring between 2011 and 2014, had government banks playing the lead roles. We provide evidence of that by comparing government and private bank loans to individuals who borrow from both. Also, public employees were more exposed to this credit expansion than private sector employees. Comparing these groups of workers, we isolate the increase in debt-to- income ratio caused by a credit supply shock and show that it resulted in a relative reduction of consumption. This happened mainly among the borrowers belonging to the lowest income quintile.



After the global financial crisis, government banks in Brazil boosted credit provision to households, generating a sharp increase in household debt which was followed by the most severe recession in recent Brazilian history in 2015-2016. Using a novel individual-level data set including matched credit registry and employer-employee information, we show that individuals with higher debt-to-income growth during the boom experienced lower subsequent credit card expenditure during the recession. To identify the credit-supply effect, we exploit individuals borrowing from both government-controlled and private banks. We show that, during the late stages of the boom period, government banks increased their lending more than private banks to the same individual. To study the effect of this credit supply shock on individual consumption, we exploit variation in the sector of employment of each borrower. Individuals employed by the public sector were disproportionately targeted by payroll loans offered by government banks and experienced larger decline in credit card spending during the subsequent recession.

JEL classification: D14, E21, G21, G28

Keywords: credit booms, household credit, payroll loans, credit card expenditure