Inequality, debt, and bankruptcy: Evidence from insolvents' balance sheets

Following the 2008 crisis, a number of authors (e.g. Rajan, 2010; Acemoglu, 2011) have linked income inequality and debt to financial distress. The basic argument is that increasing inequality leads to higher debt levels of the poor, so that they can attempt to match the consumption of the rich, which in turn leads to financial distress (e.g. bankruptcy). These authors have linked income inequality and financial distress in an attempt to explain the well-known stylized fact from Piketty and Saez (2003 and updated data) that income inequality in the US, as measured by the share of income to the top 1% or top 0.1%, peaked in the periods before the 1929 and 2008 financial crises, which were associated with very high levels of personal debt.

In spite of the large amount of discussion on the links between inequality, debt and financial distress, individual micro based empirical evidence on these relationships are very rare. This paper provides new individual level evidence on the cross sectional impact of income inequality on debt and bankruptcy. We use a new and unique data base containing detailed balance sheet data of essentially every personal bankruptcy filing in Canada from 2005 to 2010. These individual balance sheet data include the dollar amounts of all debts of each bankruptcy filer, including all mortgages, credit cards, secured and unsecured debts. These data were provided to us by Canadian Bankruptcy regulator, the Office of the Superintendent of Bankruptcy (OSB). There are almost half a million individual bankruptcy filings in the database.

Our main finding is that increased income inequality is indeed associated with higher levels of debt across bankruptcy filers, and in particular higher levels of unsecured and credit card debt. One way to interpret our results is that bankrupts living in high inequality areas will have had greater pressures to increase their debts before bankruptcy in order to match their consumption with that of their richer neighbors. Thus, these individuals are likely to have relatively higher levels of debt when they file for bankruptcy. Alternatively, bankrupts living in low inequality areas will have had fewer pressures to increase debt to match their neighbors' consumption, because their neighbors would not be that much richer. Thus, these individuals will be more likely to declare bankruptcy for reasons other than excessive debt and will have lower levels of debt when they file for bankruptcy.

In addition, because we are able to distinguish between mortgage debt and credit card debt of every bankruptcy filer, we can also examine the relative importance of income inequality in explaining these different kinds of debt across bankruptcy filers. We find that while income inequality impacts both the amounts of mortgage and credit card debt of bankruptcy filers, only 23.5 % of bankruptcy filers had mortgage debt outstanding compared to 89 % of all bankruptcy filers having credit card debt outstanding. Thus, in the Canadian context, we can show that credit card debt is a far more widespread factor in bankruptcy compared to mortgage debt.