The response of household debt to income inequality shocks: a heterogeneous approach

ABSTRACT

This article explores the impact of changes in income inequality on household indebtedness using Pedroni's heterogeneous panel VAR. As a result, we find evidence in support of large cross-country heterogeneity in the responses of household leverage to income inequality shocks. We also find that such heterogeneity stems from differences in the strength of financial regulations and supervision.

KEYWORDS:

• Income inequality
• household debt
• heterogeneity
• structural panel VAR

JEL CLASSIFICATION:

• D31
• E21

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

¹ Rajan, in his book Fault Lines (2010), argued that in response to rising income inequality in the USA over the past three decades, policymakers facilitated access to low-income mortgage loans. The resulting credit boom resulted in a sharp surge in housing prices, thereby triggering the financial crisis of 2008.

² This methodology has several additional advantages in tackling this topic. This approach first helps in addressing omitted variable bias as alternative bivariate approaches may leave out important information such as income, wage, interest rates, etc. This is in particular an efficient methodology for resolving the scepticism that country-level data with such diversity in political and economic factors may yield biased results. Furthermore, it can isolate the causal relationship between the variables of main interest without any need to incorporate internal instrument.

³ The sample covers the following countries: Argentina, Australia, Austria, Belgium, Brazil, Canada, Chile, China, Colombia, the Czech Republic, Denmark, Finland, France, Germany, Greece, Hong Kong, Hungary, India, Indonesia, Ireland, Israel, Italy, Japan, Korea, Luxembourg, Malaysia, Mexico, the Netherlands, New Zealand, Norway, Poland, Portugal, Russia, Singapore, South Africa, Spain, Sweden, Switzerland, Thailand, Turkey, the United Kingdom and the United States.

⁴ The sample covers the following countries: Argentina, Australia, Brazil, Canada, China, Colombia, Denmark, Finland, France, Germany, India, Indonesia, Ireland, Italy, Japan, Korea, Malaysia, the Netherlands, New Zealand, Norway, Portugal, Russia, Singapore, South Africa, Spain, Sweden, Switzerland, the United Kingdom and the United States.

⁵ For a detailed description of this decomposition, see Pedroni (2013) and Góes (2016).

⁶ Actually, we find that our results are robust with respect to the ordering of the variables. The results will be available upon request.

⁷ Based on the results of several sorts of panel unit-root tests, we find that all the variables used in our estimations, i.e. Gini indices, income shares of top 1 per cent of earners and household-debt-to-GDP ratios, turn out to have unit roots. Therefore, we utilize the log-differenced values.

⁸ We must consider the limited number of observations due to cross-sectional nature of the dependent variables. The ordinary least-squares methodology can yield misleading estimates if the underlying assumptions are not true. It is particularly highly sensitive to the presence of outliers in estimations involving a small number of observations.