Keeping Up With the Novaks? Income Distribution as a Determinant of Household Debt in CESEE

• Published date: 01 April 2022
• Author: Mariya Hake,Philipp Poyntner
• Date: 01 April 2022
• DOI: http://doi.org/10.1111/roiw.12540

© 2021 The Authors. Review of Income and Wealth published by John Wiley & Sons Ltd on behalf of

International Association for Research in Income and Wealth

S224

KEEPING UP WITH THE NOVAKS? INCOME DISTRIBUTION AS A

DETERMINANT OF HOUSEHOLD DEBT IN CESEE

by Mariya Hake

International Finance Corporation (IFC), Oesterreichische Nationalbank (OeNB)

AND

PHiliPP Poyntner*

Vienna University of Economics and Business

This paper constitutes an initial attempt to shed light on the role of income distribution for household

debt in Central, Eastern, and Southeastern Europe (CESEE). Using household- level data from the

OeNB’s Euro Survey for the period 2008– 2018, we address the question whether interpersonal com-

parisons (“Keeping up with the Novaks”) are associated with the probability of having a loan and

planning to take out a loan. Applying multilevel probit modeling to consider the hierarchical structure

of the data, our results support the notion that higher income inequality is negatively correlated with

the probability of having a loan at the bottom of the distribution, and positively at the top. We show

this impact for almost all components of household debt, but evidence is strongest for mortgage and

foreign currency loans. Loan plans are associated with income inequality at the very top of the income

distribution.

JEL Codes: D1, D3, G5

Keywords: CESEE, household loans, income distribution, multilevel models, relative income

1. introduction

The global financial crisis that started in 2008 has increasingly drawn atten-

tion to the importance of, and the threats arising from, household sector debt for

macroeconomic stability and GDP growth. Therefore, policymakers and research-

ers alike have turned their attention to the factors driving household indebted-

ness. Our analysis constitutes a comprehensive endeavor to relate debt and income

Note: The authors would like to thank Christian A. Belabed, Susanne Höfler, Stefan Humer,

Martin Šuster, Anna K. Raggl, Aleksandra Riedl, Kilian Rieder, Julia Wörz, Zuzana Fungácová, sem-

inar participants at the Bank of Albania, BOFIT, the Vienna University of Economics and Business,

NOeG, the Biannual Conference of the Economics Section of the German Association for East

European Studies, the IARIW- HSE Special conference “Experiences and Challenges in Measuring

Income and Wealth in Eastern Europe and CIS Countries,” and two anonymous referees for helpful

comments and valuable suggestions. The views expressed are strictly those of the authors and do in no

way commit the Oesterreichische Nationalbank (OeNB), the International Finance Corporation (IFC),

the World Bank Group or the Eurosystem.

*Correspondence to: Philipp Poyntner, Vienna University of Economics and Business,

Welthandelsplatz 1, A-1020 Vienna, Austria (philipp.poyntner@wu.ac.at).

Review of Income and Wealth

Series 68, Number S1, April 2022

DOI: 10.1111/roiw.12540

This is an open access article under the terms of the Creative Commons Attribution License, which

permits use, distribution and reproduction in any medium, provided the original work is properly

cited.

bs_bs_banner

Review of Income and Wealth, Series 68, Number S1, April 2022

S225

© 2021 The Authors. Review of Income and Wealth published by John Wiley & Sons Ltd on behalf of

International Association for Research in Income and Wealth

distribution for households in the countries of Central, Eastern, and Southeastern

Europe (CESEE) for the period 2008– 2018.

What is the relation between household debt and income inequality? Two

hypotheses have been forwarded in the theoretical literature. First, positional

income concerns have been recognized as a factor influencing consumption since

Veblen (1899) and Duesenberry (1949), a phenomenon referred to as “Keeping

up with the Joneses.” As debt can be used to finance consumption, higher income

inequality could lead to an increase in consumption- related debt for households

who see their reference households’ income increase. Therefore, we investigate if

a “Keeping up with the Novaks” effect - as we fittingly rename the “Keeping up

with the Joneses” effect - can be observed in CESEE. Second, income inequality

could be a “signaling” factor indicating the creditworthiness of borrowers to lend-

ers (i.e., banks). Accordingly, ceteris paribus, with a more unequal income distri-

bution, lenders would become more risk- averse and tend to lend to more affluent

households rather than households on the lower end of the distribution (Coibion

et al., 2014).

A limited number of papers have addressed the empirical link between

household debt and income distribution, focusing primarily on advanced econo-

mies (e.g., Iacoviello, 2008; Coibion et al., 2014; Kumhof et al., 2015 for the US,

Loschiavo, 2016 for Italy, Brown et al., 2016 for the United Kingdom). Two very

recent papers are closely related to ours. Jestl (2019) suggests a positive impact of

income inequality on consumption- related household indebtedness in a sample of

EU countries, but not for the CESEE countries as included in our paper. In addi-

tion, Hake and Poyntner (2019) constitutes an initial analysis of the correlation

between the probability of being indebted and income inequality according to the

income position.

Notably, the analysis of the household indebtedness- income inequality nexus

for European Emerging Economies is an interesting case as this region has shown

diverging developments as compared to advanced economies. On one hand, the

levels of household debt in CESEE have remained below the levels experienced

in other parts of the world (e.g., the US, OECD, Euro area countries). On the

other hand, the increase in household indebtedness has been coupled with a rela-

tively high level of income inequality, with the Gini coefficient close to 0.5 in some

countries.

Against that background, our paper is one of the first to explore the link

between indebtedness and income inequality in CESEE. We use data from a house-

hold survey performed in ten CESEE countries in the period 2008- 2018.1 The data

offer important advantages as compared to other data sets and the studies men-

tioned so far. First, they are comparable across a relatively large set of countries

and thus alleviate concerns about biases coming from different sampling methods.

Second, our data set encompasses 11 years in a repeated cross- section (and not a

1The CESEE country aggregate in this paper includes EU member states (Bulgaria, Croatia, the

Czech Republic, Hungary, Poland, and Romania) and (potential) EU candidate countries (Albania,

Bosnia and Herzegovina, North Macedonia, and Serbia).

Review of Income and Wealth, Series 68, Number S1, April 2022

S226

© 2021 The Authors. Review of Income and Wealth published by John Wiley & Sons Ltd on behalf of

International Association for Research in Income and Wealth

panel)2 at the individual level including both the pre- crisis and post- crisis period

making our results more robust to time variation concerns. Additionally, our focus

on loan intentions allows us to disentangle demand and supply factors.

In our preferred specification, we apply an upward- looking measure of income

inequality, the relative reference income, which gives the average income of richer

households as compared to the household’s own income (in line with Drechsel-

Grau et al., 2014). We apply multilevel methodology to account for the correlation

of responses from individuals from the same region. Finally, we provide a concise

test to assess which are the most suitable reference groups also including a spatial

aspect.

To foreground our findings, a higher relative reference income3 is negatively

correlated with households’ likelihood of having a loan for households at the bot-

tom of the income distribution. This is consistent with the signaling channel

prompting banks to redirect their loans to richer households. For households at the

top of the distribution, the relation between the relative reference income and debt

is positive, consistent with both the signaling channel and the “Keeping up with the

Novaks” channel. We show this impact for almost all components of household

debt, but evidence is strongest for mortgage and foreign currency loans. The mar-

ginal effect of an increase in the relative reference income by one unit4 is about 2

percentage points for the top of the decile. The sign and magnitude of this effect

are robust to different definitions of comparison groups. The results for the bottom

of the distribution are driven by low inequality regions, whereas the results for the

top of the distribution are driven by high inequality regions, hinting at a threshold

effect. These results are robust to various alternative income inequality measures

such as the Gini coefficient, income percentile ratios, or income shares. In addition,

the analysis explores loan intentions and thus tries to disentangle, to the extent

possible, demand and supply factors for household indebtedness. We find support

for a positive correlation between loan intentions and income inequality at the top

of the income distribution, supporting evidence for the “Keeping up with the

Novaks” effect.

The rest of the paper is structured as follows. In Section 2, we start with a brief

discussion on the theoretical and empirical literature dealing with the influence of

the income distribution on household debt. The description of the data and the

methodological setup follow in Sections 3 and 4, respectively. In Section 5, we turn

our attention to an analysis with respect to loan purpose and currency denomina-

tion. Section 6 zooms in on the impact on households’ loan intentions. Sections 7

and 8 focus on alternative measures of income inequality and test the suitability

of reference groups on the regional level, respectively. The last section concludes.

2The data could only be used as panel data if data are aggregated at regional or country level.

Because this would result in a significant loss of information, we chose to work with the microdata at

individual level and pool the data.

3The relative reference income is defined as the average income of richer households relative to the

household’s own income. A high relative reference income therefore indicates that the household’s ref-

erence group income is large relative to that household’s own income. Detailed information on this

variable can be found in Section 3.

4The median of the relative reference income is 2.6 in our sample.