Inequality, leverage and crises: Theory and evidence revisited

• Author: Yang Zhang,Chun Kwok Lei,Pui Sun Tam,Xinhua Gu
• DOI: http://doi.org/10.1111/twec.12806
• Date: 01 August 2019
• Published date: 01 August 2019

2280

|

wileyonlinelibrary.com/journal/twec World Econ. 2019;42:2280–2299.

© 2019 John Wiley & Sons Ltd

Received: 28 October 2018

|

Revised: 20 February 2019

|

Accepted: 10 April 2019

DOI: 10.1111/twec.12806

ORIGINAL ARTICLE

Inequality, leverage and crises: Theory and evidence

revisited

XinhuaGu

|

Pui SunTam

|

YangZhang

|

Chun KwokLei

Department of Finance and Business Economics,Faculty of Business Administration,University of Macau, Macau, China

Funding information

Research Committee, University of Macau, Grant/Award Number: MYRG2017‐00215‐FBA and MYRG2018‐00140‐FBA

KEYWORDS

asset bubble, financial crisis, GMM regression, inequality, leverage, theoretical framework

JEL CLASSIFICATION

E20; E25; E51; N1

1

|

INTRODUCTION

The Great Recession has spurred a new wave of research on the causes of financial crises. A central

issue under discussion is whether high inequality can be adduced as a causal factor behind the major

crises starting in 1929 and 2007. Rajan (2010) argues that rising inequality creates political pressure

to maintain living standards of the voting public despite their stagnant incomes. Related policy mea-

sures are not reducing inequality levels but encouraging deficit spending via expanded credit. Rajan's

political‐economy argument is confirmed by other authors' observations (e.g., Tridico, 2012; Wisman,

2013). An inevitable consequence of consumer credit growth under rising inequality is the rapid accu-

mulation of private indebtedness (Iacoviello, 2008; Klein, 2015). When such indebtedness starts to be

perceived as unsustainable, a sudden shock such as asset bubble bursting can become the trigger for a

major crisis (Kumhof & Ranciere, 2010; Kumhof, Ranciere, & Winant, 2015). This line of argument

is referred to as the Rajan hypothesis in the literature.

This hypothesis simply reopens the question of the link between inequality and crises, for

the Rajan hypothesis had existed long before Rajan (2010) (van Treeck, 2014). Previously, many

authors pointed out that rising inequality and growing credit had an intimate bearing on ensuing

disasters (Eccles, 1951; Galbraith, 1954; Pollin, 1988; Palley, 1994; Olney, 1999; Frank, 2007;

Barba & Pivetti, 2009; Stiglitz, 2009; UNCE, 2009). Recently, more studies have emerged in

support of the Rajan hypothesis (Galbraith, 2012; IMF & ILO, 2010; Kumhof, Lebarz, Rancière,

Richter, & Throckmorton, 2012; Lucchino & Morelli, 2012; Palley, 2012; Reich, 2010). While

other analyses of crises point to the roles of financial market deregulation, easy monetary policy

|

2281

GU etal.

and current account imbalances (Keys, Mukherjee, Seru, & Vig, 2010; Obstfeld & Rogoff, 2009;

Taylor, 2009), the latest studies are still in line with the Rajan hypothesis (Gu & Huang, 2014;

Kirschenmann, Malinen, & Nyberg, 2016; Kumhof etal., 2015; Stockhammer, 2015). These stud-

ies assert that credit expansion, albeit able to stimulate aggregate demand under high inequality

for some time, would be ultimately unsustainable since it is only a stopgap policy for the problem

of income distribution.

The Rajan hypothesis, while sparking lively debate about the effect of inequality on crises, has

its critics. Rajan (2010) did not emphasise the explosion of wealth for the very top (5% or 1%) of the

income distribution. Top inequality makes it possible to have privileged access to political decisions

via lobbying activities and campaign donations, and this in turn further heightens distributional and

debt problems (Hacker & Pierson, 2010; Volscho & Kelly, 2012). Moreover, the concomitant rise in

income inequality and financial fragility may be due to coincidence rather than causality (Acemoglu,

2011; Krugman, 2010). Furthermore, it is doubtful whether the Rajan hypothesis, applied mainly

to the 2007 US crisis, can be applicable universally to other countries or other crises. Bordo and

Meissner (2012) question the generality of this hypothesis by estimating a long panel with many

cross‐sections. They conclude that financial crises are preceded by credit booms that, however, are

not closely related to rising inequality. Similarly, Atkinson and Morelli (2010) find that high inequal-

ity may not be a common factor causing systemic crises in different countries other than the United

States.

It is necessary to clarify the above controversy by deepening the related research. The Rajan hy-

pothesis actually rejects conventional consumption theories since they have no link of personal con-

sumption with permanent income inequality and propose no policy action for stimulating demand in

response to high inequality (van Treeck, 2014). There is empirical evidence suggesting that rising

inequality may be due to the permanent rather than transitory components of income (Kopczuk, Saez,

& Song, 2010). The lack of attention to inequality among theorists seems to be related both to low

inequality during the first three post‐war decades and their enthusiasm about increased credit avail-

ability as efficient markets’ response to the higher household demand for insurance against the greater

dispersion of transitory income (Greenspan, 1996; Krueger & Perri, 2003). Using permanent income

inequality and relative income hypothesis, Rajan (2010) calls for a renaissance of consumption theory

because inequality that has risen again since the 1980s can no longer be ignored. This serious macro-

economic risk should be incorporated into theoretical models as in our work by considering the effects

on financial systems of permanent inequality and income classes.

Only a few studies along this line appear in the theoretical literature. Kumhof etal. (2015)

complement the Rajan hypothesis with a DSGE model for the links between inequality, debt and

crises. Although they provide a novel approach to the issue, alternative theories are still needed to

deepen our understanding of the complex impacts of income inequality on financial fragility. While

Kumhof etal. make differences by allowing for heterogeneity across income groups compared with

others (Iacoviello, 2008; Krueger & Perri, 2006), they did not address policy interventions that

cause consumption inequality to deviate from income inequality; the link between both types of in-

equality is an important issue in the recent literature (Attanasio & Pistaferri, 2016). In this paper, we

develop a macroeconomic model for this link based on key stylised facts and take into account the

government's role in affecting the inequality–debt–crisis nexus; this role is empirically confirmed

by other authors (Bartels, 2008; Igan, Mishra, & Tressel, 2011). Our model is made analytically

tractable, with a closed‐form solution derived to yield intuitive comparative static results.

This paper also presents a renewed examination of empirical evidence based on our theoretical predic-

tions. The extant works have examined the potential determinants of financial crises by following tradi-

tional approaches in a large literature on credit cycle (Allen, Babus, & Carletti, 2009; Bordo & Meissner,