Consumption Responses to a Large Shock to Financial Wealth: Evidence from Italy

Date01 April 2020
Published date01 April 2020
DOIhttp://doi.org/10.1111/sjoe.12339
Scand. J. of Economics 122(2), 762–789, 2020
DOI: 10.1111/sjoe.12339
Consumption Responses to a Large Shock to
Financial Wealth: Evidence from Italy*
Renata Bottazzi
University of Bologna, 40126 Bologna, Italy
renata.bottazzi@unibo.it
Serena Trucchi
CaFoscari Universityof Venice, 30121 Venice, Italy
serena.trucchi@unive.it
Matthew Wakefield
University of Bologna, 40126 Bologna, Italy
matthew.wakefield@unibo.it
Abstract
We estimate marginalpropensities to consume from wealth shocks. We exploit large asset-price
shocks in 2007–2008 and household-level panel data to implement instrumental variables.A fall
of one euro in risky financial wealth resulted in cuts to annual total (non-durable)consumption of
8.5–9 (5.5–5.7) cents, with small effects on food spending. Effectsseem stronger for lower-wealth
or indebted households, but significant responses from wealthier households and those without
mortgages are important for our baseline results. Counterfactuals indicate financial-wealth effects
were relatively important for consumption falls in Italy in 2007–2008.The estimated effects are
consistent with a simulated life-cycle model capturing the wealth shock.
Keywords: Household consumption; marginal propensity to consume; the Great Recession;
wealth effects
JEL classification:D12; D15; E21
I. Introduction
Household consumption spending is the largest single component of modern
market-based economies, often representing more than half of GDP.1
*Weare grateful for the following financial support: MIUR-PRIN-2010-11-2010T8XAXB 006;
MIUR-FIRB-2008-RBFR089QQC-003-J31J10000060001; EU Marie Sklodowska-Curie
agreement 655770. Wethank Orazio Attanasio, Richard Blundell, Tom Crossley, Maria Cristina
De Nardi, Richard Disney, Carl Emmerson, Ben Etheridge, Erik Hurst, Brian Kovak, Andrea
Neri, Luca Nunziata, Luigi Pistaferri, Davide Raggi, two excellent referees, and seminar
participants. R. Bottazzi and M. Wakefield are International Research Fellows of the Institute
for Fiscal Studies.
1OECD figures (OECD, 2018) for 2016 put this proportion at 53 percent in the Euro Area, 54
percent for Italy, and 67 percent for the United States (these figures are based on the series for
“household final consumption spending”).
C
The editors of The Scandinavian Journal of Economics 2018.
R. Bottazzi, S. Trucchi, and M. Wakefield 763
Therefore, it is of crucial interest and importance to understand the
determinants of consumption spending, and there is a very extensive
body of literature studying how spending responds to resource shocks and
specifically, as in this paper, to wealth shocks.
As is clear from excellent surveys by Poterba (2000) and Paiella (2009),
empirical evidence on the relationship between wealth and consumption
spending has come from both aggregate, time-series studies and studies
that exploit household-level data.2Carroll et al. (2011) discuss the potential
limitations of relying purely on aggregate data and also conditions under
which micro-data could be helpful. “Micro” studies have the potential to
add insight and to sharpen identification when detailed (and preferably
panel) data on household finances and consumption spending are available,
and when an exogenous source of variation in wealth can be found. Our
contribution to the literature comes from exploiting data with precisely these
features.
The data we use come from the Bank of Italy’s Survey on Household
Income and Wealth (SHIW), and contain rich information on the asset
holdings (values and ownership), consumption outcomes, and demographic
and economic characteristics of households. The survey sample is
designed to be representative of the Italian resident population and has
a panel component. This combination of characteristics is impressive
by international standards and makes our study of broad interest for
understanding the importance of wealth effects in consumption. Having
wealth and consumption in household-level panel data, together with a
potentially exogenous wealth shock, also facilitates identification through
instrumental variables (IV).
The wealth shock comes from the 2006–2010 period that we choose
as the object of our study. Italy’s FTSE MIB (Milano Italia Borsa) fell by
more than 60 percent between May 2007 and March 2009, with a large part
of this fall in the central months of 2008.3Households that held wealth in
stocks thus suffered a sudden, potentially large, and mostly unanticipated
shock to their financial wealth.4We use the idea that the 2008 shock to
asset values can provide a source of variation in wealth that is exogenous to
2Even studies based on aggregate time-series evidence havelong acknowledged that allowing for
different types of consumers has important implications when comparing models with the data
(see Campbell and Mankiw, 1989).
3The evolution of stock prices in Italy is documented in more detail in Figure B2 in our Online
Appendix. The figure also illustrates that the path of aggregate consumption closely shadowed
that of the stock market, and if anything slightly lagged the fall in stock prices.
4It is worth noting that, unlike in the United States and United Kingdom, house values in Italy
did not suffer large falls near the beginning of the Great Recession (Agenzia delle Entrate and
Associazione Bancaria Italiana, 2019), and this explains our emphasis on the effects of financial
wealth.
C
The editors of The Scandinavian Journal of Economics 2018.

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT