Housing wealth, financial wealth, and consumption: New evidence from micro data

https://doi.org/10.1016/j.regsciurbeco.2008.06.002Get rights and content

Abstract

Fluctuations in the stock market and in house values over the course of recent years have led to renewed macroeconomic policy debate as regards the effects of financial and housing wealth in the determination of consumer spending. This research assembles a unique matched sample of household data from the Survey of Consumer Finance and the Consumer Expenditure Survey to estimate the consumption effects of financial and housing wealth. The micro-data permit numerous innovations in the assessment of wealth effects, including an analysis of the impact of wealth on both durable and non-durable consumption and a comparison of wealth effects as derive from gross versus after-debt measures of financial and housing wealth. Further, the research seeks to assess robustness of those estimates to deviations from trend and volatility in financial and housing wealth and among credit constrained and non-credit constrained households.

Overall, research findings indicate relatively large housing wealth effects. Among homeowners, the housing wealth elasticities are estimated in the range of .06 over the 1989–2001 period. In marked contrast, the estimated elasticities of consumption spending with respect to financial wealth are smaller in magnitude and are in the range of .02. Further, the estimated wealth elasticities appear robust to deviations from trend and volatility in the wealth measures. Research findings support the hypothesized behavioral distinction in household consumption spending across durable versus non-durable categories. Consumption propensities also diverge sharply across the credit constrained and non-credit constrained households. Finally, there is little difference in wealth elasticities derived from measures of home equity versus house values.

Research findings suggest the possibility of sizable reverse wealth effects. For example, a 10 percent decline in housing wealth from 2005 levels translates into a 1 percentage point decline in real GDP growth, a sizable reduction relative to the approximate 4 percent real GDP growth evidenced in prior years. Results of the analysis point to the sizable economy-wide risks associated with the recent retrenchment in house values.

Introduction

Recent years have witnessed widespread media attention and economic policy debate regarding the consumption effects of fluctuations in household financial and housing wealth. As is well-appreciated, stock prices evidenced pronounced volatility over the course of the 1990s, running up by 450 percent before falling back by a full one-third during 2000–2001. The stock market collapse destroyed more than $8 trillion in paper wealth and was arguably a cause of the 2001 recession. In contrast, U.S. house prices approximately doubled over the decade of the 1990s and then doubled again during 2000–2005 in many metropolitan areas. In 2005, those gains were widespread as 25 U.S. states recorded double-digit house price increases. Indeed, home equity grew by about $9.6 trillion during 2001–2004 to comprise more than one-half of the wealth of the typical U.S. household (Belsky & Prakken, 2004).1 In a recent paper, Greenspan and Kennedy (2005) estimated home equity extraction at $383 billion in 2001 and $552 billion in 2002, of which $174 and $214 billion, respectively, consisted of gross cash out refinance activity. According to Greenspan and Kennedy (2005), homeowners extracted an additional $300 billion in home equity through cash-out refinancings in 2003. The refinance boom of recent years was supported by generational lows in mortgage interest rates and innovations in financial and mortgage markets that enabled households to access their wealth in cheaper, faster ways.2 More recently, in the wake of the 2006–2007 bursting of speculative bubbles both in housing and in the capital markets, house prices have recorded substantial declines. Similar retrenchment was evidenced in mortgage re-finance activity, in the wake of marked reductions in home equity and in related withdrawal or re-pricing of home equity lines of credit and related mortgage products. Those dramatic trends led analysts at the Federal Reserve and on Wall St. to ascribe a critical role to housing wealth in the determination of cyclical swings in consumption activity.3,4

A well developed literature in finance has established a link between consumption and wealth shocks (e.g., Poterba and Samwick, 1996, Juster et al., 1999). These models predict that unexpected wealth shocks change the permanent income of households and thereby affect the life-cycle pattern of savings and consumption (Lettau and Ludvigson, 2004). A companion literature has argued that shocks to different forms of wealth can elicit varying consumption responses (e.g., Iacoviello, 2004, Lettau and Ludvigson, 2004, Piazzesi et al., 2004, Case et al., 2005, Lustig and Van Nieuwerburgh, 2005) and empirical studies have generally borne this out (e.g., Case et al., 2005, Benjamin et al., 2004).

This research assembles a unique matched data sample from the Survey of Consumer Finance and the Consumer Expenditure Survey to estimate the consumption effects associated with real estate and financial wealth. The highly-detailed micro data enable us to shed new light on household consumption behavior in several important ways. Specifically, we assess household responses among different categories of consumption spending and to various components of financial and real estate wealth. Further, the research evaluates variability in consumption spending to changes in the market value of household asset holdings, as is customary in the empirical literature, and to changes in wealth net of debt, as is consistent with theory. The analysis also examines household responses over time and in response to volatility and trend deviations in the underlying wealth measures, so as to assess the robustness of the estimated elasticities to the marked fluctuations in stock market and real estate valuations evidenced over the 1989–2001 period. Additional estimates are presented, including those pertaining to the robustness of wealth estimates across households grouped by age and by credit constraint in consumer debt markets.

The research proceeds as follows. The next section provides background and a review of relevant literature. The dataset and empirical specifications are described in Section 3. Section 4 presents the statistical results, and Section 5 discusses implications of statistical findings for macroeconomic activity.

Section snippets

Background and literature review

Recent literature has sought to nuance our understanding of the link between consumer behavior and shocks to household wealth. In that regard, Lettau and Ludvigson (2004) stress that unexpected wealth shocks must be perceived as permanent to affect consumption spending. The authors present evidence that households do not respond to transitory shocks by adjusting consumption patterns.5

Data and model

As noted above, our research relies on a dataset that was expressly developed so as to allow appropriately nuanced specification of the wealth-related hypotheses. That dataset links detailed individual-level consumption information with similar quality wealth data and accordingly is substantially better suited to the questions at hand than the data used in prior studies. The data are drawn from two surveys. The U.S. Bureau of Labor Statistics' Consumer Expenditure Survey (CEX) has since 1980

Results

The estimated income, financial wealth, and housing elasticities as derived from the cross-sectional models (Eq. (2)) are displayed in Table 3. As suggested above, that specification estimates consumption elasticities associated with the market value of real estate and financial assets. Table 4 specifies the estimating equations in terms of net wealth measures (e.g., asset values net of mortgage or other debt as described in Eq. (3), above). The estimates are computed for each of the 1989,

Supplemental findings

In supplemental analyses, we also investigated the effects on household spending of (a) quantity constraints on credit extensions and (b) borrower credit quality. As regards the former, the theoretical literature suggests that household borrowing capacity, which can be influenced both by a household's credit rating and level of outstanding debt, should play a role in shaping how changes in different forms of wealth affect consumption (Iacoviello, 2004, Piazzesi et al., 2004). If borrowing

Conclusion

This research assembled a unique matched data set from individual files of the Survey of Consumer Finances and the Consumer Expenditure Survey to estimate the consumption effects associated with housing and financial wealth. Estimates are provided for all survey years of the Survey of Consumer Finances from 1989–2001, so as to assess any significant drift in estimated elasticities as might derive from the larger business cycle, evolution in mortgage finance and the like. Further, year-specific

References (34)

  • CannerGlenn et al.
  • CarrollC.

    Housing wealth and consumption expenditure

  • CaseKarl E.

    The real estate cycle and the regional economy: the consequences of the Massachusetts boom of 1984–1987

    Urban Studies

    (1992)
  • CaseKarl E. et al.

    Comparing wealth effects: the stock market versus the housing market

  • ChoiJames J. et al.

    Consumption-wealth comovement of the wrong sign

  • DvornakN. et al.

    Housing wealth, stock market wealth and consumption: a panel analysis for Australia

  • DynanK. et al.

    Does the stock market wealth matter for consumption?

  • Cited by (231)

    View all citing articles on Scopus

    The authors thank Dolores Conway, Yuming Fu, Fred Furlong, Richard Green, Don Haurin, Selo Imrohoroglu, John Krainer, Edward Leamer, Erica Morris, Albert Saiz, Mark Spiegal, Selale Tuzal and Jim Wilcox for comments on an earlier draft of this paper. We also are appreciative of feedback from the editor. We owe special thanks to Jim Kennedy of the Federal Reserve Board for discussion and inputs to the analysis. The authors also thank Jane Lincove and Abishek Mamgain for excellent research assistance.

    View full text