Estimating the Marginal Propensity to Consume Using the Distributions of Income, Consumption, and Wealth 
In contrast to most recent studies of economic inequality, which separately examine income, consumption, and wealth inequality, this paper, using Panel Study of Income Dynamics (PSID) data from 1999 through 2013, considers the relationship between the three factors to determine whether the effects of changes in income on consumption are more pronounced at higher or lower levels of wealth. The authors—the first to use the PSID to estimate the marginal propensity to consume (MPC) by wealth—find that the MPC is indeed lower at higher wealth quintiles, suggesting that lower-wealth households respond more to changes in income than do higher-wealth households. They also find that the overall MPC lies at the low end of the range examined in other research, but this discrepancy is to be expected, the authors conclude, because the PSID documents two-year changes in income and consumption, while other studies use changes over shorter periods of time.
Key Findings
- The marginal propensity to consume (MPC) is lower at the higher wealth quintiles. For low-wealth households, the MPC is 10 times larger than it is for wealthy households.
- Whereas Alan Krueger, in his 2012 Council of Economic Advisors address on the consequences of inequality, suggested that if an additional $1.1 trillion were earned by the bottom 99 percent of US households instead of the wealthiest 1 percent, annual aggregate consumption would be about $440 billion (5 percent) greater, the authors find that such a transfer would instead generate a little more than half that sum in additional consumption: $230 billion.
- Based on the two-year changes in income and consumption documented by the Panel Study of Income Dynamics, the overall MPC is about 10 percent.
Implications
Heterogeneity in the marginal propensity to consume (MPC) has substantial implications for government fiscal policy when it means, as this study finds, that aggregate consumption would be higher if income were transferred from high-wealth to low-wealth households. The conclusion here that the MPC is lower at higher wealth quintiles further suggests that low-wealth households cannot smooth consumption as much as high-wealth households do. Low-wealth households’ consumption therefore responds more to changes in income than does that of wealthy households, which smooth consumption so that even if income varies significantly over the life cycle, consumption is less variable from year to year. This suggests that an increase in income inequality increases wealth inequality while likely reducing aggregate consumption, with important implications for economic growth.
Abstract
Recent studies of economic inequality almost always separately examine income, consumption, and wealth inequality and, hence, miss the important synergy among the three measures explicit in the life-cycle budget constraint. Using Panel Study of Income Dynamics data from 1999 through 2013, we examine whether these changes are more dramatic at higher or lower levels of wealth and find that the marginal propensity to consume is lower at higher wealth quintiles. This suggests that low-wealth households cannot smooth consumption as much as other households do, which further implies that increasing wealth inequality likely reduces aggregate consumption and limits economic growth.