If someone handed you $10 right now, what would you do with it? Would you decide to spend it right away? Or would you stash it away? Or some combination of the two? The answer to that question would in part reveal your marginal propensity to consume, or MPC for short. This statistic goes a long way toward understanding the consumption behaviors of families. More or less, it tells us how much more spending happens when a household or individual gets more income or wealth. Economists have long debated what determines the marginal propensity to consume and its level to understand changes in demand in the economy.

A new working paper looks at how the amount of wealth inequality can affect the marginal propensity to consume and the resulting implications for policy. The authors, economists Christopher Carroll of Johns Hopkins University, Jiri Slacalek of the European Central Bank, Kiichi Tokuoka of the Japanese Ministry of Finance, and Matthew N. White of the University of Delaware, built a model that tries to replicate the dynamics that determine the amount of wealth inequality in an economy. In figuring out the dynamics that lead to the current levels of wealth inequality in the United States, the model also reveals the marginal propensity to consume among households across the wealth spectrum of the nation.

Carroll and his co-authors find an aggregate MPC, or average MPC for all households, ranging between 0.2 and 0.4. Their estimate is on the high end of other estimates. Their results mean, to return to the question posed above, if I gave you $10, you’d spend between $2 and $4.

Not everyone would spend these extra bucks the same way, of course, because not everyone has the same marginal propensity to consume. The authors find a wide dispersion in the MPC across the wealth distribution. For the most part, less wealthy households have much higher MPCs than wealthier households. But the economists find that the ratio between wealth and income is the key determinant of the MPC.

There are actually quite a few households in their model that have a fair amount of wealth, but a low wealth-to-income ratio, which in turn results in a high marginal propensity to consume. These households may be the “wealthy hand-to-mouth” that economist Greg Kaplan and Justin Weidner, both of Princeton University, and Giovanni L. Violante, of New York University, have written about. In contrast, a household that has a lot of liquid assets, such as investments in the stock market they could easily withdraw, tend to have a much lower MPC.

So what’s the actual real world importance of estimating marginal propensities to consume? Knowing which households are the most likely to spend an extra dollar can help make fiscal policy more effective. According to Carroll and his co-authors, any fiscal stimulus targeted toward individuals in the bottom half of the wealth distribution would be 2 to 3 times more effective than just a blanket stimulus.

Given the high and rising levels of wealth inequality in the United States, this finding should give some guidance to policy makers. Ignoring the distribution of wealth can undermine their ability to get the economy back on track.