Who Spends Extra Cash?

April 13, 2014

Imagine we dropped cash on every household in the country. Who would spend it? Who would save it? The answer to this question matters a great deal given the rise in inequality before the Great Recession. It also matters because the economy is likely to be demand-constrained during severe downturns, especially if the economy hits the zero-lower bound on nominal interest rates. If all households reacted to a cash windfall the same, then the distribution of income or wealth wouldn’t matter much for cyclical policy.

We argued in a previous post that  lower income/wealth households have a much higher propensity to spend out of cash windfalls. Another study supporting this claim is the Jappelli and Pistaferri (2013) study forthcoming in the American Economic Journal: Macroeconomics. They used answers to a 2010 survey in Italy that asked consumers how much of an unexpected cash windfall they would spend. The first notable result is that the average marginal propensity to consume out of a cash windfall shock was 48%. So 48% of the cash windfall would be spent on average.

Even more interesting, the authors found that the MPC was much larger for households that had lower “cash on hand.” Cash on hand is measured as household disposable income plus financial assets minus debt. So more casually speaking, households with high “cash on hand” are rich and those with low “cash on hand” are poor. Here is the key chart:


For the poorest households, the marginal propensity to consume was close to 70%. For the richest households, the MPC was only 35%. So even more evidence that lower income/wealth households spend a much larger fraction of cash windfalls.

The abstract makes the following policy conclusions:

“The results have important implications for the evaluation of fiscal policy, and for predicting household responses to tax reforms and redistributive policies. In particular, we find that a debt-financed increase in transfers of 1 percent of national disposable income targeted to the bottom decile of the cash-on-hand distribution would increase aggregate consumption by 0.82 percent. Furthermore, we find that redistributing 1% of national disposable income from the top to the bottom decile of the income distribution would boost aggregate consumption by 0.33%.”

In the face of a severe economic downturn, we think a better policy would be debt forgiveness, which would act as a transfer from higher income households who tend to own equity in the financial sector to lower income households who tend to have a lot of debt relative to assets or income. But the basic principle is the same: the distribution matters when devising cyclical policy.

Bookmark and Share

19 Responses to Who Spends Extra Cash?

  1. Anwer Khan on April 13, 2014 at 8:22 ami

    I agree with debt forgiveness in downturns. Likewise, I would let high-income households capture more of the gains from economic booms. For maximum stability we should restructure debt contracts / banking such that this happens automatically without any nonlinearities.

    • BigBozat on April 14, 2014 at 11:35 ami

      Anwer Hhan said:
      “Likewise, I would let high-income households capture more of the gains from economic booms.”


      Maybe your definition of ‘high-income households’ (and/or ‘economic booms’) is radically different than mine — or perhaps we live in radically different places or times — but, from what I can see it would be awfully hard for high-income households to capture more of the gains from economic booms than they already have/do…

      In the U.S., between 1969 and 2008 ALL growth in average incomes went to the Top 10%, while income for the Bottom 90% declined (most of which, BTW, was actually captured by the Top1%)…

      Too long of a secular period for you to consider a ‘boom’? How about the Clinton era? Between 1992 and 2000, average incomes grew by $13,682, of which 70% was captured by the Top 10%… not as eye-popping as the longer trend, but still remarkably inequitable (IMHO)…

      It hasn’t always been like that and – I’d argue – it isn’t at all necessary to produce prosperity. If we split the post-WW II era into two equal length periods:

      Between 1946 and 1977, average incomes grew (in real terms, using 2008 dollars) $18,673… the Top 10% captured 30% of that growth, while the Bottom 90% captured 70%. Conversely, between 1977 and 2008 average incomes grew by [only] $11,699 [i.e., marginally slower growth]… of which ALL (100%!) growth was captured by the Top 10%, while income declined for the Bottom 90%.

      (Sources for all of the above stats, see: http://www.stateofworkingamerica.org/who-gains/#/?start=1969&end=2008 -and/or- http://www.econ.berkeley.edu/~saez/TabFig2008.xls#sthash.eMrpOPqf.dpuf)

      IMO, high-income households already capture more of the gains from growth than is healthy… allowing them to capture even more seems like unthinking plutocratic apologism.

      • Anwer Khan on April 15, 2014 at 5:41 pmi

        I agree that the secular trends are bad and that we should be concerned about inequality. But in this post the focus seems to be on cyclical policy, and my point is that if debt forgiveness reduces the severity of downturns, then the same principle applied in reverse (during booms) also enhances stability.

    • Tom in MN on April 14, 2014 at 10:35 pmi

      “Likewise, I would let high-income households capture more of the gains from economic booms” — Done, if you will settle for 90% to them.

  2. Fair Economist on April 13, 2014 at 2:27 pmi

    I don’t think debt forgiveness is the best tool, although it does play a role. The people with large debts are generally not the poor, with the highest propensity to consume, but the middle and even upper middle class who have either overstretched or hit hard times. Particularly for that section of the middle class, gains from debt forgiveness are almost always just “saved” because they’re not likely to incur more debt in the near future and it doesn’t actually give them anything to spend.

    One approach I’d like to see is an attack on what you might call “implicit debt”. To get by in our society requires a lot of ongoing expenses in the form of housing, transportation, and food. The effect of a payment that you absolutely have to make to continue to survive has similar effects to a legally mandated dabt payment, even if it’s formally different, and can lead to similar situations of snowballing debt and forestalled personal investment. In particular, improving public transportation and creating more concentrated low-income housing could allow more to live without the cost of maintaining a car, which is absolutely crushing for people with low or insecure incomes.

  3. jcb on April 13, 2014 at 5:39 pmi

    A very interesting piece, and I couldn’t agree more that debt forgiveness is–would have been–the most effective response to the Crash of 2007. That’s why the temporary nationalization of the large banks, rather than providing them with bailouts, would have the single most effective means for containing the deflation of the bubble, and stimulating the economy.

    In particular, bankruptcy would have deflated the wealth in bank assets held by the wealthiest consumers (but not the federally guaranteed deposits of most consumers). More important, mortgages for owner-occupied housing under a certain value could have been written down to market value through federally controlled (under bankruptcy) bank foreclosure and repurchase for qualifying households.

    Guaranteeing both the solvency of the banks and the contract value of their mortgages was a self-defeating policy that gave the financial sector no further incentive for stimulus.

    As Fair Economist writes above, providing basic public services–healthcare, education, transportation, social security–is another means of increasing the propensity to consume on a broad level.

  4. Steve Roth on April 14, 2014 at 9:31 ami

    I built a little model of this MPC/redistribution effect that you might find interesting:


  5. Tyler on April 14, 2014 at 9:44 ami

    The poor should be exempt from payroll taxes.

  6. Steve Roth on April 14, 2014 at 10:00 ami

    Sorry I realize I’ve pointed you to that before. Thanks for continuing this discussion. Good stuff.

    On the transfer from creditors to debtors, higher inflation seems a more palatable policy. It transfers real buying power (to the tune of tens or hundreds of billions a year for each extra point of inflation), with nary a dollar transfer between accounts.

    Why doesn’t the Fed do that? It’s run by creditors.

    • jcb on April 14, 2014 at 11:26 ami

      I don’t understand the fascination with “higher inflation” as a policy solution for deficient aggregate demand. I get that inflation modifies the relative positions of debtors and creditors to the advantage of the former. A good thing.

      But since “inflation” has many components and is not simply manna spread equally over every category of income and expenditure, why is it desirable? Does anyone think if the price of, say, oil inflates that the rate of wage increases will march in tandem? Inflation isn’t primarily a monetary phenomenon at all. It requires an object of expenditure. This may be excess of demand over supply; it may also be excessive, speculative buying of a necessity for which the elasticity of demand is low.

      Inflation and deflation are price signals of real economic forces at work. But even assuming that both can be summoned by monetary policy, why is either per se desirable or undesirable? You need to beg a question (i.e. assume a conclusion) to think it one or the other.

    • Anwer Khan on April 14, 2014 at 2:21 pmi

      Inflation does have that effect, but the Phillips curve tells us that it happens at the wrong times. If we want stabilization we should index debt contracts to nominal GDP.

    • cas127 on April 15, 2014 at 1:24 pmi

      “It’s run by creditors.”

      Golly…*that* sure explains well over a decade of ZIRP.

  7. MacCruiskeen on April 14, 2014 at 1:46 pmi

    Do these studies account for the difference between what people tell surveys what they will do and what they will actually do? Every year we before the xmas holiday, shoppers are surveyed about how much they plan to spend, and these surveys are not good predictors of actual spending.

  8. Mary on April 14, 2014 at 4:51 pmi

    The poorest decile cannot obtain credit and so debt forgiveness won’t touch them. If you are unaware of this, you need to investigate the financial realities of life at the bottom. These are people with no credit cards and no bank accounts. They have never had a car loan. They do not have college loans. Another issue is that if debt forgiveness became routine, it would risk becoming a moral hazard.

    • jcb on April 15, 2014 at 2:34 pmi


      I agree with your first point. The lowest decile ought to be the recipient of the largest amount of transfer payments from the government–income supplements, social security, unemployment insurance extension, medicaid. I say, “ought to be.” But nothing would have supported aggregate demand more than mortgage relief, and it did not happen.

      As for “moral hazard,” we already have it: it’s called “credit forgiveness” and it applies only to large financial institutions.

    • Anwer Khan on April 15, 2014 at 5:46 pmi

      Even the people who do not qualify for credit will benefit from increases in aggregate demand, because they will have an easier time finding work. And before we talk about moral hazard, let’s consider how much influence individuals have on aggregate demand.

  9. Fed Up on April 14, 2014 at 10:39 pmi

    “Imagine we dropped cash on every household in the country. Who would spend it? Who would save it?”

    A better idea would be to apply this to QE.

    The fed buys a bond from Warren Buffett or Apple. The chances of either one spending it on a consumption good or investment good is extremely low.

    Do you know this tends to contradict the hot potato people?

  10. Fed Up on April 14, 2014 at 10:40 pmi

    “On the transfer from creditors to debtors, higher inflation seems a more palatable policy. It transfers real buying power (to the tune of tens or hundreds of billions a year for each extra point of inflation), with nary a dollar transfer between accounts.”

    What if prices go up, but wages do not?

  11. Kaleberg on April 15, 2014 at 11:21 pmi

    1) We knew this. It was common knowledge in the 19th and 20th centuries. It was repeatedly empirically demonstrated and measured. It was frequently cited as an argument for the New Deal and countercyclic spending. It’s nice to know that water still runs downhill and that we are once again allowed to say so.

    2) While debt reduction was a big thing in the Old Testament, it would miss helping a lot of people. There are huge numbers of Americans who have never had access to debt, and they are at the bottom end of the income/wealth scale so they would spend the most.

    (This isn’t 100% true if you allow people to capitalize their bodies. Poor people will spend every penny they can get, often on deferred maintenance like getting a tooth fixed or a pair of eyeglasses. In that sense, the ACA is a form of debt reduction and given how it has cut insurance premiums for so many, we should start hearing how it is a backdoor stimulus in three or four years.)