Why the Income Distribution Matters for Macroeconomics

March 27, 2014
By

A central argument we have made on this blog and in our book is that the distribution of income/wealth matters a great deal for thinking about the macro-economy. Convincing some of this fact is not easy — many continue to work within a modeling framework in which all distributional considerations are assumed away, the so-called “representative-agent” framework.

Perhaps the easiest way to see the importance of the income distribution is to examine how households respond to a windfall of cash or wealth. Do they spend the money, or do they save it? And does the spending response to a windfall of cash depend on the income of the household?

The answer is a resounding yes: low income households spend a much higher fraction of cash windfalls than high income households. In the parlance of economics, low income households have a much higher marginal propensity to consume, or MPC, than high income households.

This is one of the most well-established facts in empirical research in macroeconomics. Here is a summary:

  • Low income households spend a much higher fraction of fiscal stimulus rebate checks. See the evidence here and here. For the 2001 stimulus checks, low income households spent 63% more of their rebate check than high income households.
  • During the Great Recession, for the same decline in housing wealth, low income households cut back on spending much more. See here. The spending response of low income households is twice as large as that for rich households.
  • During the housing boom from 2002 to 2006, we see the same result. For the same increase in home values, low income households very aggressively borrow and spend. In contrast, high income households are almost completely unresponsive.
  • If a credit card company increases the limit on an individual’s credit card, individuals that are close to the limit spend a much larger fraction of the increased limit. See the evidence here. Individuals right up against the limit spend $0.45 for every $1 of increase in the limit, whereas individuals with plenty of room on their credit card spend only $0.07 for every $1. While this is not exactly a difference across income levels, credit card utilization rates are highly correlated with income so it applies pretty well.

 

The implications of the differences in spending propensities across the population are enormous, especially if we believe that inadequate demand explains economic weakness during severe recessions. For example, facilitating debt forgiveness or progressive fiscal stimulus rebates will likely boost spending during the most severe part of a recession.

But perhaps even more interesting are the implications for the secular stagnation hypothesis, which holds that we are in a long-run stagnating economy because of inadequate demand. Is it a coincidence that the secular stagnation hypothesis is being revived exactly when income inequality is accelerating? If a higher share of income goes to the wealthiest households who spend very little of it, then perhaps these two trends are closely related.

Bookmark and Share

26 Responses to Why the Income Distribution Matters for Macroeconomics

  1. Francesco on March 27, 2014 at 8:20 ami
  2. Josh on March 27, 2014 at 8:49 ami

    When I was studying this in grad school, I always thought the models we looked at ignored the feedback mechanism of wealth distribution. If, for instance, I own all the productive capacity of an economy, every dollar I spend is returned to me in some percentage as income in the next period. The opposite extreme (probably the vast majority in modern economies) receive nothing in the next period, as they own nothing. I could have been mistaken in my understanding, but we always seemed to solve for solutions that did not include the marginal future income of purchases today. It seemed to me that this was a pertinent factor and would suggest that optimal growht would encourage a wider ownership of society’s capital.

  3. Enno Schroeder on March 27, 2014 at 9:28 ami

    That spending propensities vary over income groups is obvious for someone reasoning on the basis of common sense.

    Graduate programs in economics are pretty effective in replacing commons sense with a mindset that is preoccupied with imaginary worlds, rather than real ones. In this sense graduate programs in economics necessitate your post.

    Differential spending propensities are at the heart of the post-Keynesian research program, but graduate students of economics are hardly ever exposed to it.

  4. Manfred on March 27, 2014 at 11:01 ami

    Amir & Atif, my apologies for sounding a discordant note, but you both being faculty members of two of the most elite of the elite universities in the world, you have to come up with something much better. You simplify so much (at least in this post), that it is misleading, and as Economics (or Finance) professors at elite school you should, must, know better.
    Take, for example, the first bullet point – that lower income people spent 63% more of the rebate checks etc. So? Of course, poorer people will spend more, but so what? The important question is, was it effective to stimulate the economy as a whole on a permanent basis? There are very bright people working a elite departments like yours that say “no”, it was not effective. Therefore, why does it help the policymaker to know that poorer people spend a higher fraction of some silly stimulus check? But you do not mention anything about this.
    Another example: take the first sentence in the last paragraph: “…implications for the secular stagnation hypothesis, which holds that we are in a long-run stagnating economy because of inadequate demand.” How about that perhaps, just maybe, we are *not* stagnating because of “inadequate demand” – how about we are stagnating because of silly and burdensome government policies that only throw wrenches into the workings of the economy. Again, there are very prestigious people, very smart professors, working at elite schools like yours, that hold such thesis, but you mention nothing.
    Finally, a third example: the last sentence of the last paragraph: “If a higher share of income goes to the wealthiest households who spend very little of it, then perhaps these two trends are closely related.” What do you mean that the richest households “spend very little” of it? What exactly do you think the richest of the rich do with their money? Hold it under a mattress? Put it into some type of garage like Scrooge? Or do they invest it, and create jobs (and thus, help relieve your “indadequate demand”)? What exactly do you think Mark Zuckerberg does with his money, or Bill Gates? Just sit on it and stare at it?
    I am sorry, but such statements are below a professor from highly distinguished departments of Finance and/or Economics. If you want to educate the public, you must do much better.

    • ISLM on March 27, 2014 at 1:01 pmi

      Some citations would help to support your conjectures, in particular the point about “burdensome government policies.”

      • Manfred on March 28, 2014 at 4:34 pmi

        Try Casey Mulligan (a colleague of Amir Sufi in Chicago) and his work on the Affordable Care Act. Try Nicholas Bloom in Stanford and his work on policy uncertainty.

    • Anonymous on March 27, 2014 at 3:15 pmi

      You should put some numbers into your answers as the authors of the post did. If not, you are just telling a story.

      • Manfred on March 28, 2014 at 4:35 pmi

        What – the authors of the original blog are not telling a story? Of course they are.

    • Richard on March 27, 2014 at 6:29 pmi

      Manfred:

      Are you so ignorant about economics that you have to rely on “prestige” to tell whether an argument is a good one or not?

      1. “There are very bright people working a elite departments like yours that say “no”, it was not effective.”

      And what is their proof? Are they saying that the multiplier is less than 1 or something else?

      2. “how about we are stagnating because of silly and burdensome government policies that only throw wrenches into the workings of the economy”

      So test that. Not just across time, but also countries (does the change in the level of “burdensome government policies” explain the difference in economic performance between the UK and Germany in recent years, for instance?) If you don’t, you are just buying in to a pet theory that fits your prejudices.

      3. “Or do they invest it, and create jobs”

      You do realize that investment isn’t spending, don’t you? Plus, what do they invest them in? If they invest them in treasuries in a world where banks are awash in liquidity, why do you think that would create more jobs than increased demand? Also, why do you hold the preconceived assumption that it is investment that creates jobs rather than demand?

      • Manfred on March 28, 2014 at 4:29 pmi

        Richard,
        you call me ignorant of economics, but you display your ignorance all over.
        I may delve later into the details of your answer to me, but as of now, I will say this: Investment is not spending? What is it then? Of course, it is spending; of course it is part of “aggregate demand”, any decent first principles textbook will tell you this; GDP = C + I + G + NX, where the “I” part is investment. There is a reason why this simple equation in all textbooks is called the “expenditure approach”.
        I do not follow Bill Gates around and look what he invests in. I only know that he (and other like him, like Mark Zuckerberg, Larry Ellison, etc) created huge companies, huge payrolls, and created prosperity for thousands and thousands of people. Something that the good professors of the original blog seem to forget.

        • Richard on March 30, 2014 at 11:15 pmi

          “I only know that he (and other like him, like Mark Zuckerberg, Larry Ellison, etc) created huge companies, huge payrolls, and created prosperity for thousands and thousands of people.”

          Did they create that without demand for the products that their companies put out?

          That is the crux of the question. Without a prosperous and broad-based consuming middle-class, what are the returns on entrepreneurship?

    • Longtooth on March 28, 2014 at 10:51 pmi

      Manfred’s argument is that A should agree with B, or if A is not in agreement with B then A should at least cite B’s opposing arguments.

      Since A did neither in this case then Manfred concludes A is misleading.

      I actually can’t disagree with Manfred’s logic as a generality. However, propaganda is in fact designed with intent to mislead by citing a one-sided argument without disclosing, and in fact intentionally not disclosing opposing facts which would put the propaganda (and the propagandist) in a less favorable light.

      But in this case, i.e. this blog article, to cite the opposing arguments by the other smart people from elite universities would require a magnum-opus since the authors would have to being with Smith, Ricardo and go on from there to cite opposing arguments by smart people from elite universities.

      Manfred’s argument is therefore out of context and without merit… unless Manfred could reasonably expect a magnum-opus in a blog article. I submit that Manfred’s expectations are therefore unworthy of further discussion in any context.

    • Al on March 30, 2014 at 4:49 ami

      Most of your complaints are misreadings of the article. The C in MPC stands for consumption. It does not imply that the wealthy put their money in mattresses (which is a rather random interpretation). As to Nicholas Bloom, and alternative explanations, the authors discussed some of these points in their analysis of state level data:
      http://www.frbsf.org/economic-research/publications/economic-letter/2013/february/aggregate-demand-state-level-employment/

  5. Pookah Harvey on March 27, 2014 at 11:16 ami

    Michael Pettis in his blog article “Economic consequences of income inequality” quotes from former Fed Chairman (1932-48) Marriner Eccles, from his memoir, Beckoning Frontiers (1966):

    “As mass production has to be accompanied by mass consumption, mass consumption, in turn, implies a distribution of wealth – not of existing wealth, but of wealth as it is currently produced – to provide men with buying power equal to the amount of goods and services offered by the nation’s economic machinery. Instead of achieving that kind of distribution, a giant suction pump had by 1929-30 drawn into a few hands an increasing portion of currently produced wealth. This served them as capital accumulations.

    But by taking purchasing power out of the hands of mass consumers, the savers denied to themselves the kind of effective demand for their products that would justify a reinvestment of their capital accumulations in new plants. In consequence, as in a poker game where the chips were concentrated in fewer and fewer hands, the other fellows could stay in the game only by borrowing. When their credit ran out, the game stopped.”

    It would seem that productivity and median income must stay in balance to keep a high rate of GDP growth.

    • State of Thought on March 28, 2014 at 4:33 pmi

      Sadly, the Eccles quote seems to have been the main worthwhile part of Pettis’ post. Aside from that good quote, there were a several questionable presumptions without any cited support … making that particular entry in the inequality discussion rather dubious.

  6. Steve Roth on March 27, 2014 at 12:10 pmi

    Thanks for these great links to the empirical literature. Got my reading cut out for me.

    On the theory: the arithmetic of this is inexorable — roughed out in a simple model here:

    http://www.asymptosis.com/does-upward-redistribution-cause-secular-stagnation.html

    As the post says, this is only one economic effect being illustrated. But the potential power of this effect seems massive — especially over decades — and it seems to be almost universally ignored and even pooh-poohed by economists, even liberal economists.

    http://www.asymptosis.com/a-liberal-is-someone-who-doesnt-know-how-to-take-his-own-side-in-an-argument.html

    http://www.asymptosis.com/underconsumption-income-wealth-and-capital-gains.html

  7. Lazy guy on March 28, 2014 at 5:40 ami

    The wealthy may not consume the windfall, but will instead use it for investment. Investment ultimately determines potential GDP (think of a Solow model). So, arguing for progressive taxation is just asking for a less productive economy in the long run.

    • Richard on March 28, 2014 at 9:05 ami

      You’re assuming a capital constrained world. In a world where central banks flood the world with liquidity, there’s not much evidence that “the wealthy using wealth for investment” would increase productivity, as companies can get access to easy capital right now.

      Plus, from Wikipedia:
      “In the Solow-Swan model the long-run rate of growth is exogenously determined – in other words, it is determined outside of the model by assumptions not explained by the model. Given any constant savings rate, the economy will always converge to a steady state rate of growth, which depends only on the rate of technological progress and the rate of labor force growth. In the steady state a country with a high savings rate and a country with a low savings rate will experience the same rate of growth in output/worker.

      A country that increases its savings rate (as a share of output) will experience faster growth temporarily until it converges again to the steady state rate. As a consequence, the model predicts that if poor countries will grow rapidly if they raise the share of output (GDP) they invest in capital. Since there is a practical limit on the maximum share of GDP that can be invested, eventually savings rates must stabilize. When this happens the Solow-Swan model predicts that economic growth per worker will converge to the lower, steady-state rate determined by technological progress.”

      So what you say isn’t even true.

      • Lazy guy on March 28, 2014 at 11:31 pmi

        Yes, in the long run the growth rate is determined by technology. I’m talking about whether the standard of living is comparable to that of Singapore or Mexico.

    • MN on March 28, 2014 at 10:35 ami

      Lazy Guy: “so arguing for progressive taxation is just asking for a less productive economy in the long run”.

      This is only true if wealthy households are omniscient philosopher-kings that rationally allocate their wealth towards the most productive ends. This is patently untrue. Not only is owning multiple, unused mansions spectacularly unproductive investment, but so are many of the investments made whenever a bubble starts inflating. If we followed your claim to its logical conclusion, we would have to argue that public education financed with progressive taxation is “just asking for a less productive economy in the long run”. Is that really what you think?

      • Lazy guy on March 28, 2014 at 11:38 pmi

        My point relates to the trade-off between consumption and investment. Favoring consumption diminishes potential GDP in the long run, because it hinders capital formation. Public education is a form of investment, so spending on public education arguably boosts long run potential GDP. Spending tax dollars in a way that is primarily consumption, such as Medicare, is a better example of something that lowers GDP in the long run.

        • MN on March 29, 2014 at 11:03 ami

          Lazy Guy, fair enough. But I think you can also make a strong case for public health being an investment in the same way as education.

          And this is the problem with hazy concepts like “capital formation”. What does that actually mean? Human capital? Factories? Rising market capitalizations via buybacks? All three could be considered capital formation, but they are very different.

  8. Eric Zuesse on March 28, 2014 at 12:44 pmi

    I have in draft and am currently seeking comments upon a paper that presents an empirically supported case for replacing the Pareto Welfare Criterion with a new welfare criterion, which is supported by the very same mounting body of empirical studies that disconfirm the PWC. This paper argues that the PWC is the basic reason for economists’ having difficulty recognizing that inequality of wealth can impact the output of an economy. Basic microeconomic theory embodies the PWC, which is a false assumption; and this is the reason why economists have difficulty acknowledging this. If interested, I invite you to comment upon it; reach me at hyacintheditions@mail.com and I’ll be happy to email to you the .pdf.

  9. Mark A. Sadowski on March 29, 2014 at 8:02 ami

    “But perhaps even more interesting are the implications for the secular stagnation hypothesis, which holds that we are in a long-run stagnating economy because of inadequate demand. Is it a coincidence that the secular stagnation hypothesis is being revived exactly when income inequality is accelerating? If a higher share of income goes to the wealthiest households who spend very little of it, then perhaps these two trends are closely related.”

    In underconsumption theory recessions and stagnation arise due to inadequate consumer demand relative to the production of new goods and services. Consequently it is precisely the amount of savings relative to the income that is derived from the production of new goods and services that is connected to the matter of whether underconsumption is actually taking place. So changes in wealth, and the income derived from the buying and selling of assets, are completely peripheral given the very nature of underconsumption theory.

    The handful of studies that produce aggregate savings rates consistent with the NIPA/SNA accounts, have shown that savings rates at times have actually been inversely related to income. Maki and Palumbo (2001) came up with the following estimates by quintile for the the U.S. for the year 2000 (Table 2):

    Top quintile : (-2.1%)
    Fourth quintile : 2.6%
    Middle quintile : 2.9%
    Second quintile : 7.4%
    Bottom quintile : 7.1%

    http://www.federalreserve.gov/pubs/feds/2001/200121/200121pap.pdf

    In the largest, and by most accounts, best study to date on the question of inequality and national savings, by Schmidt-Hebbel and Serven (2000), involving World Bank data on 82 countries over 1965-1994, they generally found no significant correlation between measures of income inequality and national savings.

    The lone exception was in a regression that used the income share ratio of top 20% to bottom 40%. The correlation was significant at the 5% level, but was negative. In other words, it suggested that greater inequality resulted in lower national savings.

    http://darp.lse.ac.uk/papersdb/Schmidt-Serven_(JDE2000).pdf

    So much for the cross-sectional evidence. How about the time series evidence?

    Well we all know that inequality bottomed in the US between 1968 and 1978 depending on the measure one uses (Gini Index, income ratios, income concentration etc.) and has soared since then. Since those with high incomes now have a much larger share of total income than they did in the 1970s, if those with high incomes save more than those with low incomes, obviously savings has soared, right? No.

    http://research.stlouisfed.org/fred2/graph/?graph_id=152807&category_id=0

    Personal savings as a percent of GDP (blue line) peaked at 9.4% in 1971 and fell as low as 1.8% in 2005. It has increased since then but that is due to the recession, not to a dramatic surge in inequality.

    Well maybe the rich were in effect having corporations save on their behalf (red line). It’s more volatile, but there too we see that domestic business savings as a percent of GDP peaked at 5.1% in 1965 and fell to a low of 1.1% in 2008.

    Thus it’s not surprising when we sum the two, net private savings (green) fell from a high of 13.0% of GDP in 1973 to a low of 4.0% of GDP in 2007. Adding in capital consumption to make it gross private savings (orange) results in a peak of 23.2% of GDP in 1975 and a low of 16.2% in 2000.

    In short, during the time that income inequality soared, savings plunged. So both the cross sectional and the time series evidence fail to support the idea that those with high incomes save more than those with low incomes.

  10. Mark A. Sadowski on March 29, 2014 at 8:38 ami

    “Low income households spend a much higher fraction of fiscal stimulus rebate checks. See the evidence here and here. For the 2001 stimulus checks, low income households spent 63% more of their rebate check than high income households.”

    The results of Johnson, Parker and Souleles (2001), showing differences in spending of the 2001 stimulus check by income level (Table 6 on page 40), are not statistically significant.

    The results of Johnson, Parker and Souleles (2008), showing differences in spending of the 2008 stimulus check by income level (Table 5), are statistically significant, but only for the low income group.

    Moreover there are other studies on the effects of fiscal stimulus rebate checks on spending, that look not only at income, but non-liquid financial wealth.

    Here is one by Sahm, Shapiro and Slemrod (2009):

    http://www.nber.org/papers/w15421.pdf?new_window=1

    In particular look at Table 7. They find no significant difference in plans to spend the 2008 Tax Rebate by income level, although those in the highest income category were more likely than any other group to spend the rebate.

    They do find a significant difference (p-value = 0.078) by wealth class (stocks), but again, they found those that were in the highest wealth category were by far the most likely to spend the rebate.

    So I would suggest that the results of studies on the effects of fiscal stimulus rebate checks on spending are far less conclusive, and far more mixed than you are suggesting.