Does the trickle-down part of supply-side economics ever work? In a Financial Times article last month, Rana Foroohar pointed out that the recent track record of trickle-down economics may not be terribly impressive:

The argument is that trickle-down economics will somehow magically start working to bolster growth, even though there is no evidence over the past 20 years that this has been the case. Tax cuts in 2001 and 2003 during George W Bush’s administration did not juice growth, nor did any Obama era cuts.

Unfortunately, any discussion on the topic quickly breaks down into what is often a political debate and not a discussion about the economics of income distribution. Either one is in the camp that believes that the economy is automatically better off from tax changes that benefit the wealthy, or one is in the camp that dismisses the whole thing as voodoo economics. The conversation generally ends there.

This is an important topic and one that can have a substantial impact on the future growth prospects for an economy. The issue matters especially to the United States, an economy with a level of income inequality, among the highest in the world, more typical of developing economies than of advanced economies. But rather than being an open-and-shut case when it comes to the United States, I would argue that there have been times in which the economic impact of income inequality has supported long-term growth and other times when it has inhibited growth.

Foroohar’s aforementioned article points out that “Americans do not mind paying taxes,” but instead mind if tax-paying is unfair; Americans are pragmatic enough, in general, to consider long-term growth prospects as relevant in the definition of what is and isn’t unfair. One of the policies the Trump administration has proposed—and similar policies have been proposed directly by nearly every Republican administration since Ronald Reagan and indirectly perhaps by most Democratic administrations since Jimmy Carter—is to cut taxes on the very wealthy. The Hartz reforms in Germany in 2003–2005 had a similar impact, although rather than lower taxes on wealthy Germans, these reforms caused wage growth to slow relative to GDP growth, boosting corporate profits by roughly 50 percent. The net effect has been quite similar, and although income inequality in Germany is quite low, it probably is not just a coincidence that, in recent years, income inequality has sharpened far more rapidly in Germany than in the rest of Europe.

The key point is that the rich save a greater share of their income and consume a smaller share than do the poor. Transferring $100 from poor households to rich households would immediately cause a drop in consumption. It would also cause an immediate rise in ex ante savings, but, as I will show below, whether ex post savings actually rise or not is much more complicated than most people first assume.

The argument in favor of cutting taxes on the very wealthy hinges on the impact this has on savings and investment: it allows well-off individuals to redirect their wealth into increases in productive investment, and as these investments are made, they subsequently increase the productive capacity of the economy. Because the only sustainable way the average citizen in any country can get richer is if worker productivity increases, this, in turn, leads to higher incomes for all, including the less wealthy.

In this essay, I will show that in economies in which savings are scarce, and which consequently suffer from investment levels that are lower than what the market would otherwise choose, income inequality can increase long-term growth. When savings are not scarce, however, income inequality results, over the long term, in lower growth and higher unemployment, although this can be temporarily postponed with rising debt.

Desired and Actual Investment

In order to understand the conditions under which supply-side tax cuts for the wealthy will lead to more growth or greater unemployment, we must make a distinction between actual investment and desired investment. Actual investment, as its name implies, consists simply of the current investment level, that is, all the investment projects that are being funded.

By contrast, desired investment consists of investment projects that have not been funded because capital is not available to fund them. The key here is the availability of funding, and whether or not savings are scarce. There are many potential projects in the world for which injections of investment capital would benefit these economies. This would raise productivity levels by more than the cost of funding such investment, in which case these projects would be wealth-creating and net positive ventures for these economies.

If these projects have not been implemented because savings are scarce and capital is not available to fund them at a reasonable cost, they form part of desired investment. If, however, these projects have not been implemented because of other reasons—say political opposition, or because their value consists of externalities that cannot be realized by the investors, or for other similar reasons—they are not included in the category of desired investment.

In some countries, desired investment substantially exceeds actual investment. These are usually developing countries with low credibility that have high investment needs that have not been met because of weak domestic savings and the limited availability of foreign savings. This condition characterized the United States for much of the nineteenth century and most developing countries today.

In other countries, desired investment is roughly in line with actual investment. This is the case for most advanced economies, in which credibility is high and access to finance at reasonable costs is fairly automatic, especially now, when the world is flooded with excess savings and suffers from limited demand.

Finally, in a few countries, actual investment may exceed desired investment. These are countries that have followed what I have called the Gershenkron development model, in which savings have been forced up, the investment decision has been centralized, and high investment levels have been maintained long after a given economy has reached its optimal investment level. This was the case in Japan in the 1980s, and perhaps still is today to some extent, and it is also the case in China today.

What Happens When Taxes for the Wealthy Are Cut?

In any economy, the two main sources of demand are consumption and investment. There is a third source of demand, the trade surplus, which is the excess of foreign consumption and investment over foreign production, but this is usually dwarfed by domestic consumption and domestic investment. An economy will grow sustainably mainly if there is sustainable growth in domestic consumption and domestic investment. Sustainable growth in consumption is largely a function of growth in household income.

Sustainable growth in investment depends on growth in desired investment and the excess, if any, of desired investment over actual investment. Consumption growth usually drives growth in private sector investment, for obvious reasons, in such a way that the two often grow or shrink together.

The consequence of a tax cut on the wealthy is basically a transfer of wealth from average households to wealthy households, and if it is to generate higher growth this transfer must sustainably increase demand in the form of higher consumption or higher investment.

  1. If taxes on the wealthy are cut with no change in fiscal expenditures, the net result is that the wealthy pay a smaller share of total expenditures, so—with the tax cut—wealth is effectively transferred from average households to wealthy ones.
  2. Because the richer the household, the smaller the consumption share of income, the net impact of such a transfer is that overall consumption declines, along with the consumption share of GDP. As the consumption share of GDP declines, by definition, the savings share must rise.
  3. But total savings need not rise. Here is one of the great confusions that tends to mar nearly every discussion about the impact of tax cuts on the wealthy. A decline in consumption will only be accompanied by an increase in savings if GDP remains constant, but GDP will only remain constant if a decline in consumption is matched dollar for dollar with an increase in investment. Total savings only rise, in other words, if total investment rises.
  4. Will investment rise? It turns out that if desired investment exceeds actual investment, the additional savings will be borrowed and invested, and investment will rise by the same amount by which consumption declines. The net effect is to convert consumption demand into investment demand, so total demand is unchanged. There is consequently no reduction in current GDP, which means that savings rise.
  5. As consumption is transferred into investment, current living standards decline for ordinary households, but future living standards rise as the greater investment leads to faster productivity growth. At first, ordinary and poor households are worse off and wealthy households are better off, in other words, but as the benefits of higher investment are realized, wealth trickles down to ordinary and poor households so all are better off.
  6. But this isn’t necessarily what happens. If desired investment is in line with actual investment, the existence of a higher ex ante level of savings will not increase investment because businesses have already invested all they want to invest. If that is the case, investment will not rise, except temporarily in the form of unwanted investment as inventory rises (lower consumption leaves goods unsold, which makes it seem at first as if point number 4 above is occurring).
  7. This means that consumption demand is not converted into investment demand (except perhaps temporarily), and total demand declines, bringing GDP down along with the decline in total demand. Although there is a decline in consumption, in other words, there is no commensurate increase in savings.
  8. But the rich do save more than the poor, so how is it possible that there is no increase in total savings if there is a transfer of wealth from average households to wealthy ones? The savings of the wealthy have definitely gone up.
  9. But total savings don’t rise because something must happen to reduce savings elsewhere. As consumption declines, those who produce consumer goods and services must cut back, and as they do so, they must fire workers. These workers shift from zero or positive savings to negative savings, so that the reduced savings of the newly unemployed counter the increased savings of the rich.
  10. There are three things that may temporarily interfere with this process. First, as consumption drops, and before businesses that produce consumer goods and services cut back and fire workers, they will see inventory rise or profits decline. Rising inventories cause (unwanted) investment to rise so at first it seems as if points number 4 and number 6 have taken place. But as declining profits reduce business savings, these will counter the rise in the savings of the rich.
  11. Second, the additional wealth of the rich pours into real estate markets and sets off a speculative real estate boom. This causes a rise in real estate development. In such a case, investment rises temporarily along with debt and at first it seems again as if point number 4 has occurred.
  12. Third, the additional wealth of the rich pours into the banking system and banks respond by lowering lending rates and relaxing lending standards. Riskier or more optimistic households that had previously been unable to borrow for consumption are now able to do so, and so consumption does not decline even as money is transferred from average households to wealthy ones. In that case, total demand is unchanged. There is consequently no reduction in GDP, which means that the savings of the rich rise as the savings of the ordinary and poor decline, and GDP is maintained by a rise in debt.
  13. To summarize, if desired investment exceeds actual investment, transfers from ordinary and poor households to wealthy households will immediately cause investment to rise at the expense of consumption, as living standards drop. But the overall economy will increase its total production of goods and services and, depending on how future higher income is distributed, ordinary and poor households will eventually see their income and consumption rise.
  14. If desired investment is in line with actual investment, however, investment will not rise sustainably. At first, either overall consumption will not drop because a rise in debt-fueled consumption will counter the drop in disposable income among ordinary and poor households, or overall investment will rise because of a temporary increase in unproductive investment—that is, unwanted inventory or excess real estate development—either of which also implies a rise in debt. Eventually, however, unemployment will rise, in such a way that as consumption drops, it is not replaced by higher investment but rather is accommodated by lower GDP. Overall savings will remain unchanged because the higher savings of the wealthy will be matched by the lower savings of the unemployed.
  15. I have discussed how the current and capital accounts cause savings to adjust in several earlier blog entries (including here, here, here, and here, for example) and the economic consequences of income inequality in my book, The Great Rebalancing, and several earlier entries (for example, here).

Desired Investment Exceeds Actual Investment

When it comes to determining whether or not supply-side tax cuts benefit the economy, the key is the relationship between desired investment and actual investment. For some reason, most participants in the debate have failed to understand that this is the key to whether or not policies that force up the savings rate benefit an economy, but the logic is pretty clear. If desired investment exceeds actual investment, trickle-down economics works, and everyone eventually can benefit from tax cuts for the wealthy because they lead to an increase in productivity-enhancing investment that ensures that GDP growth in the future will be higher than it otherwise would be.

This was the case in the United States for much of the nineteenth century and is usually true for developing economies with high investment needs and low savings. The United States in the nineteenth century can be thought of as part of a single system that also included Great Britain (and the Netherlands and France to lesser extents). Significant income inequality in Great Britain did not lead to a corresponding increase in British domestic investment but, as excess savings flowed into the United States, it set off higher investment in the United States, a country with substantial investment needs and insufficient savings.

The United States benefitted, of course, because it was able to raise its productive investment to much higher levels than it could have managed had it needed to rely only on domestic savings. The British benefitted because income inequality did not result in higher unemployment or higher debt but rather in a higher trade surplus (since exporting capital is the same as importing demand).

Desired Investment in Line With Actual Investment

If desired investment is in line with actual investment, however, the higher ex ante savings of the wealthy does not lead to greater productivity, more growth, or increased wealth eventually trickling down to all households. It can temporarily lead to no change in GDP as long as nonproductive investment rises (mainly in the form of inventory) or increases in debt-fueled consumption keeps total consumption unchanged. But either way, GDP can only be maintained with an unsustainable rise in the debt burden and, over the longer term, GDP growth must drop.

In that case, rather than implement tax cuts for the wealthy or other policies that increase income inequality, the economy is better off with the reverse. If wealth is transferred from wealthy households to ordinary and poor ones, either consumption will rise without a corresponding reduction in investment because unemployed workers will be put back to work, or consumption will remain constant but it will no longer be fueled by a rise in the debt burden. Either GDP growth rises and wealth actually trickles up, in other words, or it is maintained without a corresponding rise in the debt burden.

By the way, it is important not to ignore the impact of higher consumption on investment. As consumption rises, it turns out that desired investment will rise too, as businesses must produce more goods and services to satisfy higher demand for their products. In that case, GDP growth will rise even more as the increase in GDP is not wholly driven by an increase in consumption. As the income of ordinary and poor households rises, in other words, part of this will be diverted to more consumption, but part of it also will be diverted to more investment, which in turn will increase productivity.

We saw how this happened in Germany, or rather, we saw how a reduction in consumption can actually cause investment to decline, not to rise. After the labor reforms of 2003–2005 effectively pushed down the household share of GDP in favor of business profits, German savings rose. But this happened not because higher ex ante savings were channeled into higher investment, but rather because they were exported to peripheral Europe, so Germany—which until then had been running current account deficits—began to run surpluses the sizes of which soon rose to levels that were historically unprecedented. In fact, as the German savings share of GDP rose, the German investment share of GDP actually declined as German businesses responded to lower consumption at home with lower private sector investment.

This is why former U.S. President Franklin Delano Roosevelt’s brilliant Federal Reserve chairman, Marriner Eccles, a self-made millionaire, insisted in the 1930s that wealth had to be redistributed downward. He did this not because he wanted to give away his wealth but rather because he wanted to keep it. This is the same reason Warren Buffet too wants to reverse income inequality. It is much better to own a factory, Eccles insisted and Buffett would probably agree, when people are eager to buy the things it produces, and can afford to, than when people can’t. During his 1933 testimony to Congress, Eccles quoted with approval an unidentified economist, probably William Trufant Foster:

It is utterly impossible, as this country has demonstrated again and again, for the rich to save as much as they have been trying to save, and save anything that is worth saving. They can save idle factories and useless railroad coaches; they can save empty office buildings and closed banks; they can save paper evidences of foreign loans; but as a class they cannot save anything that is worth saving, above and beyond the amount that is made profitable by the increase of consumer buying. 

It is for the interests of the well-to-do, to protect them from the results of their own folly – that we should take from them a sufficient amount of their surplus to enable consumers to consume and business to operate at a profit.

Actual Investment Exceeds Desired Investment

There are cases in which actual investment exceeds desired investment, strangely enough. China, as I pointed out above, is the most obvious such case because of its dependence on massive increases in investment to generate enough domestic demand to keep the economy running, but there are other examples.

This usually means that investment has long stopped being productive and is being wasted—and on a massive scale that may well be unprecedented in China’s case. When actual investment exceeds desired investment, it is vitally important to reduce investment growth as rapidly as possible and, in order to prevent demand from dropping and unemployment from rising, to increase consumption as rapidly as possible. This is exactly what rebalancing means in China’s context.


Trickle-down economics does indeed work, as does its opposite, trickle-up economics, depending on underlying conditions that are not hard to specify. The key is the relationship between desired investment and actual investment. When the former exceeds the latter, policies that increase income inequality will generally cause savings to rise and expenditures to shift from consumption to investment; this leads to higher future growth that will eventually more than compensate ordinary and poor households for the increase in income inequality.

When desired investment is broadly in line with actual investment, however, there is no trickle-down effect. Policies that increase income inequality must permanently lower growth in the long run, although, in the short run, lower growth can be postponed by an increase in the debt burden.

In advanced economies, like those of the United States and Europe, there is no savings constraint on desired investment, so income inequality can only result in higher debt or higher unemployment and slower growth. It is only in developing countries that income inequality may boost growth, although in countries that have pursued the Gershenkron model of forcing up domestic savings, like China has, actual investment can substantially exceed desired investment. This makes the reduction of income inequality or the channeling of wealth from the state to ordinary and poor households an urgent matter.