Journal Issue

Accepting the Reality of Secular Stagnation

International Monetary Fund. Communications Department
Published Date:
March 2020
  • ShareShare

Related Material

Show Summary Details


A FUNDAMENTAL DIFFERENCE between natural science theories and social science theories is that natural science theories, if valid, hold for all times and places. In contrast, the relevance of economic theories depends on context. Malthus’s theory of food availability was valid for the millennia before he formulated it, but not after the industrial revolution. Keynes’s ideas were much more valid during the Great Depression than during the inflationary 1970s.

I am increasingly convinced that current mac-roeconomic theories, with their premise that monetary policy can determine the rate of inflation, may be unsuited to current economic reality and so provide misguided policy prescriptions. They failed to anticipate Japan’s deflationary slowdown that began in 1990, or the global financial crisis, slow recovery, and below-target inflation during a decade of recovery, or the sustainability of high levels of government debt with very low real interest rates. Understanding these developments and crafting policies that respond effectively will likely require that economists develop what might be called a “new old Keynesian economics” based on Alvin Hansen’s Depression-era idea of secular stagnation. This article summarizes the case for new approaches to macroeconomics by highlighting important structural changes in the economy of the industrial world, explains the secular stagnation view, and draws some policy implications.

The investment dearth

Barring a change in current trends, the industrial world’s working-age population will decline over the next generation, and China’s working-age population will decline as well. At the same time, trends toward increased labor force participation of women have played out with, for example, more women than men now working in the United States.

These demographic developments eliminate the demand for new capital goods to equip and house a growing workforce. This trend is reinforced by the observation that the amount of saving required to purchase a given amount of capital goods has declined sharply as the relative price of equipment, especially in the information technology (IT) space, has sharply declined. A $500 iPhone today has more computing power than a Cray supercomputer did a generation ago. In addition to capital goods’ having lower prices, the downward trend in their prices encourages delaying investment.

Moreover the IT revolution has been associated with a broader demassifcation of the economy. E-commerce has reduced the demand for shopping malls, and the cloud has reduced the demand for office space by eliminating the need for fling cabinets, allowing offices to be personalized with a fick of the switch, down to family photos on the walls.

I am increasingly convinced that current macroeconomic theories, with their premise that monetary policy can determine the rate of inflation, may be unsuited to current economic reality.

Fracking for oil and natural gas requires far less capital than traditional drilling techniques, and IT makes targeting of exploration much easier, further reducing investment demand.

Technology now permits sharing of everything from apartments (Airbnb) to planes (NetJets) and dresses (Rent the Runway) to cars (Uber) in ways that would not have been imaginable a decade ago. Rising generations look to live in sparsely furnished apartments rather than large homes.

Many argue that monopoly power has increased—at least in the United States—tending to discourage new investment. And increasingly promiscuous distribution of the veto power has slowed public infrastructure investment, which on a net basis is running in the United States at less than half of previous levels.

The upshot of all these developments is that investment demand has been substantially reduced, regardless of interest rate levels.

The savings glut

At the same time that investment demand has fallen of, a number of factors have combined to increase saving. A larger amount of income is accruing to higher-income people who have a greater propensity to save. Increased corporate profitability, coupled with lower interest rates, means more corporate retained earnings.

Increases in uncertainty associated with growing doubts about government’s ability to meet pension obligations and more risk of future tax increases also raise saving. Similarly, reductions in expected future income growth increase the need for future saving.

Strengthened financial regulation and its legacy mean households find it more difficult to borrow and spend, leading to an increase in aggregate saving. This can happen either because of consumer protection—as when, for example, higher down payment requirements reduce mortgage borrowing—or because of regulatory burdens on financial intermediaries, through, for example, higher capital requirements.

So structural changes in the economy have operated both to raise saving and to reduce investment.

Secular stagnation

With a somewhat different list of factors in mind, the Harvard economist Alvin Hansen labeled the failure of private investment to fully absorb private savings “secular stagnation” because of the threat that it would mean insufficient demand.

There are a number of things we would expect to see if secular stagnation has been taking hold in recent years. First, a high supply of savings and a low level of demand should mean low interest rates. Indeed, real rates by almost any measure have been trending downward over the last 20 years, even as budget deficits have increased. This is what we have seen with real-term interest rates negative in the industrial world despite major run-ups in government debt.

Second, one would expect that difficulties in absorbing savings would lead to reduced growth and difficulty in achieving target inflation. This is what has been observed. At present markets do not expect any country in the industrial world to hit a 2 percent inflation target. Despite unprecedentedly low interest rates and deficits at record levels after more than a decade of recovery, growth has been tepid. Notably—contrary to the views of those who explained low rates after the recession by pointing to “headwinds”—central banks have found it impossible to raise rates and still count on the momentum of recovery.

Third, disappointing growth has coincided with inflation’s surprising again and again on the downside. Economists teach beginning students that reduced quantity and reduced price suggest a decline in demand. If, as many suggest, the dominant reason for stagnation is disappointing productivity performance, we would expect to see prices rise rather than fall. Absent extraordinary policy settings, deflation might be setting in.

Fourth, a period of slow growth and deflation has also been a period of asset price inflation. US stock markets have risen fourfold since the crisis, and real housing prices are almost back to previous peak levels. This is as one would expect with secular stagnation, as abundant savings pushed into existing assets, increasing, for example, price-to-earnings ratios on stocks and price-to-rent ratios on real estate and decreasing term premiums on long debt.

I am not aware of any other theory that can explain sluggish growth in the face of hyperexpansionary policies and rapid acceleration in private sector credit growth. Lack of productivity growth would be expected to lead to increased product price inflation and reduced asset price inflation. Increased risk and uncertainty would tend to lead to decreased rather than increased asset price multiples. Any temporary consequence of the financial crisis would lead to reduced credit expansion and a steep yield curve rather than what we have observed.

What is to be done?

Demography can be destiny. Much else is moving with demography to create an environment of abundant savings with an absorption problem. This is the mirror image of the macroeconomic problems we have dealt with for decades.

Central banks, to be true to their mandates, need to raise rather than lower inflation. Ensuring that economies fulfill their potential is a challenge that logically comes before increasing their potential. Financial stability is as much at risk from low rates as high rates. The medium-term issue is the full absorption of savings rather than the crowding-out of investment.

At the same time, central banks are unlikely— with rates already negative in Japan and Europe and below 2 percent in the United States—to have much room at least by historical standards to respond to adverse shocks. Typically recessions in the industrial world have been addressed by decreases in rates on the order of 5 percentage points.

The beginning of meeting new challenges is recognizing them. That means accepting the reality of secular stagnation and focusing policy debates on the challenges it poses. FD

LAWRENCE H. SUMMERS is the Charles W. Eliot University Professor at Harvard University.

Other Resources Citing This Publication