Unlike Keynes’s 1930 Treatise on Money, his General Theory offered no coherent theory of inflation or the price level but instead treated nominal and real income as the same thing. He suggested that “an increase in the quantity of money will have no effect whatever on prices, so long as there is any unemployment.” That idea was later formalized in the Phillips Curve tradeoff, whereby lower unemployment could supposedly be achieved through higher inflation. The results of that policy bias were the disastrous inflationary recessions of 1974–75 and 1980–82.
In 1978, future Nobel laureate Robert Lucas Jr. wrote an obituary for these ideas, “After Keynesian Economics,” along with Thomas Sargent of the University of Minnesota. They showed that “Keynesian … predictions were wildly incorrect and that the doctrine on which they were based is fundamentally flawed.” The hubris of expert demand‐management through fiscal policy should have suffered a permanent loss of credibility 30 years ago. But memory is short.
One reason Keynesian theorizing never quite disappears is that our national‐income model deliberately incorporates Keynesian concepts. Keynes described the overall economy in terms of how money is spent rather than how it is earned. He divided national income into a few arbitrary accounting categories, describing income (denoted by the letter “Y”) as being spent for consumption (“C”), investment (“I”), government (“G”), and net imports (“X”). Ignoring foreign trade, as Keynes usually did, this yields the famous equation: Y=C+I+G. “The decisions to consume and the decisions to invest,” he wrote, “between them determine incomes.”
The theory remains much too popular — because it is much too simple. Keynes’s discussion of consumption makes no distinction between durable and nondurable goods, and regards consumption as dependent on current income alone, not wealth. Yet young people clearly consume out of human capital (expected future income) and seniors consume out of accumulated financial capital.
How many times have we read the demand‐side fallacy — namely, that economic growth “depends on” consumption, because consumption accounts for 70 percent of GDP? To say that income growth depends on consumption would be absurdly circular even in Keynesian terms, because Keynes argues that consumption depends on income. In reality, Keynes attributed sudden gyrations in income to changes in investment. This is a real theory of the business cycle, which may be both the best and least understood part of Keynes’s work. Recessions arise, he said, “where investment is being made in conditions which are unstable and cannot endure, because it is prompted by expectations which are destined to disappointment.” Think of highly leveraged investments in Las Vegas condos a few years ago by those who thought they could resell at a higher price before the teaser rate on the mortgage went higher.
If such wrongheaded private investment collapsed, Keynes worried, fear could keep investment depressed for a long time. So he proposed offsetting the drop in private investment with government purchases. When it came to public works, the more wasteful the better — because unproductive investments would not crowd out private investment: “If the Treasury were to fill old bottles with banknotes, bury them at suitable depths in disused coal mines which are then filled up to the surface with town rubbish, and leave it to private enterprise on well‐tried principles of laissez‐faire to dig the notes up again … there need be no more unemployment.”
Such reasoning lay behind the infamous “multiplier,” which the late Harry Johnson described as an “inexhaustibly versatile mechanical toy.” Because people employed in burying and digging up bottles will supposedly employ other people by spending their paychecks, the initial increase in government spending was thought to have a multiple effect on total spending. And that, said Keynes, will lead to an “increase in employment and hence in real income.” But checks received for producing nothing are not real income. Real income per worker depends on real output per worker — incentives to produce, not incentives to spend.
If there is no multiplier effect, the multiplier is one — a billion dollars of government spending adds a billion to national income, but no more. Keynes offered a hypothetical example suggesting the multiplier could be ten if people promptly spent 90 percent of added income on consumer goods. That is how he came to imagine that “public works even of doubtful utility may pay for themselves over and over again at a time of severe unemployment if only from the diminished cost of relief expenditure.”
Recent research finds multipliers to be very small at best, if not negative. In 2002, the IMF published “The Effectiveness of Fiscal Policy in Stimulating Economic Activity — a Review of the Literature” by Richard Hemming, Michael Kell, and Selma Mahfouz. They found that “short‐term multipliers average around a half for taxes and one for spending, with only modest variation across countries and models.”
The C+I+G rubric is a tautology — true by definition. Yet it seduces people into confusing the uses of income (spending) with the sources of income (production). One person’s spending is another person’s income, but that does not mean the mere act of spending money creates real income. If that were true, then every poor country could become rich by simply dropping money from helicopters.
Aside from transfer payments, the sources of income are payments for producing something consumers or taxpayers are willing to pay for. Saving generally involves spending too (buying stocks or bonds), with sellers receiving what the buyers spend. The fear‐driven urge to keep savings piling up in cash (rather than stocks, bonds, or property) could become deflationary. But that is why the Fed has been meeting that surge in the demand for money with additional supply, while also making it unrewarding for potential investors to just sit on T‐bills.
Today’s business press is full of Keynesian stories about the “paradox of thrift” — fretting about consumers’ saving more and therefore buying less. Yet it’s quite possible for both savings and consumption to rise at the same time if income or wealth rises. And incomes and wealth will rise if conditions become more favorable for producers, including workers. Lower inflation, for example, was already raising households’ real disposable income by October and November of 2008.
I sometimes joke about having had trouble with Keynesian accounting in school — because I always wanted to subtract G. It’s not just a joke. Government purchases of real resources absorb labor, land, equipment, and materials, and thereby raise the cost of production for private businesses, damaging the profitability of private investment. Government transfer payments are a disincentive for those who receive the benefits and for the taxpayers who pay.
Alberto Alesina of Harvard published a major long‐term study of fiscal policy changes in 18 economies in The American Economic Review, September 2002. What they found was that “fiscal stabilizations that have led to an increase in growth consist mainly of spending cuts, particularly in government wages and transfers, while those associated with a downturn in the economy are characterized by tax increases.” Ireland experienced miraculous economic growth after cutting spending by an amount equal to 7 percent of GDP (the equivalent of the United States’ eliminating two Pentagons’ worth of spending) in the late 1980s, then slashing marginal tax rates on profits and capital gains. As an IMF report explained, Ireland also “significantly reduced the exceptionally progressive nature of the progressive tax structure and increased work incentives.” By contrast, Japan ran budget deficits that averaged 5.8 percent of GDP from 1993 to 2005, and the economy was stagnant.
In 1997, Christina Romer, Obama’s choice to head the Council of Economic Advisers, found that a U.S. tax increase amounting to 1 percent of GDP reduces real GDP by nearly 3 percent within three years, with employment falling 1.1 percent and housing and business investment by 12.6 percent. Explaining the persistence of the damage from tax increases, she suggests “tax changes could have large supply‐side effects.”
Confronted with such inconvenient facts, Keynesians spin a different theory. When the economy recovers (as it always does), they say that is because budget deficits stimulate demand. If the economy later slumps, they’re likely to say that it is because budget deficits crowd out investment. If the dollar goes up, some Keynesians are sure to argue that budget deficits attract foreign investment. If the dollar goes down, they’ll say it’s because budget deficits create fears of inflation. If inflation goes up, that will be considered proof that budget deficits are inflationary. If inflation goes down, that just proves budget deficits are not large enough. The answer is always the same; only the questions change.
A theory that can explain everything explains nothing. If Keynesian theorists refuse to accept any evidence as contradicting their theory, they are practicing a secular theology, not science.
A dozen years of massive public‐works spending in Japan were associated with the weakest economic performance of any major economy of the time. Yet Paul Krugman now speculates that “even in Japan … public spending probably prevented a weak economy from plunging into an actual depression.” That leaves Keynesians with no testable hypothesis. They predicted that more G in Japan would produce much more Y, through the magic of multipliers. Whenever their predictions fail, Keynesians insist their theory is correct but reality has gone awry.
When the government spends money, or gives it away in rebate checks, that undoubtedly “stimulates” those who receive the cash. But it has the opposite effect on those who pay the bills. Karl Marx, in his 1852 critique of Louis Bonaparte, explained that “the people are to be given employment: initiation of public works. But the public works increase the people’s tax obligations.… Taxes are the life source of the bureaucracy.… Strong government and heavy taxes are identical.”
When the government borrows from Peter to pay Paul, taxpayers then have an obligation either to pay interest to Peter (forever) or to repay the debt. Government money does not become free money simply because it was borrowed. True, the U.S. government has been able to borrow at extremely low interest rates lately, but that never lasts for long. The extra debt the new administration plans to dump on the backs of future taxpayers will have to be rolled over at higher interest rates, sooner or later (more likely sooner) and the rising cost of debt service will be borne by taxpayers unless the government defaults (which is already the subject of some speculation).
The CBO estimates that the 2009 budget deficit will be $1.2 trillion, or 8.3 percent of GDP. Obama’s “recovery and reinvestment” plan is expected to increase the deficit by $825 billion over two years. Assuming that sum is split evenly between 2009 and 2010, it would raise the estimated deficit to 11.3 percent of GDP this year and 7.6 percent in the following year. And that’s not counting another $350 billion for the TARP slush fund.
A deficit of 11.3 percent of GDP would be nearly twice the previous peacetime record of 6 percent, set in 1983. Japan’s deficit was nearly as high in 1998, however, reaching 10.7 percent of GDP. Did that jump‐start Japan’s economy? No, it did not.
The Obama team must not believe their lavish spending plans will do anything useful. They have claimed that spending an extra $825 billion over two years will add or save 3 million jobs. Those had better be terrific jobs, because the cost to taxpayers amounts to $275,000 per job. Why spend so much for so little? Deep recessions are invariably followed by strong recoveries, without gargantuan federal spending schemes. In fact, a 1999 study by Christina Romer showed that the average length of recessions from 1887 to 1929 was 10.3 months — without any Keynesian spending schemes — while the average recession from 1948 to 2000 lasted 10.7 months.
After the 1975 recession, employment grew by 6.2 million jobs in two years and 10.2 million in three. After the 1981–82 recession, employment grew by 5.5 million jobs in two years and 10.1 million in three. The population was much smaller in the past, making Obama’s target of 3 or 4 million jobs appear even less ambitious.
Before rushing to add another trillion dollars in TARP and stimulus spending to a deficit already above a trillion, is it too much to ask for some shred of evidence that “fiscal stimulus” ever worked?
Paul Krugman offers only one dubious success story. He says, “The Great Depression in the United States was brought to an end by a massive deficit‐financed public works program, known as World War II.” But in “What Ended the Great Depression?” Christina Romer found that “monetary developments were very important and fiscal policy was of little consequence.… Even in 1942, the year that the economy returned to its trend path, the effects of fiscal policy were small.” Sending most young men overseas to fight certainly reduced the unemployment rate, but wartime price controls and rationing exaggerated measures of real income during the war.
What about post‐war fiscal policy? Alan Auerbach of the University of California at Berkeley surveyed that topic in 2002, concluding that there is “little evidence these effects [of fiscal policy] have provided a significant contribution to economic stabilization, if in fact they have worked in the right direction at all.”
Andrew Mountford of the University of London and Harald Uhlig of the University of Chicago have a new statistical study of the U.S. fiscal experience, “What Are the Effects of Fiscal Policy Shocks?” (It can be found at nber.org.) They compare their results with several of the latest studies, including one co‐authored by Romer: