Stimulus by Spending Cuts: Lessons from 1946
Many, probably most, Americans are skeptical of the vast stimulus efforts the federal government has undertaken in an effort to alleviate the economic downturn. After all, through early 2010, employment has fallen by 8.4 million jobs despite passage of two stimulus bills totaling nearly one trillion dollars in early 2008 and 2009, passage of the $700 billion Troubled Asset Relief Program (TARP), and the extraordinary expansionary monetary actions by the Federal Reserve. But now with serious anxiety regarding the impact of the nation's unprecedented deficits and a potential surge in inflation, a second concern is arising: would any nascent recovery be thwarted if the government was to withdraw the stimulus and return to a semblance of financial normalcy? There's good news on that point. Just as history tells us that stimulus packages are ineffective in bringing about recovery, so it also tells us that "de-stimulus" — moving in the direction of monetary and fiscal contraction — likewise need not have severe adverse effects on employment, income, stock prices, and other macroeconomic variables.
The Obama administration projects a $1.6 trillion budget deficit — almost 11 percent of our GDP — for the 2010 fiscal year. This deficit is the size of total federal spending just 13 years earlier (1997). And this follows a 2009 fiscal year deficit of over $1.4 trillion. At the same time the Federal Reserve has injected another $1.5 trillion in liquidity through various lending programs since the Great Recession began in late 2007. We might call this the "Great Stimulus," but those words are terribly misleading. It hasn't been much of a stimulus, given the rise in unemployment to double digits for only the second time since the 1930s and the general lack of confidence economic agents seem to have in the future economy (the conference board's "Present Situation Index" of consumer confidence hit its lowest level in 27 years in February 2010). Nor is it all that "great": when compared to the size of the economy, the recent stimulus does not even begin to approach that of World War II.
Between 1943 and 1945 government deficits ranged between 21 and 27 percent of GDP — in comparison to the size of today's economy this would be the equivalent of annual deficits of around $3 trillion to $4 trillion. During these three years, the national debt rose from 50 percent of GDP to over 120 percent. Furthermore, the United States Bureau of the Budget estimated that at the wartime peak 45 percent of the nation's civilian labor supply was supported by government spending on the war effort while another 12 million citizens (18 percent of the total labor force) were employed directly by the military.
Of course it is often said that World War II provides the empirical proof that a Keynesian-style government stimulus can bring an ailing economy back to full employment. During the 1930s, the argument goes, government simply did not spend enough to end the Great Depression. AfterPearl Harbor, policymakers finally put the stimulus pedal to the metal with massive deficit spending and highly expansionary monetary policy — the money supply doubled between 1941 and 1945 — to finance wartime production. Unemployment fell from nearly 20 percent in the late 1930s to 3.1 percent in 1942 and 1.2 percent in 1944. John Maynard Keynes himself implied that the return to full employment in the face of massive expansionary policy validated his theory, saying that economic "good may come out of evil" if we heeded the lessons of the wartime stimulus by using the same methods to combat downturns during peacetime.
But the real economic lesson to come out of the World War II era was not that the conscription of nearly a fifth of the labor force into grueling and dangerous working conditions abroad and the imposition of a command economy at home — complete with rationing, price controls, and government allocation of many aspects of life — could bring unemployment down. Soviet-style command economies had many problems, but unemployment was not typically one of them.
Instead, the true lesson from the period can be ascertained from the events of 1945-1947 when the largest economic "stimulus" in American history was dramatically and quickly unwound, months before most people anticipated it (because the atomic bomb brought a sudden unexpected end to the war). No other episode more clearly supports the notion that the best economic stimulus is for the government to get out of the way.
THE DEPRESSION OF 1946
Historically minded readers may be saying, "There was a Depression in 1946? I never heard about that." You never heard of it because it never happened. However, the "Depression of 1946" may be one of the most widely predicted events that never happened in American history. As the war was winding down, leading Keynesian economists of the day argued, as Alvin Hansen did, that "the government cannot just disband the Army, close down munitions factories, stop building ships, and remove all economic controls." After all, the belief was that the only thing that finally ended the Great Depression of the 1930s was the dramatic increase in government involvement in the economy. In fact, Hansen's advice went unheeded. Government canceled war contracts, and its spending fell from $84 billion in 1945 to under $30 billion in 1946. By 1947, the government was paying back its massive wartime debts by running a budget surplus of close to 6 percent of GDP. The military released around 10 million Americans back into civilian life. Most economic controls were lifted, and all were gone less than a year after V-J Day. In short, the economy underwent what the historian Jack Stokes Ballard refers to as the "shock of peace." From the economy's perspective, it was the "shock of de-stimulus."
If the wartime government stimulus had ended the Great Depression, its winding down would certainly lead to its return. At least that was the consensus of almost every economic forecaster, government and private. In August 1945, the Office of War Mobilization and Reconversion forecast that 8 million would be unemployed by the spring of 1946, which would have amounted to a 12 percent unemployment rate. In September 1945, Business Week predicted unemployment would peak at 9 million, or around 14 percent. And these were the optimistic predictions. Leo Cherne of the Research Institute of America and Boris Shishkin, an economist for the American Federation of Labor, forecast 19 and 20 million unemployed respectively — rates that would have been in excess of 35 percent!
What happened? Labor markets adjusted quickly and efficiently once they were finally unfettered — neither the Hoover nor the Roosevelt administration gave labor markets a chance to adjust to economic shocks during the 1930s when dramatic labor market interventions (e.g., the National Industrial Recovery Act, the National Labor Relations Act, the Fair Labor Standards Act, among others) were pursued. Most economists today acknowledge that these interventionist polices extended the length and depth of the Great Depression. After the Second World War, unemployment rates, artificially low because of wartime conscription, rose a bit, but remained under 4.5 percent in the first three postwar years — below the long-run average rate of unemployment during the 20th century. Some workers voluntarily withdrew from the labor force, choosing to go to school or return to prewar duties as housewives.
But, more importantly to the purpose here, many who lost government-supported jobs in the military or in munitions plants found employment as civilian industries expanded production — in fact civilian employment grew, on net, by over 4 million between 1945 and 1947 when so many pundits were predicting economic Armageddon.
Household consumption, business investment, and net exports all boomed as government spending receded. The postwar era provides a classic illustration of how government spending "crowds out" private sector spending and how the economy can thrive when the government's shadow is dramatically reduced.
Employment is closely related to the productivity-adjusted real wage. When the labor costs of making a widget fall, employers find it profitable to make more widgets and hire more widget-makers. Those costs fall when productivity rises (more widgets produced per hour of work), when the price of widgets rises (increasing the margin between revenues received and cost of production), or when money wages fall. In the immediate postwar era, prices and productivity were generally rising, more than offsetting modest increases in money wages.
The data today suggest that the selfcorrecting and healing forces of markets are beginning to work again. Worker productivity is generally increasing, and money wages are stagnant or rising less than the rate of inflation, meaning real wages are falling. In a productivity-adjusted sense, the wage decline appears to be substantial.
After a lag to be sure this trend is real and sustaining, this should lead to an upsurge in new hiring. In other words, unemployment will start falling not because of the stimulus spending, but in spite of it. And just as the stimulus money created few if any new jobs, its withdrawal will destroy few if any jobs. To be sure, some specific jobs will be lost, but others will be gained as the negative effects of government borrowing are eased somewhat. To better illustrate the crowding out effect of government spending, economists often refer to FrÃ©dÃ©ric Bastiat's 1848 essay, "What Is Seen and What Is Not Seen."
The illusion that new employment results from the stimulus package is understandable because the jobs created by it are visible, whereas jobs lost due to the stimulus are much less transparent. When several hundred million dollars are spent building a 79-mile per hour railroad from Cleveland to Cincinnati, we will see workers improving railroad track, building new rail cars, and so on. In fact, we can directly count the number of jobs supported by stimulus dollars and report them on a website (www.recovery.gov currently reports that 608,317 workers received stimulus monies in the 4th quarter of 2009). At the same time, however, the federal spending invisibly crowds out private spending.
This happens regardless of how higher federal spending is financed. Tax financing (not done in this case) reduces the after-tax return to workers and investors, leading them to reduce the resources they provide. Deficit-financing (borrowing) tends to push up interest rates and, more generally, eats up dollars that would otherwise have gone toward private lending and investment. Inflationary financing (roughly the Fed printing money — a fear in this situation) reduces investor confidence, lowers the real value of some financial assets, and leads to falling investment. Of course we do not register these "job losses" on the mainstream statistical radar because they are jobs that would have been created, absent the government spending, but never were — hence their invisibility.
There are no free lunches in the world. Stimulus efforts of modern times, perhaps most notably that of Japan during the 1990s, which actually led to reduced economic growth and long-term higher unemployment, show the futility of the Obama administration's current approach. Furthermore, a recent study by Claudia Sahm, Matthew Shapiro, and Joel Slemrod shows that the Bush stimulus policies in 2001 and 2008 ad no significant impact on the economy. Other recent work by Robert Barro and Charles Redlick examines long-term macroeconomic data and confirms the notion that government spending crowds out that of the private sector. Barro predicts that the long-term effect of the current stimulus will be negative.
Markets, by contrast, have marvelous healing properties. If unemployment is too high, declines in the productivityadjusted real wage make it attractive to hire workers again, lessening the problem. If investors are slow in borrowing, falling interest rates entice them to take on credit. These sorts of things are happening in the American economy today, but government- imposed shocks can derail any recovery. This happened in the Great Depression as the economy finally began to recover after a major slowdown in government interference in the labor market between mid 1935 and early 1937. However, these gains were reversed by the Supreme Court's surprise ruling (which followed Roosevelt's threat to pack the Court) upholding the constitutionality of the National Labor Relations Act. Real wage rates rose sharply in the months that followed.
Unemployment, which had fallen to around 13 percent on the day of the court ruling, was back above 20 percent a year later. When market processes lead us to see light at the end of the tunnel, the government sometimes adds more tunnel. Recent examples of this phenomenon can be seen in the newly passed health care legislation and the proposal for a cap-andtrade environmental regime. The new health care legislation will enormously increase labor costs, as would cap and trade. Nervous employers, wanting to avoid the possibility of taking on sharply rising labor expenses, demur in hiring workers that they would in a more neutral policy environment.
Furthermore, the multi trillion dollar eficits to finance the stimulus as well as government bailout money from TARP have to be financed, and the possibility that the Federal Reserve would engage in inflationary financing of this new federal debt has clearly unnerved many investors. Since the November 2008 election, the price of gold has risen 50 percent because of growing inflationary fears. Yet another example is the government's continual extension of unemployment benefits beyond the customary maximum 26 weeks (most recently at the beginning of March).
While most would agree that unemployment insurance provides short term relief to those who must seek new work, many studies confirm what common sense says we should expect — the longer the time frame people are eligible for such benefits, the longer it takes for unemployment rates to fall. In 2009 the average duration of unemployment nearly doubled, and today, well over 40 percent of those unemployed have been out of work over six months. While the poor labor market is to blame for much of this jump in duration, there can be no doubt that incentives to obtain new employment have been, and will continue to be, tempered by governmental action which has extended unemployment insurance to many through the end of 2010.
Finally, it is clear that the government stimulus has not provided any kind of positive placebo-type effect on consumer and business confidence. As mentioned earlier, survey data show that such measures of confidence continue to linger around the lowest levels seen in a generation. In fact, a simple econometric model consisting of two explanatory variables — government spending as a percent of total output and the rate of inflation, can explain the vast majority of the changes in stock market prices in modern times — and stock market valuations are a good indicator of confidence.
Stock prices fall with growing government involvement in the economy or with rising inflation. The sharp rise in the government's share of output in the last decade and the threat of greater inflation in the next one are important factors behind the 30 percent decline in the inflation-adjusted Dow Jones Industrial Average since 2000. Eye-popping deficits of the past year have lowered optimism about the future, kept stock prices depressed, and reduced key elements in new investment spending. These negative side effects of the stimulus spending are certainly slowing down the recuperative process that market forces are attempting to generate.
The conversation has begun regarding the nation's exit strategy from the unsustainable fiscal and monetary stimulus of the last two years. Our soaring national debt will not only punish future generations but is also causing concern that our creditors may bring about a day of reckoning much sooner (the Chinese have recently become a net seller of U.S. government securities). There are fears that the Fed's policy of ultra-low interest rates may bring new asset bubbles and begin the cycle of boom and bust all over again.
And unless the Fed acts to withdraw some of the monetary stimulus, many fear a return of 1970s era double-digit inflation. On the other hand, there are widespread fears that if we remove the stimulus crutch, the feeble recovery may turn back toward that "precipice" from which President Obama has said the stimulus policies rescued us. History and economic theory tell us those fears are unfounded. More than six decades ago, policymakers and, for the most part, the economic profession as a whole, erroneously concluded that Keynes was right — fiscal stimulus works to reduce unemployment. Keynesian- style stimulus policies became a staple of the government's response to economic downturns, particularly in the 1960s and 1970s.
While Keynesianism fell out of style during the 1980s and 1990s — recall that Bill Clinton's secretary of treasury Robert Rubin turned Keynesian economics completely on its head when he claimed that surpluses, not deficits, stimulate the economy — during the recessions of 2001 and 2007-09 Keynesianism has come back with a vengeance. Both Presidents Bush and Obama, along with the Greenspan/Bernanke Federal Reserve, have instituted Keynesian-style stimulus policies — enhanced government spending (Obama's $787 billion package), tax cuts to put money in people's hands to increase consumption (the Bush tax "rebate" checks of 2001 and 2008), and loose monetary policy (the Federal Reserve's leaving its target interest rate below 2 percent for an extended period from 2001 to 2004 and cutting to near zero during the Great Recession of 2007-09 and its aftermath).
What did all of this get us? A decade far less successful economically than the two non- Keynesian ones that preceded it, with declining output growth and falling real capital valuations. History clearly shows the government that stimulates the best, taxes, spends, and intrudes the least. In particular, the lesson from 1945-47 is that a sharp reduction in government spending frees up assets for productive use and leads to renewed growth.
Jason E. Taylor is professor of economics at Central Michigan University. Richard K. Vedder is distinguished professor of economics at Ohio University and adjunct scholar at the American Enterprise Institute.
This article originally appeared in the May/June 2010 edition of Cato Policy Report.