Viewpoint: Sorry, Keynesians, More Spending Doesn't Boost Jobs

By W. MICHAEL COX AND RICHARD ALM

Political leaders continue to peddle the snake oil that we can spend our way back to prosperity.

The Obama administration has pushed government's share of GDP past 40%, the highest ever without a major war. Europeans are grousing about austerity, seeing crippling debt not as a comeuppance but as an obstacle to the spending needed to revive a moribund economy.

In the 1930s, with the world in the Great Depression's death grip, British economist John Maynard Keynes argued that massive government spending would boost demand and put the unemployed back to work. Over the next eight decades, Keynesian stimulus became the standard remedy for weak economies — even though it has never worked.

To test the efficacy of Keynesian policies, we looked at the annual changes in U.S. government spending as a share of gross domestic product from 1901 to 2011, measured relative to the growth trend of 1.76%. Then we determined whether the higher spending had lowered unemployment rates (see chart above).

Many Americans believe President Franklin D. Roosevelt's Keynesian conversion beat back the Great Depression. It's pure myth. In the 1930s, the United States doubled government outlays relative to GDP. The unemployment rate didn't fall; instead, it jumped from 3.2% in 1929 to 25.2% in 1933 — an outcome contrary to Keynes' doctrine.

Yet the policy's failure hasn't fazed Keynes' acolytes. They argue that U.S. policy was too timid and even more government spending was needed to cure the Depression. They point to World War II, where government spending rose to 50% from 20% of GDP and unemployment fell to 1.2%.

Only two periods of rising government spending have been associated with falling unemployment — 1917-19 and 1941-45. They're both times of major world wars, where millions of adults were plucked from the civilian labor force to serve in the military.

The share of the adult population on active duty rose from 0.3% in 1916 to 4.5% in 1918 and from 0.5% in 1940 to 12.3% in 1945.

In short, unemployment fell not because of government spending but because of government conscription — hardly a good way to cure joblessness or evidence of a Keynesian miracle.

At all other times during this 110-year sweep of U.S. history, government spending and unemployment rates have moved in the same direction. In the 1920s, both trended downward. The Depression decade saw them rise in tandem.

From the 1950s through the 1970s, spending and unemployment moved up and down together. In the 1980s-90s, they had another nicely choreographed decline.

Keynesian proponents could claim the positive correlation stems from increases in government spending to create jobs as unemployment rises. However, the pattern persists even with a lag, meaning that government spending programs have made unemployment worse.

So now we come to our current Keynesian episode. An $825 billion stimulus package, passed in February 2009, authorized spending for infrastructure, health care, education, energy efficiency, scientific research and dozens of other projects. Rebates and tax cuts sought to rouse skittish consumers.

At the time, proponents projected that the stimulus would keep the jobless rate below 8%. While government went on its pending spree, unemployment kept rising, peaking at 10% in October 2009. After three years of the stimulus, funded with borrowed money, unemployment hadn't yet gone down to 8%.

Yet another Keynesian policy failure hasn't led to a reappraisal of the belief that spending can solve the economy's problems.

To some, the slow recovery means that the stimulus wasn't strong enough, or the economy was in worse shape than anyone thought. Neither argument faces the glaring truth — Keynesian stimulus doesn't work the way its adherents say it does.

Keynesian-style demand stimulus assumes businesses receiving new orders will quickly go out and add workers.

Why might firms not do that?

First, they may regard the new demand as temporary and choose to squeeze more productivity out of the existing workforce rather than incur the cost of hiring and training new employees. Output per employed worker rises during recessions, providing strong support for this notion.

Second, unemployed workers may lack the skills and training to perform the tasks newly demanded in business. This is particularly true when massive technological change makes the job skills of yesterday obsolete.

Government programs that subsidize the unemployed only make the problem worse by enabling people to stay outside the workforce longer, where their skills atrophy or fall further behind the needs of the new workplace.

Third, more government spending ultimately means higher taxes. Many households will prepare to meet their future obligations by saving more and consuming less. So government spending crowds out private spending, negating policymakers' attempts to increase overall demand.

Fourth, and perhaps most important, demand stimulus doesn't create jobs. Firms are in business to make a profit, not to increase employment. They'll add workers only if it's the profitable thing to do. If salary plus benefits are too costly, the company will not hire.

Too often, taxes and government-mandated benefits saddle firms with substantial hiring costs, blocking the incentive to hire people to meet any demand, permanent or transitory.

In short, demand doesn't create jobs; incentives do. Nothing in massive government spending addresses incentives to hire.

Stimulus is doomed to fail.

Click here to comment on this viewpoint.

# # #