All Commentary
Friday, October 23, 2009

Old, Bold Futility

In economic analysis and policy formulation, profundity is not to be confused with complexity. And simple logic is not the same as simplicity. Reliance in thought and communication on shortcut slogans and mottos yields not solution but fiasco.

With employment slumping, many would have us believe in a simplistic “bold economic recovery program.” With vast public-works projects to build and rebuild so-called infrastructure—highways, bridges, sewers—we could allegedly create jobs.

We may have neglected our infrastructure. Perhaps we would be more productive if we invested more in maintaining and expanding such basic capital. But the condition of our infrastructure commonly is not the central issue. The more immediate question is whether government spending would increase total employment.

Public-works programs clearly would directly create specific jobs. But expansion of public-works jobs is likely to cause jobs to contract elsewhere, with little net change in aggregate employment. Other employment will contract because federal expenditures on public-service jobs must somehow be financed. If the federal government is to increase spending for public works, then it must reduce other government expenditures or raise taxes or deficit-spend.

If government reduces spending for other programs, such as national defense, then fewer workers will produce guns. Or if government reduces transfer payments to selected individuals, then fewer workers will produce the butter that beneficiaries would have demanded. In either case, decreases in federal expenditures in one area to finance increases in public works will not increase total employment.

Instead of curtailing other spending in order to increase expenditures on public works, government might raise taxes and thereby curtail private consumption and investment. With private spending down and government spending up, private jobs are traded for public jobs, and total employment is not increased.

If, alternatively, government finances the expansion of public works by enlarging its already enormously swollen deficit, then it will either have to borrow existing money or create new money to pay its increased bills. If it borrowed existing money from the community, government would compete more intensely with businesses and households for financial resources. More money would then go to government and less to private firms and individuals. Government spending of borrowed funds would displace private spending of private funds. Again, some jobs are traded for others.

Finally, what if government financed its increased deficit spending by creating new money? In this instance, spending need not fall in other areas, so new public-works jobs would not reduce other employment—initially. But down the line, increased money creation, if continued fast enough long enough, will raise prices. And higher prices—a pernicious form of taxation—would threaten future productivity and employment. We would then obtain a bit more force-fed employment today at the expense of less employment and more misery tomorrow.

Inflation is not a function solely of rapidly increasing money stock, to be sure. The rate at which money is spent—the velocity of monetary circulation—also affects aggregate community expenditure. Velocity has trended downward over the last decade and fell abruptly in late 2008, which partially offset the rising amount of money. But in a period of substantial inflation, we will see both the amount of money and its rate of spending rise. When expenditure goes up faster than the production of goods—and it is vastly easier to create money than to create goods—we have inflation.

No matter how government finances it, greater spending for public works is not likely to increase the total number of jobs significantly. Economic thinking leads to this conclusion, but does history support it?

More Spending, No More Jobs

The economy is not a laboratory in which experimentation readily yields conclusive evidence. But we can look at how past government spending in general has related to total employment.

Data for the last 80 years indicate that big increases in federal spending have not typically been associated with large increases in employment. Indeed, the loose relationship that existed over those decades is quite the opposite. Faster growth of employment has been associated with slower growth of federal spending, and slower growth (or reduction) of employment was more likely when federal spending grew faster.

From 1929–33 government outlays increased by almost half—over 48 percent—while unemployment surged more than eightfold.

In the New Deal period of 1933–39, government spending rose at an annual rate of over 12 percent, doubling in just six years, while employment increased at only 2.5 percent. The unemployment rate was nearly 25 percent in 1933, and was still more than 17 percent in 1939. The then secretary of the Treasury acknowledged in 1939 the basic failure of the spending program.

In 1962–66 and in 1975–81, employment expanded at an annual average of 2.2 percent in each period, but government spending rose 6 percent in the earlier years and at more than double that rate in the later period.

From 2001 to the present, government outlays have risen at an annual rate of nearly 7 percent, while employment has expanded at a measly 0.8 percent. In the past year and a half, employment actually fell.

No one has found a significant, consistently positive relationship between government spending and total employment in the United States or any other country. The reason is simple. The money spent by government has to come from somewhere. Certainly, there are things that can be done to promote a vigorous economy. But it should be apparent that government cannot simply and surely spend us to rapidly expanding employment.

  • William R. Allen is an economist, professor and author. He is known for his authorship of economic literature alongside frequent co-author Armen Alchian.
    Allen obtained his A.B. (Bachelor of Arts) from Cornell College (1948) and his Ph.D. from Duke University (1953).

    He instructed at Washington University prior to joining the UCLA faculty in 1952. He has been a visiting professor at Northwestern University, the University of Wisconsin, the University of Michigan, Southern Illinois University, and Texas A&M University, and he has been on the faculty of the Colorado School of Banking.