Industrial Recovery Act, the Wagner Act, the Fair Labor Standards Act, and legislation creating the Civilian Conservation Corps and the Works Progress Administration are but a few examples of the burst of new federal government initiatives designed to bring about “relief, recovery, and reform”—and lower unemployment—during the New Deal.
The crowning manifestation of government activism was the Employment Act of 1946, which declared the eradication of unemployment a national priority. Public-policy efforts to end unemployment did not end with the Employment Act. Activism peaked in the 1960s and 1970s. Not only was further legislation passed emphasizing the importance of full employment as a national goal (e.g., the Humphrey-Hawkins Act), but numerous new institutions were created as an outgrowth of the War on Poverty (e.g., the Jobs Corps and the Office of Economic Opportunity). To deal with persistent unemployment, a variety of new public-assistance programs, such as Medicaid and food stamps, were created. While the Reagan era of the 1980s brought a stifling of new unemployment initiatives, little was done to dismantle the apparatus of federal programs, and the activist monetary and fiscal policies, that had been created over the previous decades. Just as earlier Republican presidents Eisenhower and Nixon had not attempted to dismantle the New Deal and the Great Society, so President Reagan, by far the most conservative modern American president, did relatively little to undo the “safety net.” The legacy of macropolicy expansionism established in the New Deal era was left largely intact.
While interest in the unemployment problem tended to rise and fall with the business cycle, concern was not exclusively directed to cyclically related unemployment. Thus in the prosperous 1920s, there was mounting concern about technological unemployment. Likewise, the concept of what we now term “frictional unemployment” was developed.3 Similarly, in the equally prosperous 1960s, structural unemployment was a concern—the mismatch of skills of those unemployed with the skills needed for jobs available.
In the 1970s, economists coined a term to refer to the noncyclical forms of unemployment: “the natural rate of unemployment.” Much of the modern discussion about employment is related to the natural rate of unemployment and its determinants. Thus the twentieth-century discussion of the unemployment problem is partly a by-product of concerns over economic fluctuations, but also in part of concern over idleness that reflects other noncyclical factors.
Why the rise in interest in unemployment? To a large extent, the issue grew in importance with the urbanization of America. Before the Civil War, most Americans were engaged in farming. Most farmers, in turn, were either self-employed, or slaves who could not become unemployed almost by definition. After 1890 or 1900, cheap or free public land in the West was no longer readily available, and the proportion of Americans working for wages had grown strikingly with industrialization and urbanization. In the nineteenth century, the urbanized proportion of the American population rose from 6 to 40 percent of the total, and it became a majority by 1920.4 Whereas no more than 5 percent of the labor force was engaged in manufacturing in 1800, by 1920 that proportion approached one-third.
The decline in the relative importance of self-employed individuals increased the vulnerability of workers to unemployment, and as a consequence the incidence of involuntary joblessness rose with the passage of time. Although good annual data are unavailable for almost all of the nineteenth century, it is unlikely that the nation suffered from a double-digit unemployment rate, at least on any sustained basis, until the 1890s. It was the Great Depression of the 1930s, however, with a full decade of double-digit unemployment, that led to the overwhelming preoccupation with the unemployment question.
Unemployment increasingly evoked both intense humanitarian concerns over the plight of individuals and broader concerns over the efficiency and effectiveness of the aggregate economy. Workers who lost their jobs “through no fault of their own” often found themselves with a dramatic drop in income and increasing financial pressures. They suffered dramatically, both economically, and also psychologically, as their self-esteem declined with each day of idleness. Nationally, the loss of 10 or more percent of the most productive resource meant a decline in total output, and a reduction in capital formation needed to augment longer-term increases in material well-being.
Variations in Unemployment over Time
The problem of unemployment was not merely that it was growing dramatically in size over time, but that its magnitude varied significantly over time and space, often striking individuals unexpectedly before they could prepare for it. Over the first nine decades of the twentieth century, the official annual unemployment rate averaged 6.6 percent, but varied between 1.2 percent and 24.9 percent.5 Even if one looks at decade averages of unemployment (see table 1.1), the variation in rates is still huge. The 1930s stands out as an extraordinary aberration—mean unemployment rates well over double the next highest unemployment decade (the 1980s). Yet even the 1980s witnessed a mean decade unemployment rate over 50 percent higher than in four other decades (the 1900s, 1920s, 1950s, and 1960s.)
TABLE 1.1 VARIATIONS IN UNEMPLOYMENT IN THE U.S., BY DECADES, 1900–1990
The bad news from table 1.1 is that the lowest mean unemployment rates tended to come early or in the middle of the century: the mean unemployment rate rose steadily by decade after the 1950s, and even the 1950s figure was higher than observed in the very first decade of the century. Whereas unemployment rates of 5 percent or less tended to be the rule rather than the exception in the period 1900–29 and also in the era from 1940 to 1969, by the 1980s unemployment rates that low simply did not occur.
The good news from table 1.1 is that the absolute variation around the mean or median unemployment rate was lower in the last half of the century than in the first half, reaching a nadir in the 1960s, when the highest unemployment rate for the decade was a mere 3.2 points above the lowest rate, and the standard deviation barely exceeded one. The good news with respect to unemployment variation is tempered, however, by the fact that moderate increases in fluctuations were observed after the 1960s.
Moreover, the official data for the first decades of the century were constructed from decennial census “benchmark” data, and are thus subject to considerable error. Christina Romer has made a compelling case that the variation in unemployment in the early decades of the twentieth century was much less than the official numbers indicate, and that the postwar unemployment instability is understated, suggesting that the observed reduction in unemployment instability shown in table 1.1 is in fact in large part a statistical artifact.6
It is tempting to try to draw broad-brush conclusions about trends in unemployment by comparing data for various subperiods in the twentieth century. The conclusion one reaches, however, is highly sensitive to the periods chosen. If one arbitrarily divides the ninety years of available data in half, the evidence shows that the mean unemployment rate in the first half, 1900–44 (7.9 percent), is considerably higher than in the last half, 1945–89 (5.49 percent). Moreover, unemployment variation (as measured by the standard deviation), is dramatically lower in the latter period. On the basis of this evidence, the inclination is to accept the conventional wisdom that the unemployment record has been clearly superior in the period of governmental activism that followed the Employment Act of 1946.
Yet one could just as rationally divide the twentieth century into three equal periods of thirty years each: 1900–29, 1930–59, and 1960–89. The first period is an era of very limited direct governmental involvement impacting on unemployment, the middle period is a transitional era where an interventionist policy came to be increasingly accepted, and the latter period is one of consistent
