[page-number of print edition: 37]


Theodore Caplow, Charlottesville

Figure 1 shows the U.S. unemployment rates from 1900 to 1998, covering all but one year of the twentieth century. During the first part of the century, unemployment oscillated wildly - especially from its low point in 1926 in the midst of hectic post-war prosperity to its peak in 1933 when slightly more than one-fourth of the labour force was counted as unemployed. The rate

Figure 1

The U.S. unemployment rate: 1900-1998

[page-number of print edition: 38]

advanced to another low point in 1945 when the effect of withdrawing thirteen million people from the labour force for service in the armed forces and the extreme manpower demands of wartime industry reduced unemployment nearly to zero. Since 1950 the fluctuations have been much milder and in the post-war decades there was a consensus among American economists that a minimum unemployment rate of at least 5 percent was necessary to avoid runaway inflation. As it turned out when unemployment dipped below 5 percent in 1997 and stayed below 5 percent throughout 1998, inflation remained close to zero and none of the projected ill effects appeared.

Before and after the Great Depression of the 1930s, unemployment was almost exclusively a blue-collar affliction. In 1924 with the general rate of unemployment at a very low level, it was nevertheless reported that nearly two-thirds of the male factory workers in a sample of families in Middletown, the Midwestern industrial city studied by Robert and Helen Lynd, had at least one spell of unemployment during the first nine months of 1924. None of the white-collar workers in their sample had that experience (Lynd and Lynd 1929). Blue-collar workers were susceptible to lay-off without notice as, with certain limited exceptions, they still are in the U.S. while managers, professionals, and other white-collar workers were effectively immune from dismissal without cause and remained so until a comparatively recent time.

Figure 2 compares the unemployment experience of the United States with that of the three large developed nations which are closest to the U.S. in political culture and in level of modernisation. What leaps out from the figure are the similarities in the recent experience of these four nations? All of them show less oscillation of unemployment in the second half of the century than in the first. All of them experienced comparatively low unemployment in the 1960s and a renewed surge of unemployment in the 1980s. During the early part of the century the discrepancies were more conspicuous. The U.S. and the UK experienced high unemployment immediately after World War I.

[page-number of print edition: 39]

France and Germany, with their labour forces significantly reduced by wartime casualties, did not. The U.S. and France both experienced a surge of prosperity in the middle 1920s that reduced their unemployment rates to the low points of the century

Figure 2

Unemployment rates of large developed nations

[page-number of print edition: 40]

while Germany and the UK were suffering relatively severe economic dislocations. Similarly we note that unemployment in France was not significantly raised by the Great Depression and that Germany experienced a surge of unemployment after World War II that was not matched by the victorious allies. The most important discrepancy that appears in Figure 2 is that for most of the century the posted American unemployment rate was higher -sometimes very much higher - than those of the other three countries. The differences were particularly marked in prosperous years. In 1962, for example, with the German unemployment rate at an almost invisible 0.5 percent, the French rate at 1.3 percent, and the British rate at 2.9 percent, the U.S. was celebrating what was then considered a very favourable rate of 5.6 percent.

The reversal of the U.S. situation that took place in the early 1980s with respect to France and the UK and that persisted into the 1990s when Germany, too, developed a severe problem of long-term unemployment is of historic significance. According to Pugliesi (1998) „the 1990s has seen a general worsening of the [European] labour market situation: unemployment has hit countries such as the Federal Republic of Germany and Sweden and has kept increasing in Southern European countries, e.g. Italy and Spain".

Both the relatively higher American rates of earlier years and the relatively lower American rates that prevail today are closely linked to differences in employer/employee relationships among these countries, particularly the sharp contrast between traditional provisions for employment security in Britain and France and the virtual absence of such provisions in the United States , with the UK occupying an intermediate position. It often comes as a surprise to foreign observers to learn that ordinary workers in American enterprises can be, and often are, dismissed without notice after years of steady employment - a procedure that would be legally and politically impossible in other advanced industrial countries. During that long period of time when American em-

[page-number of print edition: 41]

ployment rates were prevailingly higher than those in peer economies, it was widely believed that employment insecurity was the price that the U.S. paid for its relatively high Gross National Product.. The argument was that the lack of employment security made it easier to restructure enterprises and to maximise their productivity. Europeans during that same period pointed to the social benefits of employment security while acknowledging the possibility that job tenure might, under some circumstances, be an impediment to economic growth.

Recent developments have somewhat shifted the ground under these views. On the one hand, employment security has conspicuously lessened in the United States for several different reasons. In the early part of the century, blue-collar workers could be, and often were, laid off and rehired from week to week. White-collar workers, especially middle managers in large-scale enterprises enjoyed virtual life tenure; dismissals were rare and only for grave cause. That situation changed abruptly in the restructuring frenzy that seized the American economy in the 1980s when thousands of corporate mergers occurred and almost every corporate merger involved down-sizing that affected the managerial and administrative cadres. Suddenly, middle-aged middle managers were being dismissed with as little notice as factory workers. Meanwhile factory workers, too, have been losing rather than gaining job security as the bargaining position of labour unions has declined nearly to the vanishing point . The demand for guarantees of long-term employment that loomed so large on the want lists of union negotiators twenty-five years ago have been largely abandoned. Only a few sectors of the economy now offer as much job security as formerly, notably the government service. But here, too, the passion for down-sizing has taken hold. While outright dismissals of government employees are rare, the constriction of job and promotion opportunities by leaving positions vacant and reduc-ing administrative levels is commonplace. The armed forces represent a particularly interesting case. Soldiers were formerly of

[page-number of print edition: 42]

fered premiums for reenlisting when their terms of service expired but in the past few years it has become routine to deny reenlistment to soldiers with excellent records in order to reduce the size of the armed forces and to impose involuntary retirement on officers with excellent records for the same reason.

Given these circumstances, one would expect to find high unemployment in the United States, or at least higher rates than those prevailing in national economies that still offer a relatively high degree of job security. How does it happen, then, that the latest U.S. rates are so low in relation to the French, German, and British rates and how, for that matter, do we explain the absence of those inflationary pressures that economists unanimously predicted if unemployment in the U.S. should ever drop below 5 percent?1

To resolve these mysteries we need to look more attentively at the U.S. unemployment rate, understand how it is calculated, and see why it cannot be safely compared with the rates of other countries unless a number of contextual factors are taken into account. A national system of unemployment compensation was not introduced in the United States until 1935, much later than in the three nations we are using for comparison. It was launched, not as a federal programme, but as a state programme mandated by the federal government. By 1937 all forty-eight states, the then territories of Alaska and Hawaii, and the District of Columbia had passed the requisite laws. All contributions collected by the states are deposited in an unemployment trust fund in the U.S. Treasury, but each state maintains a separate account. A state may withdraw money from its account only to pay unemployment benefits. Each state has independent responsibility for the content and development of its unemployment insurance scheme. A network of joint federal/state bureaus called employment commissions administer unemployment benefits while acting as employment agencies for those unemployed who receive insurance benefits and also for those, actually a slight majority, who receive none. In order to be counted as unemployed,

[page-number of print edition: 43]

workers must be ready, able, and willing to work, must be registered as job-seekers with one of the employment commissions, and must accept suitable work if offered. Elaborate rules govern the definition of suitable work and define eligibility. In the state of Virginia, for example, the applicant for unemployment benefits must have earned at least $2,850 over two quarters out of the first four of the last five completed quarters. The maximum benefit payable is $228 per week, a figure which if annualised would put most recipients below the poverty line but, in fact, very few beneficiaries receive the maximum payment and none for an entire year. The average weekly benefit is considerably lower than the maximum and benefits for most recipients are limited to thirteen weeks. They can be doubled to twenty-six weeks in cases of demonstrated hardship, but not further extended. A good deal of paperwork must accompany the application. The recipient must submit to the employment commission during every week in which he or she receives benefits a certified record of having applied for work. The minimum is two serious contacts a week, and these records are frequently checked. The system is self-supporting. Unlike other entitlements, unemployment benefits are paid for entirely by a direct tax on employers and employers are taxed in proportion to the number of their former employees who apply for benefits.

The count of unemployed is not restricted to the people who manage to collect benefits under these rather stringent conditions, but it is limited to those who are registered with the employment commissions as actively looking for work and can demonstrate to the satisfaction of officials that they have been doing so in good faith. People who cease to search for jobs are no longer counted as unemployed. Thus, in the U.S. statistics, long-term unemployment is not captured at all. It does not exist. Discouraged workers or those for whom there is no plausible reemployment opportunity are likewise excluded from the count. Consequently, it may be fairly said that the U.S. rate understates

[page-number of print edition: 44]

unemployment in comparison with the rates of most other countries, but that assertion remains essentially empty unless the particular rules of calculation that are used in a given other country are taken into account. While most European countries are less restrictive than the U.S. in counting the unemployed, their calculation rules are far from uniform among themselves.

Another factor affecting the comparability of the American rate with those of other countries is the relatively large size of two groups who are not counted in the labour force at all and thus necessarily excluded from any count of the unemployed. Somewhat more than two million men and women, mostly men, currently serve in the U.S. Armed Forces. Nearly two million men and women, mostly men, are currently confined in U.S. prisons and jails. Both of these populations are much larger proportionately than the equivalent populations in other advanced industrial nations. Such considerations do not deprive Figure 2 of informative value because the downward distortion to which the U.S. rate is subject has been in place for more than sixty years, and the changing relationship of the U.S. trend to those of other countries is not much affected by the realisation that comparison in absolute terms is difficult.

Even if the absolute level of the U. S. unemployment rate is questionable because of the omission of long-term unemployed and discouraged workers and the exclusion of special populations, it is still very interesting to decompose the general rate and examine various sub-populations. We noted, for example, that within the past two decades white-collar workers have lost their former immunity to arbitrary dismissal. But it is equally important that blue-collar workers today still experience about twice as much unemployment risk as white-collar workers and that within the white-collar groups, sales and clerical personnel have about twice the risk of managers. Education, race, and age generate other differences. High school dropouts have about twice the risk of high school graduates who in turn have about twice the risk of college graduates who in turn have a significantly

[page-number of print edition: 45]

higher risk of unemployment than the holders of advanced degrees. The higher educational qualifications come close to providing complete immunity from unemployment. Age is a factor, as well. Men younger than twenty-four have about twice the risk of unemployment of men over twenty-four, although a particular young man’s exposure to unemployment is strongly affected by school enrolment and school leaving and by his enlistment in or discharge from the armed forces (Mare, Winship and Kubitschek, 1984). Women under twenty-four have about three times the risk of unemployment of older women. That particular comparison is complicated by the circumstance that women, although now constituting nearly half of the American labour force, move in and out of the labour market much more frequently than men and that the question of whether a given woman is unemployed or has merely returned to household duties is often impossible to resolve. Marriage, like education, provides considerable protection against unemployment for both sexes. Race and ethnicity are powerful factors. Blacks have nearly twice the unemployment risk of whites. That gap has persisted with little change for many years (Sundstrom, 1997). Moreover, there is evidence that the impact of unemployment on family stability, especially the risk of divorce or separation, is significantly greater in black than in white families (Starkey, 1996). These sub-group differences are cumulative so that, for example, white male college graduates at age forty-five have an unemployment rate close to zero while black high school dropouts at age nineteen have a rate close to 50 percent. The subject of differential unemployment risks has been well explored in a recent dissertation by Wang (1998) and by Jencks (1991), Massey and Fenton (1993).

Social scientists have traditionally recognised four distinct types of unemployment: 1) frictional unemployment; 2) seasonal unemployment; 3) cyclical unemployment; 4) structural unemployment.2. Given the way the U.S. system works, each type impinges somewhat differently on the enumeration of unem-

[page-number of print edition: 46]

ployed persons.

Frictional unemployment occurs when workers change jobs. Even with a high level of labour demand, the movement from one job to another takes time, sometimes quite a lot of time when it involves retraining or migration from one part of the country to another. Hence, even under the most favourable circumstances, there will always be a pool of workers in the process of changing jobs, and in an economy where restructuring is almost continuous, that pool will be an appreciable fraction of the labour force, perhaps as much as 2 percent.

Seasonal unemployment involves a considerable number of diverse occupations ranging from those of very low to very heigh occupational status. For example, the farm labourers who harvest crops are necessarily limited to a few weeks in the working year if that is their sole occupation. Many such labourers are imported from Mexico or Central America and return home after the harvest season so that they cannot be counted as among the American unemployed. A considerable number are legal residents and they can be counted if a) their earnings are sufficient to qualify for unemployment insurance or, b) they sign up to search actively for employment in some other occupation. However, it appears that most seasonal farm workers are never counted among the unemployed during their periods of inactivity if only because so many of them use the opportunity to become illegal residents instead of returning home and are likely to stay clear of official agencies for some time thereafter. At the other end of the scale, established film actors who may work only a few weeks in the year, have no motive to register as unemployed since they seek employment through quite different channels than those offered by the employment commissions and are not attracted by the stingy benefits of unemployment insurance. By contrast, less established actors often depend on unemployment insurance to carry them from job to job. The relatively short benefit period fits their pattern of intermittent employment very well. The same is true of many of seasonal employments in

[page-number of print edition: 47]

summer or winter resorts and in occupations like logging and rodeo riding that can only be pursued in certain seasons.

Cyclical unemployment is of course what produced the enormous peaks displayed by the unemployment rate in the 1930s and most of the lesser peaks and valleys since. There can be no question that the severity of these employment cycles has been lessening, as Figure 2 plainly shows. The United States experienced its most recent surge of cyclical unemployment in 1982-83. Some economists believe that sophisticated fiscal policy has permanently damped down the wild swings of the business cycle. Other economists regard that optimistic view as nonsense. Only time can resolve the question.

Structural unemployment is what seems to be afflicting the major European economies in the past few years (ILO 1996) and perhaps, in a more hidden form, the U.S. economy as well. Structural unemployment involves one or more of the following conditions:

1) There are particular sectors of the labour market from which workers cannot easily move into other jobs. In the early 1980s, for example, the so-called rust belt of heavy industrial districts stretching from Pennsylvania to Illinois suffered a far more serious recession than the rest of the United States and it was particularly characterised by the plight of skilled workers in steel mills and manufacturing plants who could not easily transfer their skills to other kinds of work. A similar problem afflicted the Texas oil fields in the early 1990s and again today.

2) In some sectors with diminishing employment the number of vacancies continues to shrink in relation to the number of job seekers. A notable case extending over two decades has been the decline of employment in the U.S. merchant marine, occasioned by the inability of American ship owners to compete against foreign flag vessels subject to lower taxes, lower wages, and more accommodating safety regulations. As the number of jobs for

[page-number of print edition: 48]

American seaman and sea officers shrank, elaborate measures were introduced to spread the available employment opportunities by, for example, requiring seamen to remain on the beach for the same length of time as their previous sea-going assignment. But as the number of openings continued to decline, most members of this workforce with their specialised skills and long experience, were forced into other lines of work or into long-term unemployment.

3) Technological change, in a highly modernised economy, continuously renders certain types of skill obsolete and requires the holders of those skills to seek other work. But in innumerable instances, the highly skilled worker whose occupation disappears is unlikely to find a substitute occupation that offers equivalent pay and benefits. The union printers are a case in point. Not very long ago they were the aristocrats of blue-collar labour with a powerful union, highly protective work rules, and a strong apprenticeship system. The advent of computerised composition made the linotype machine as obsolete as the spinning wheel and did so almost overnight. Many of the older union printers took early retirement and were never counted among the unemployed. Most of the younger printers eventually found other lines of work and they, too, are not counted although they may reasonably regard themselves as victims of structural unemployment.

The more flexible the economy and the greater the rate of technological innovation, the more structural unemployment will develop and be concealed. The net effect is both good and bad. On the one hand the economy becomes more productive. On the other hand the distribution of earnings becomes more unequal. This brings us to globalisation, the last element in the equation. Globalisation puts the worker in advanced economies in direct competition with workers in less developed economies who demand and obtain much lower wages per unit of output. Within the last ten or fifteen years, entire industries have migrated out of the United States to low-wage regions in Asia and Latin

[page-number of print edition: 49]

America. Very little clothing of any kind is still manufactured in the U.S. Furniture, toys, kitchen ware, farm machinery, household appliances, electronic components that were formerly mass-produced in the U.S. are now manufactured abroad and returned for sale in the American market. The positive consequence is that ordinary consumer goods are much cheaper than they would otherwise be. And from the standpoint of the managers of enterprises, globalisation has an even more important favourable effect. It enables the U.S. economy to maintain a low level of unemployment without generating any inflationary pressure. In other words, the scarcity of job applicants in the domestic market does not induce a demand for higher wages because of the ease with which jobs can be shifted overseas. The negative aspects of this situation are equally apparent. While there is no lack of employment opportunities, average earnings in the U.S. have lagged conspicuously behind rising productivity and continue to do so. McGhee (1996) writes that, „in spite of a prosperous economy, Americans are experiencing the greatest sense of job insecurity since the Great Depression" because of several kinds of structural unemployment.

Thus, we confront the paradox that while the current U.S. economy delights Wall Street and Washington because it combines low unemployment and low inflation, it nevertheless leaves a larger proportion of the national population in abject poverty and a wider gap between the earnings of managers and workers than that of any other advanced economy.


1For sophisticated analyses of the long-term relationship between wage-inflation and unemployment, see Hyclak and Johnes 1997 and Volgy, Schwarz and Imwalle 1996.

2For an old but excellent description of these types, see Gordon 1965.

[page-number of print edition: 50]


GORDON, R. A. (1965) ‘Employment and Unemployment’. International Encyclopedia of Social Science. 5:49. New York: Macmillan.

HUNT, Michael (1988) ‘More Universalism, Less Structural Mobility: The American Occupational Structure in the 1980s.’ American Journal of Sociology, 93:6:1358-1400.

HYCLAK, Thomas and Geraint JOHNES (1997) ‘Wage Inflation and Unemployment’. In: Britain and America 1892-1991’. Labour, 11:2:373-390.

GLOBAL UNEMPLOYMENT CRISIS Continues, Wage Inequalities Rising’ (1996) World of Work 18:12:4-5.

JENCKS, Christopher (1991) ‘The Jobless Male Underclass.’ pp. 39-61. In: C.Jencks and P.E. Peterson eds, The Urban Underclass. Washington: The Brookings Institution.

MARE, Robert D., Christopher WINSHIP and Warren N. KUBITSCHEK (1984) ‘The Transition from Youth to Adult: Understanding the Age Pattern of Employment.’ American Journal of Sociology.

MASSEY, Douglas S. And Nancy A. DENTON (1993) ‘American Apartheid: Segregation and the Making of the Underclass’. Cambridge: Harvard University Press.

MCGHEE, George C. (1996) ‘The Scourge of Unemployment and What Might be Done about It.’ Proceedings of the American Philosophical Society 140:4:459-468.

PUGLIESE, Enrico (1998) ‘New Mass Unemployment and Mobilization in European Cities.’ University of Naples. Available on e-mail from pugliesi@ds.unina.it

STARKEY, James L. (1996) ‘Race Differences in the Effect of Unemployment on Marital Instability: A Socioeconomic Analysis.’: Journal of Socio-economics 25:6:683-720.

[page-number of print edition: 51]

SUNDSTROM, William A. (1997) ‘Explaining the Racial Unemployment Gap: Race, Region, and the Employment Status of Men, 1940.’ Industrial and Labor Relations Review 50:3: 460-477.

WANG, Amy (1998) ‘Gender and Racial Differences in Education and Unemployment: A Structural and Longitudinal Approach.’ Dissertation Abstracts International (Humanities and Social Sciences) 58:11:4449-4450.

© Friedrich Ebert Stiftung | technical support | net edition fes-library | Februar 2000

Previous Page TOC Next Page