West Virginia has long been associated with high unemployment, much to the detriment of its residents and to the dismay of government and business. Unemployment has helped to create one of the highest levels of poverty in the nation, and has arguably done immeasurable harm to the quality of life for several generations of West Virginia residents. State and local governments have consequently been repeatedly forced into a defensive position, struggling not only to provide some support for those affected by unemployment but also to provide the basic services required by the population at large in the face of an unemployment-induced declining tax base. The unemployed cannot pay much in the way of taxes, and the out-migration of the better educated, skilled and trained to search for jobs elsewhere, has been endemic to the state for decades. For business, unemployment is a two-edged sword. On the one hand it does significantly help to reduce the wage demands of workers, particularly so in the nonunionized sectors of the state's economy and increasingly so in unionized businesses, but on the other hand it also significantly restrains consumer demand. Of the two, the latter appears to have the upperhand.

The Unemployment Rate

The annual unemployment rates for West Virginia and the United States from 1960 to 1988 are illustrated in Figure 1. With the exception of 1975 and 1976, a short-lived boom period due largely to the strength of the coal industry following the rise of energy prices, the unemployment rate has consistently been greater than the national figure. In the recession of 1961, West Virginia's rate was 13.6%, nearly 7% more than the national rate. During that year, the state had the highest rate in the country. In 1983, at the height of the most recent recession, the state's rate was 18%, almost 9% more than the national rate. In that recession, West Virginia had thethird-highest rate in thecountry. Even towardtheend of the two longest expansionary periods in the history of the nation's economy, West Virginia retained its unenviable status. In 1969, when the lowest unemployment rates since World War 11 were recorded for both state and nation, West Virginia still had the highest rate in the nation at 5.6%. In 1988, the state's rate of 9.9% was fifth-highest. During the 1970s, the state's position was significantly better. In the recession of 1975, for instance, West Virginia's unemployment rate was a full percentage point below the national figure and had a ranking approximately halfway among the states. Unfortunately, the pattern during the most recent business cycle looks very much like that of the 1960s cycle.

The ratefor males in 1983 was more than 21%, and for those aged 16-19 it was more than 39%. Even by 1988, the latter had only dropped to about 27% compared to a national figure of 16%. It is hardly surprising that so many of the state's young people wish to leave West Virginia upon finishing their education.

Despite the fact that the state's unemployment rate is generally high relative to the national average, one might imagine that there are regions in West Virginia where unemployment is significantly less of a problem. Showing county unemployment rates for 1982 and 1987 respectively, indicates that some counties do have less of an unemployment problem, but their number is very small. In 1982 (Figure 4), the height of the last nationa economic recession, only two counties show unemployment rates below the national rate of 9.5%, namely Jefferson and Monongalia. The former is very much influenced by the fastgrowing Baltimore-Washington, D.C., metropolitan area, and the latter's economy is very much protected by state government employment opportunities at West Virginia University. In 1987 (Figure 5), during the national economy's recent expansion, only one additional county joined this group, that being Berkeley.

A much larger group of counties have unemployment rates above the national figure, but below the statewide figure. In 1982 (Figure 4), this group (9.6%-13.9%) was located primarily in the eastern panhandle, along the state line next to Virginia, in the north and northwest of the state, and around Charleston. By 1987 (Figure 5), this group (6.2%-10.8%) had expanded into the northern panhandle, but withdrawn from much of the Virginia state line.

Counties with unemployment rates above the statewide figure have been divided into two groups in Figures 4 and 5, those with rates above the state figures and below twice the national figure, and those with rates more than double the national rate. In both 1982 and 1987, counties with the highest rates form a core region in the middle of the state surrounded by counties with somewhat lower rates. Most noticeable about these maps is the fact that the region with

the highest rates expanded in all directions from 1982 to 1987. Although the unemployment rates of most counties declined over this period, relatively speaking, the plight of many counties worsened. Twelve counties had rates of more than double the national rate in 1982, but in 1987 as many as 25 did.

Reasons for Unemployment

People become unemployed for a number of reasons. First, a person can lose his or her job involuntarily, chiefly due to being laid off. Second, a person can leave a job voluntarily. Third, a person can reenterthe labor force unemployed after a period of absence. Fourth, a person can enter the labor force for the first time unemployed. Figures6and 7on page4 illustratethe importanceof these four reasons for unemployment in both West Virginia and the United States during the period 1976 to 1988. Losing one's job is clearly the major reason for being unemployed. In West Virginia, most years this accounts for more than half those who are unemployed. Unsurprisingly, the percentage who lose theirjobs increases during an economic recession and decreases in an expansion. As Figure 6 shows, the percentage for West Virginia was much the same as the national figure in the late 1970s. In the 1980s, however, the two have diverged considerably. In 1983, for example, more than 67% of those unemployed in West Virginia lost theirjobs, 9% more than in the nation as a whole. By 1988, 53% in the state were involuntarily unemployed, 7% more than the nation and second-highest among the states.

Typically, the percentage of those unemployed who leave their jobs voluntarily is much smaller. Understandably, it tends to rise when the economy is strong and fall when the economy slows down. Most people who voluntarily leave their jobs and become unemployed do so for the purpose of looking for another, presumably better, one. Figure 6 shows this trend to be evident in West Virginia. It also illustrates the fact that when the West Virginia economy was relatively strong in the 1970s, workers were more likely to leave their jobs voluntarily than in the nation as a whole, but that in much of the 1980s when the state's economy was relatively weak workers were much less likely to leave their jobs than the national average.

Reentry into the labor force is the second most common reason for unemployment. This may simply be due to the fact that the number of involuntary unemployed is so high in the state (reasons for unemployment must sum to 100). Alternatively, it might be due to fewer people being willing to reenterthe laborforce. This would be a rational choice on the part of many who are confronted by the dim prospect of finding a job in an economy where unemployment is persistently high. This interpretation is indirectly supported by the fact that the labor force participation rate in West Virginia is easily the lowest in the nation. For whatever reason, people in the state are less willing to be in the labor force.

With the exception of the growth period of the 1970s, there is very little difference between state and nation in terms of the percentage of the unemployed who are new entrants to the labor force.

Duration of Unemployment

The unemployment rate provides no direct indication of the length of time the unemployed have been out of work. Generally, about one third of the unemployed in West Virginia have been without a job for lessthan fiveweeks, and a little over one half have been unemployed for less than fifteen weeks. What distinguishes the statefrom the nation as a whole, however, is the extent of long-term unemployment. Figure 8showsthat, with theexception of the late 1970swhen the figures for the two were very close, the percentage of those without a job for 27 weeks or more is considerably greater in West Virginia. In 1988, for example, almost 27% of the unemployed in the state had been without a job for half a year or more, whereas only 12% were long-term unemployed in the nation. Long-term unemployment was worst in the state in 1983, when the figure reached 39%. Since the unemployment rate that year was 18%, we can infer that approximately 7% of the labor force was long-term unemployed. Unfortunately, Figure 8 indicates that the problem of long-term unemployment has not declined during the current expansion at anywhere near the rate at which it grew during the previous recession. Among the states, West Virginia has the highest long-term unemployment rate by a considerable margin.

Joblessness

The unemployment rate is only a partial indicator of the degree of joblessness in an economy. It measures the number of people actively belonging in the labor force who have no work at all in the formal economy. To get a truer picture of the extent to which people are not employed, we should modify the original rate in three ways. First,thosewho are not officially counted among the labor force but do work in the informal economy would have the effect of reducing the official unemployment rate if they were added to the labor force as employed workers. Consistent, reliable figures of these unofficially employed people are virtually impossible to obtain for rather obvious reasons. Common, everyday experience suggests that the number of such people in West Virginia is probably fairly high.

The unemployment rate can be modified in a second way to get a better idea of the degree of joblessness. Those who are in the full-time labor force and looking for a full-time job but are only able to find part-time work are officially counted as employed. These people arejobless, however, in the sense that they have not found full-timejobs. Figure 9 shows that as a percent of the labor force this group is quite large in West Virginia. Since 1976 it has always been higher than the national figure, but the discrepancy is greater in the 1980s. In fact, it peaked in 1986 when about 8% of the full-time labor force in West Virginia could only find part-time work, compared to a national figure of about 4%. If we add these data to the respective official unemployment rates, we derive the broader measure of joblessness illustrated in Figure 9. One can see that this only worsens the discrepancy between state and nation. For West Virginia, this joblessness rate peaked in 1983 at about 25% (compared to the national figure of 15%). Most recently, in 1988 the state rate was about 16% (compared to the national figure of 9%).

This measureof joblessness is unfortunately nottheend of the story. The official unemployment rate incorporates only those unemployed persons who are actively seeking work. It excludes those who are sufficiently discouraged about finding a job that they have withdrawn from the labor force (and, conceivably, those who never entered it in the first place). It is not easy to measure their number from official statistics, but we can make a very rough estimate. In their study of the national economy, Morehouse and Dembo (1984) estimated thatthe measureof joblessnesssuch asthe one shown in Figure 9 should be increased by 40% to reflect the addition of those discouraged enough to leave the labor force. If we apply the same factor to the West Virginia data, the distressing result would be a comprehensive measure of joblessness in the region of 5% for 1983 and 22% in 1988. Given the fact that long-term unemployment, which is likely to increase the incidence of discouraged jobseekers leaving the labor force, is more of a problem in the state than in the nation, it is even possible that the above figures may be underestimates. Further, albeit indirect, evidence to support the contention that joblessness may be this high in West Virginia comes from the labor force participation rate, which is easily the lowest in the country (51% for West Virginia in 1988 compared to 66% for the nation). Of course, one would have to discount an unknown number of unofficially employed from the above estimates to obtain the most accurate assessment of joblessness in West Virginia.

Causes of Unemployment: Supply and Demand

The level of unemployment in the West Virginia economy is a product of the supply of labor by the population of the state and of the demand for labor by private and public employers within the state.

We expect unemployment to be higher when the supply of labor (at a given wage) increases. There should, therefore, be a positive correlation between changes in the size of the

state's labor force and the state's unemployment rate over time. Actually, the correlation (using annual observations from 1963 to 1988) between these two factors is 0.46, contrary to expectations. What this negative correlation is almost certainly telling us is that when unemployment in the state is high it has the effect of reducing the size of the labor force. Given the state's high long-term unemployment rate and its low labor force participation rate, this is not an unreasonable conclusion. People may become sufficiently discouraged about finding work that those not in the labor force choose not to enter it and those in it choose to leave it, particularly if their unemployment benefits have expired. The negative correlation also tells us that when unemployment is lower, the size of the labor force expands. Again, this a rational response on the part of those looking for work.

The demand for labor by government employers in the state has consistently risen over time. To some extent this rise reflects population growth, but it also reflects the increased demand for government services by the state's population regardless of population change. The demand for labor by the private sector, however, is far more variable over time. It fluctuates through the business cycle for most industries, and in several industries there has also been a secular downward trend in the demand for labor. Coal mining, steel production and glass making are the three most important examples of industries in which there has been long-term decline in the demand for labor.

One important determinant of the demand for labor by private sector businesses in West Virginia is the demand for their products and services. Most product demand originates inside the state itself, but there is also a significant component of demand that originates outside the state, resulting in the export of goods and services to other states and nations. There should be a negative correlation between both in-state and out-of-state demand for West Virginia's products on the one hand and unemployment in West Virginia on the other hand. Using the annual change in the state's gross state product to represent in-state product demand for West Virginia goods and services, we find that the correlation between in-state product demand and unemployment during the period 1963 to 1988 is -0.59. In other words, when in-state product demand for West Virginia goods and services as a whole rises the unemployment rate declines, as expected. The correlation between out-of-state product demand and West Virginia unemployment should also be negative. When we correlate annual changes in gross national product (representing out-of-state product demand) with the states' unemployment rate for the same period, however, the result is a correlation of zero, suggesting no significant relationship since the early 1960s.

The same type of analysis can be undertaken with respect to the individual sectors of the economy. For example, the correlation between the demand for the state's coal and the unemployment rate in West Virginia is -0.34; the correlation between the demand for primary metals (largely steel) and unemployment is -0.38; the correlation between the demand for stone, clay and glass products and unemployment is -0.21; and between chemicals and unemployment the correlation is -0.06. Although the direction of the correlations is what one would expect, the correlations themselves are not large.

A combined analysis of the relation between labor supply, in-state product demand and out-of-state product demand on the one hand with the state's unemployment rate on the other hand, gives rise to a squared multiple correlation coefficient for the period 1963-1988 of 0.43. In other words, these three factors explain only 43% of the variation in unemployment in the state, which of course leaves 57% unexplained. Clearly, there are other factors which determine the demand for labor besides changes in product demand.

Virtually all private-sector industries in West Virginia have reduced their demand for labor over the last few decades in order to lessen their costs and retain profit levels. In coal mining, the trend began in earnest in the 1950s. Toward the end of the 1960s, the same trend began to appear in manufacturing. In both cases, the decline has continued to the present. Other research (Hanham, 1989) has examined the strategies typically used to reduce labordemand in these industries within the state. Two in particular have been most effective; namely, (1) disinvestment and (2) the substitution of capital for labor. Of the two, the former has been used mos effectively in manufacturing, and the latter most effectively in coal mining. However, both strategies have been used to some extent throughout private-sector industries in West Virginia. In both cases, the result has been a substantial increase in structural unemployment in the state. It seems fairly clear that this isthe majorform of joblessness inWest Virginia.

Unemployment and Public Policy

Although widely recognized as a major state problem, there is no government policy in West Virginia explicitly designed to prevent unemployment from occurring. Unemployment is assumed to result largely from the private actions of business and labor. Government does, however, become indirectly involved in determining the level of unemployment in a number of ways, and it is also directly involved in alleviating the effect of lost income for the unemployed through its unemployment insurance program '

Indirectly, government may influence the level of unemployment through a variety of policies which affect eitherthe supply of labor or the demand for labor. With regard to the supply of labor, the provision of educational and training programs, for example, substantially affects the supply of laborwith certain skill levels. Again with regard to the supply of labor, one controversial suggestion is that the state's policy of unemployment insurance inhibits the supply, the argument being that it provides a disincentive for the unemployed to accept work at lower wages. This is not a strong argument in light of two facts. First, there is no significant correlation between average unemployment benefit and labor force participation among states. Second, in West Virginia the average benefit is only 40% of the average wage which, in light of the fact that the latter is so low in West Virginia, does not seem likely to provide a disincentive to work, even assuming jobs are available.

With regard to the demand for labor, there is no explicit public policy aimed at reducing unemployment by providing jobs in the public sector. State and local employment nonetheless has the same effect. Next to the service sector, government employment is the only consistently growing sector in the West Virginia economy. In some counties it is the dominant sector, and without it there would be mass unemployment in many areas.

Indirectly, the state's economic development policies should also have an impact on the demand for labor by providing private-sector businesses with incentives to locate in the state and create new jobs. The success of such programs will be fairly limited, not least because there is a lot of competition from other states aiming to do exactly the same thing.

Finally, some have argued that the state's policy toward labor unions is too lenient, that unions have too much influence in government, and that union power inhibits the demand for labor by the private sector. Consequently, it has been suggested that the state should initiate right-to-work legislation, which would weaken the hold of unions in the labor force, hold down wage costs and increase the demand for labor. In light of the fact that real wage costs in West Virginia have been declining for some time, that real wage rates are already very competitive, and that there exists a relatively large pool of unemployed looking for work, it doesn't seem likely that such legislation would provide that much more of an incentive to private-sector businesses to locate in the state. Morelikely, it would simply impoverish the population of West Virginia even more.

State Unemployment Insurance

One state program which deals explicitly with unemployment is that of unemployment insurance. Established in the 1930s, this federally mandated program is designed to alleviate the effect of lost income for the unemployed who are covered bytheprogram. It is supposed to be a self-financing program based on taxes imposed on employers throughout the state. The state itself determines both the tax rate and eligibility requirements for the unemployed, although the federal government does set certain maximum and minimum standards. Payroll taxes are deposited in the Unemployment Trust Fund in the U.S. Treasury, out of which the state may withdraw funds to pay benefits to eligible unemployed. Typically, state funds are built upduring periodsof economic expansion and drawn upon during periods of recession. However, manystates in the 1970sand 1980s exhausted their accounts and have had to resort to borrowing from the federal governmeritto meet their obligations. West Virginia is one of those states. In this situation, the federal government imposesa penaltytaxon employerswithin thestateaswell as an interest charge until the loan is repaid. Needless to say, this funding problem is of great concern to West Virginia government, employers and employees. Employers regard the taxes involved as a cost burden and employees are understandably concerned about income security. Government is caught in the middle.

A key question regarding the program is what has caused itsfinancial problems. Iftheanswerto thatcan befound, then appropriate policy may bedeveloped to correct the problem. A common claim is that West Virginia's program is too generous; it costs too much and, furthermore, provides a disincentivefor people in the stateto work. Let's examine this in more detail. A comprehensive study by Vroman (1986) of state unemployment insurance programs in the 1970s and early 1980s concluded that although average benefits had increased as a whole and as a proportion of the weekly wage throughout the country, including West Virginia, this increase was more than offset by a fall in the ratio of insured to total unemployment. In other words, although benefits had become somewhat more generous, they were becoming available to far fewer people. Vroman's study also concluded that the funding problem in state unemployment insurance programs such as West Virginia's was very largely due to those states' persistently high unemployment, and not to the generosity of their programs.

Vroman's study ended in 1983. More recent data strongly suggests that his conclusions are still valid for West Virginia. In 1987, the last year forwhich comparable data are available, the average weekly benefit in this state's unemployment insurance program was $139, one dollar less than the national average. Furthermore, only 24% of those who were unemployed that year were covered by unemployment insurance. In the nation as a whole, 31% of the unemployed were covered. West Virginia's unemployment rate, however, was4.7% greaterthan the national rate (see Figure 1) and the overall joblessness rate in the state was 7.2% greaterthan the national rate (see Figure 9).

Except for a brief period in the mid-1970s, West Virginia has had the unenviable record of being one of the nation's leaders in unemployment for decades. The official unemployment rate in the state is dismaying enough, but even this understates the true extent and duration of joblessness in West Virginia by a considerable degree. What is to be done about this immense public problem? On the supply side, probably very little. Per capita income in the state is almost the lowest in the country and real wage rates are declining steadily. Policies to increase the labor supply would.simply worsen unemployment. As it is, endemic out-migration and the lowest labor force participation rate in the nation havethe effect of keeping the official level of unemployment lower than it would otherwise be, if not the general level of joblessness.

The cure lies on the demand side. As long as the state leaves this up to the private sector, however, its ability to deal with the problem of unemployment will be limited. There is little the state can do to influence the demand for the goods and services produced by the state's businesses, and there is very little the state can do to protect its businesses from the forces of an increasingly powerful international economy. West Virginia is but one of a very large numberof competitors who are desperately trying to entice private-sector employers to their part of the world, and with the developments that have recently been taking place in Eastern Europe it seems as though the competition will become even stronger. West Virginia has traditionally rejected a greater role for the public sector in managing its economy, but perhaps, ironically, it is time for the state to reconsideir this.