| 摘要 | Pity the poor labor economists. They build their careers studying one of the most complicated bazaars in the economic world–the marketplace for jobs. Not only that, but their labors have become even more high-stakes in recent decades, as an increasingly technological economy seems to be pushing more and more unskilled Americans into poverty. The popular magazine BusinessWeek poses the question this way: “What’s happening in the world’s richest, most powerful country when so many families seem to be struggling? And what can be done?”
BusinessWeek has a point. On the broadest level, the work of labor economists has a moral dimension. What would it say about the “American dream” if some families were unable to rise above the poverty level, despite their best efforts (efforts such as having a family member working full time)? And yet in tackling this dilemma, all too many economists are confounded by a cloudy understanding of the true nature of the low-wage problem and a confused approach to solving it.
Three questions confront labor economists. First, what is “working poverty” and how common is it? Second, why might a full-time worker earn less than he or she needs to support a family? Third, what can society do to help? Unfortunately, answers to these questions often prove elusive; solutions lie at the end of a road intersected with dead-ends and blind alleys.
A Question of Skill
The first step to answering the low-wage question is to determine the nature and scope of the problem. Here already unwary economists can run into trouble, because when we (meaning economists, policy makers, and the public) talk about the low-wage labor problem, we are really talking about two distinct issues. On one hand, we are confronting the marketplace for jobs, and on the other the well-being of individuals and households. This is an important distinction because the economic concepts that apply to the labor market itself, where wages and employment levels are set, are quite different from those that apply to measurements of welfare (which to an economist means “economic well-being”).
Measuring phenomena in the labor market is a tricky business, especially when it comes to comparing wages. An individual worker’s wage depends on a combination of factors that gauge the worker’s skills and the value an employer places on the ability of that worker’s skills to produce goods or services. The result is a huge spectrum of wages. Workers can have widely divergent personal characteristics. Differences in earnings power can depend on whether a worker is young and inexperienced or whether he has developed valuable skills through education and work experience. Meanwhile, different employers can value different types of education, or different previous work experiences, differently. It is customary to speak of average wages when we refer to the aggregate. Still, the range of wages in the labor market is wide, with a large share of workers earning wages near the middle of the distribution, but with some workers earning wages far higher or far lower than the average.
Given such complexity, developing a simple economic model to predict changes in wage levels might seem like a fool’s errand. And the labor economist’s task only becomes even harder when he turns to predicting the “welfare effects”–the changes in overall economic well-being–of those wage levels on families. One must first estimate how much income is necessary to provide modest levels of food, clothing, shelter, and other amenities to members of the household. Even for simple official measures of economic well-being, the economist must be mindful of how many adults and how many children are in the household, because the two categories require different consumption patterns. Only after one has done this is it possible to estimate the level of income that constitutes the “poverty line” that reporters, activists, and politicians fret about. As of 2001, that line was just below $18,000 for a family with two adults and two children.
As imprecise as all this is, however, at least one thing becomes clear immediately: A consequence of basing wages on workers’ productivity, and not on their household needs, is that some workers at the low end of the productivity, and thus the wage, distribution earn incomes too low to comfortably support a family. Certainly, many low-wage workers could not by themselves earn an income sufficient to support a large household above the poverty level. There is really no major puzzle to how one might find a household living in poverty despite being headed by a full-time worker, if that worker is unskilled. (One is especially likely to see this when the household has only one member employed full-time and at a very low wage, or when the household is large relative to the income a low-wage worker can expect to earn.)
Workers with lower skill levels (usually defined in terms of schooling and work experience) have always earned lower wages than more skilled workers. This difference has widened considerably since the 1970s. For much of the past 25 years, wages of high school graduates and workers with even less schooling have stagnated, while wages of workers with college credentials have increased substantially. The resulting increase in wage inequality has provided ample motivation for a careful look at the low-wage segment of the workforce.
Simple Supply and Demand
Much of the change is most likely rooted in the changing structure of the economy. The rise in the relative wages of skilled workers took place at a time when the relative supply of skilled workers in the labor market was expanding, although not as fast as it previously had. If one sketches out simple supply and demand curves, it becomes apparent from these observations that demand for skilled workers was expanding more rapidly than the supply of skilled labor–the price of that labor (the wage) increased even as the supply expanded.
At the same time, immigration has provided a steady flow of low-skilled workers into the economy even as the relative demand for low-skilled labor has been contracting. A large share of immigrant workers have completed far fewer years of schooling than their native peers. Census data examined by Harvard economist George Borjas show that in 2000, immigrant workers accounted for almost 40 percent of workers who are not high school graduates, an astonishing increase from 6.1 percent in 1970. Wages of workers without college credentials stagnated both because relative wages of workers with high school-level skills declined and because a growing share of less-skilled workers lacked even high school credentials.
Ironically, successful welfare reform has also played a part in the wage divergence, because it has moved large numbers of low-skilled workers into the workforce. The number of adult welfare recipients declined from a peak of over 4.5 million in 1993 and 1994 to slightly more than 1.4 million in 2002. However, here the news is not all bad: The economy has successfully absorbed most of these low-skilled workers, albeit at relative wages lower than they might have received before, which should stand as a tribute to the flexibility of the U.S. labor market. This flexibility stands in even starker relief compared to Western Europe.
Still, this increased relative supply at a time of shrinking relative demand has further contracted relative wages. It is a textbook market dynamic, but unfortunately the result is a growing number of low-skilled, low-wage households that cannot support themselves. In a society that values work and equality, this is easily seen as objectionable, and especially now, when higher-skilled workers have become so much better off compared to their lower-skilled compatriots.
The Scale of the Problem
That is the basic theory behind how working poverty arises. But how prevalent is it? Many, including Ronald Ferguson, argue that the problem is pervasive. They point out that, as Ferguson has it, “almost a quarter of the nation’s work force in 2001 earned less than $8.70 an hour,” that full-time work even at the top of this wage range would barely reach the poverty line, and that most low-wage workers are among the 42 percent of workers without college credentials. Yet although poverty for working families is a significant problem, it is not as prevalent as these numbers might suggest.
For starters, many low-wage workers do not have to support a family, and many of them are members of households where someone else is also working. According to the Census Bureau, a total of 9.2 percent of families had incomes below the poverty line in 2001. But for families with a member who worked year-round full-time, that number fell to 5 percent. Meanwhile, only 3.3 percent of families with a household head younger than 65 years old with a year-round full-time worker were in poverty. Together, these data imply that the number of families living in poverty despite having a full-time worker is actually quite small. It is much smaller than the proportion of workers whose wages are too low to enable them by themselves to earn above the poverty level for a family of four.
In addition, wage data alone don’t give a complete picture of household welfare. As Ferguson himself notes, there is a constellation of income-transfer programs designed to ease the burden of low wages. One example is the earned income tax credit, a formula-based grant designed to supplement income from earnings. A low-income working family with two dependent children can receive up to about $4,000 per year under this program. In addition, food stamps, housing assistance, Supplemental Security Income (SSI), Medicaid, and other programs to improve health and nutrition are often available to families with incomes near or below the poverty line. Many low-income families receive benefits from more than one of these programs, and the total benefits can be significant relative to wages: On average, families with children with incomes below the poverty line receive more than 30 percent of their annual income from just three programs–welfare payments, SSI, and food stamps–aid consisting of money and in-kind benefits. Such assistance makes an important contribution to filling the gap when low-wage workers do not earn enough to lift their families out of poverty.
This analysis leads to several conclusions. First, the labor market sets wages as a function of the worker’s skills and the value the employer derives from those skills, and in the aggregate the relative value of skilled labor has been rising faster than that of unskilled work. Second, there is the very real possibility of a disconnect between market-driven wages and workers’ consumption needs. Third, this disconnect is not as prevalent as popular imagination would have it. And fourth, if one views the “problem” of low-wage labor not as a problem of the wages themselves but as a problem of inadequate consumption possibilities in poor households, one realizes that there are opportunities to fix that problem outside the labor market’s wage-setting. That is to say, techniques such as income transfers can solve the real problem independent of what happens in the labor market.
Talking about Wages
Even if the real issue is household welfare and not wages per se, wages and the labor market that sets them still make for interesting study. The trick is to make sure one doesn’t lose sight of the important issue–welfare–when considering low-wage work. Unfortunately, this is a pitfall that often catches Ferguson and his unwary co-authors in Low-Wage America.
The book reports on extensive and detailed studies of low-wage workers. A major strength of these studies is the effort they make to understand changes that are taking place in the real world of low-wage work. Some essays, such as that on plastics manufacturing, describe changes in technology, competition, and marketing, and how those changes affect workers, their skills, and their jobs. The essays also highlight differences between industries, showing, for example, that new technology has contributed more directly to changes in jobs and skill requirements in bank check processing than in food service and hospitality. The authors identify many culprits in the widening wage gap. Some pressures stem from the general economic environment, factors such as globalization and advances in information technology. Other causes–deregulation in many industries, and a minimum wage held nominally steady in the face of a decline in real terms–are the products of government policies.
This book also shows how firms have responded in different ways to change, sometimes developing a profitable new market niche, sometimes making use of new technology, and sometimes adjusting to conditions that threaten their commercial survival. In fact, their observations of this adaptability lead the authors into one of their more serious wrong turns along the road to a low-wage solution. They interpret business nimbleness in the face of market conditions to imply that employers could readily respond to pressures in the labor market in ways that would allow them to pay higher wages. From that assumption it is not a far leap at all to the conclusion that one could use government fiat to raise wages directly without damage to businesses, and without making a significant investment in improvements to workers’ skills. They highlight ways in which some firms have adopted “high-performance work practices” that focus on reorganizing the workplace to eke more productivity out of low-skilled workers. Such practices might include teamwork, incentive pay, or some limited additional training. Firms adopting this approach are often described, in this book and elsewhere, as being on the “high road,” and generally pay even low-skilled workers higher wages. The commercial success of these firms is often interpreted as a sign of the broader viability of this approach.
On its face this is an appealing idea, but on further inspection it turns out to be rather superficial. The authors are certainly correct that firms are adaptable. But it is difficult to interpret that adaptability, because it is hard to judge whether differences in patterns of adjustment reflect the wide discretion of the firm, or whether those variations indicate real differences in circumstances that are not easily identified. In other words, it is not at all clear that all firms could adopt these strategies. Because different industries and firms have different production processes and skill levels among workers, one cannot assume that the success of a particular set of practices implies that other firms would enjoy the same results if they followed the same regimen.
On an even more basic level, however, the authors’ method is poorly suited to the question they take on. Looking at groups of low-wage workers and tracing changes in the job characteristics and the wages that those workers experience provides valuable information on how firms and workers adjust to aggregate supply and demand imbalances. But this research methodology can provide only limited insight into the size and sources of the aggregate imbalance, and is not especially useful for identifying the most promising approaches to restoring a new balance more favorable for less-skilled workers. The authors constantly place themselves in danger of missing the forest for the trees.
This focus on individual firms creates one particularly problematic blind spot in the study. The authors investigate many jobs, such as manning telephones in call centers, housekeeping in hotels and hospitals, and many manufacturing jobs, and conclude that these are “dead-end” jobs. They make this claim based on the fact that there is hardly any hierarchy of skills in these lines of work that could provide promotion opportunities within firms. Making beds will not teach a hotel maid the computer skills she needs to move into a job behind the reception desk. This flies in the face of the traditional definition of a “good job,” a definition still near and dear to many labor economists–the existence of a labor market within the firm that provides a well-defined job ladder.
Yet this notion of a “good job” seems increasingly antiquated in a fluid labor market like that in the United States, where between 2.5 and 3 million workers leave their jobs and a slightly larger number obtain new jobs in a typical month. (This according to the Bureau of Labor Statistics; the difference between the number obtaining and the number leaving jobs produces net job growth, on average, of about 150,000 jobs per month.) So even if there are scarce opportunities to move to a higher-skilled, higher-paid job within a given firm, there should be abundant opportunities to advance by moving between firms. Given that about a quarter of all employed people move on and off the payrolls of individual firms during the year, a need to move between firms to climb the career ladder would not seem to be a difficult barrier to surmount. Yet if climbing the ladder increasingly takes the form of jumping from one firm to another in search of a better job, especially among young workers, opportunities for low-wage workers to advance are largely missed by studies such as Ferguson’s that capture only the opportunities within an individual firm. (As a practical matter, of course, studies often show that many workers aspire to move to jobs that are not very different from the jobs they are leaving. But this has historically been the case for blue-collar manufacturing jobs as well.)
Fixing Whose Problem?
The authors’ methodology does more than just skew perceptions of the low-wage problem–it also shapes misguided policy prescriptions. To see how this is so, consider that in crafting policies to counter the deleterious effects of the wage gap one needs to consider the interests of several different groups: currently employed low-wage workers, unemployed workers with low earnings capabilities, and the public at large. However, a method that focuses only on studying the first group almost inevitably leads the authors to offer recommendations that would “help” these workers, conceivably at the expense of the other two groups.
For example, Ferguson and his colleagues suggest an increase in the federal minimum wage and a greater density of union organization. Both these measures could certainly benefit workers who are currently employed, at least as long as they kept their jobs. And adding a prohibition on temporary hiring to some union contracts could also benefit these newly unionized workers, perhaps. But it is less clear that such changes would do much good to unemployed people with low earnings capabilities, or would serve the public interest as a whole. And an artificially high minimum wage, set to accommodate household welfare needs without regard to worker productivity, would make it more difficult for the least skilled to get jobs and to benefit from acquiring work experience.
When it comes to temporary employment, it is not even clear that such a prohibition would aid currently employed low-wage workers. Many labor market specialists seem to view temporary employment as inimical to the interests of workers. Such theorists note, among other things, that many workers in temporary jobs really want to obtain permanent employment. But if temporary employment is prohibited, such jobs would not necessarily be replaced by permanent jobs. Meanwhile, temporary employment can be a productive alternative to unemployment for many low-wage workers who want to earn some money while looking for a permanent job. And, in fact, that temporary job can actually serve as a trial period for permanent employment, as the employer and the worker size each other up to determine whether a long-term relationship would work. After all, a good job match requires more than information on job duties and pay. Workers want to know what a particular job is really like, while employers want to know what a particular worker is really like. Temporary employment offers both parties an opportunity to find out.
Thus temporary jobs can make the labor market work better by lowering the unemployment rate without producing inflationary pressures–a point that the authors concede. And it can provide a path into permanent employment for unemployed workers with low earnings potential. Therefore, even if the effects on the particular group that the authors study are ambiguous, the public as a whole (as measured by indicators of overall economic health) is likely to benefit.
Where the Road Leads
If these authors do not offer a useful solution, it seems fair to ask what form such a solution might take. Policy makers need both to return to first principles in formulating an answer to the low-wage problem, and to remember that they must serve the interests of several different groups.
Low-wage labor and its cousin low household consumption pose challenges to at least three different groups. The first is low-wage workers themselves, for obvious reasons. The second are those with low earnings capabilities, with or without jobs, who face the dual challenge of entering the workforce and then making ends meet. The third is society as a whole, which is not ready to tolerate a large underclass with incomes too low to afford a decent standard of living.
Protecting the interests of all three requires a juggling act. As I have argued above, measures such as prohibitions on temporary labor might help one group, but are likely to harm at least one of the others. Meanwhile, regulations such as a higher federal minimum wage might introduce an entirely new set of problems for one group–business–that is currently in a sort of equilibrium, and might harm another group–the unemployed–even though it would benefit currently employed low-wage workers.
It appears, then, that artificially regulating the price or supply of labor in the labor market would be fraught with perils. But programs, both private and government, that focus on worker training might offer a way to intervene in the cycle of low-wage work without overly distorting the marketplace for jobs.
Going back to basic principles, a central driver of the increasing wage gap appears to be the relative supply and demand of skilled and unskilled labor. If skills are in short supply, it seems quite natural to ask whether more training by employers might make sense, and whether greater opportunities could be created for workers to move up in skill levels and pay. More skill development, in the form of better or additional schooling, or more specific training, could help to increase the wages of the less skilled and reduce the divergence in pay between skilled and unskilled labor.
Augmenting the skill supply in this way would help alleviate imbalances in the labor market in two distinct ways. First, enhancing the skills of a worker with low skills so that the worker can qualify for a job requiring a higher level of skill directly increases the earning power of the worker. Training transforms a formerly unskilled worker into a skilled worker, thus allowing the worker to earn higher pay. Second, transforming a single unskilled worker into a skilled worker has a small aggregate effect on the labor market. That is, such a change increases the relative supply of skills in the labor market, and therefore increases the relative scarcity of the remaining unskilled labor, and its wages.
Training is a constructive response to employers’ needs for workers with better skills, and to the desire of workers to increase their earning power. That might be why it has become such a popular idea. Additional investment has already been taking place in response to the increase in the wage premium for skill. A larger proportion of young people completing high school have been getting additional schooling than before; the fraction of young people enrolled in some kind of post-secondary schooling after completing high school increased from about 50 percent in the late 1970s to over 60 percent since the beginning of the 1990s, for example.
Unfortunately, so far the record is mixed when it comes to policies designed to improve training. It is essential that training actually increase workers’ earning power if it is to be beneficial. But much of our experience with federally sponsored job training programs up to now suggests that they do not pass that test. Changes in the amount and quality of training provided by employers are harder to measure. But it appears that there are some subtle positive changes underway. Even though technology has reduced demand for workers with few skills, hiring workers with fewer skills is less costly, so employers have found ways to make productive use of less-skilled labor. These methods allow the workers themselves to acquire marketable work experience in the process.
It bears emphasizing that well-crafted training programs offer a way to tackle the underlying cause of wage divergence without meddling with the dynamics of the fluid U.S. labor market, which is one of America’s greatest economic assets. This allows policy makers to avoid false solutions that would actually harm one or more of the groups that have a stake in a positive outcome.
Finally, it should be recognized that even if policies are pursued that encourage training and maintain flexibility in the labor market to adjust to new technology, strong competition, and more globalization, some families will still have difficulty earning enough money to support a decent quality of life. We have a number of different kinds of income maintenance programs that are intended to alleviate material hardship. Reliance on such programs that help to directly address the problem of inadequate income is generally preferable to policies that restrict wage and job flexibility and reduce opportunities for people to benefit from working and contributing to their own economic well-being.
Marvin H. Kosters is a resident scholar at AEI. |
