This month's featured article
Minimum Wage and its Effect on Employment
David NeumarkProfessor of Economics, University of California, Irvine
November 16, 2007: RIETI International Seminar - Summary
Minimum Wage, Employment, and Income Distribution:
Presentation by Professor David Neumark
The U.S. Congress expanded coverage of the minimum wage and raised its level during the 1960s and 1970s. In the meantime, a debate emerged about the appropriateness of this policy. In 1977, the Minimum Wage Study Commission was established. In 1981, the Commission produced a report concluding that a 10% increase in the minimum wage reduces employment of young workers, particularly teenagers, by 1%-3%. This was called the "consensus view" of minimum wage effects that most people seemed to agree on. However, this research had two limitations. First, with federal data alone, it is difficult to construct a counterfactual analysis, which is needed for policy study. It is necessary to make a comparison between what happened when the minimum wage was increased and what would have happened had the minimum wage not been increased. In addition, in the U.S. at that time, nearly all research on minimum wages used aggregate time-series data and the federal government set the level of minimum wage and its coverage. In addition, the federal minimum wage was not indexed to inflation and changes were relatively infrequent. Thus, it was difficult to distinguish the effects of minimum wage changes on employment from those of other aggregate influences. Second, because of limitations of the data available to researchers, the analysis merely covered certain demographic groups such as teens and young adults, and the more general population of workers earning at or near the minimum wage were not the primary focus of research.
Since the publication of the Commission's report, the federal minimum wage changed little over the 1980s in the U.S. As the real value of the minimum wage fell as a result of inflation, several states started to set their minimum wages independently. (The federal minimum wage is binding for almost all workers everywhere, but states can implement a higher minimum if they want.) In 1984, three states had a state minimum wage that was higher than the federal level. In 1989, 13 states had higher minimum wages. In 2007, 30 states had higher minimum wages and recently some states have started to index the minimum wage to inflation. The state-level variation in minimum wages has opened the way for employment comparisons between states where the minimum wage has been raised and states where it has remained unchanged, taking minimum wage studies into a new phase. Since then, there has been an enormous number of papers on minimum wages, which vary in data, quantitative approach, target industries, and countries. However, there is continuing debate about the interpretation of different studies done to date. Some claim that there is no evidence that the minimum wage has any negative effect on employment, others argue that the evidence generally points to a negative effect, and, finally, some maintain that there is no consensus on the direction or size of the effect on employment.
Professor Neumark examined more than 100 papers estimating the effects of minimum wages on employment, in collaboration with Dr. William Wascher of the Federal Reserve Board, attempting to draw general conclusions about the effects of the minimum wage on employment. Instead of performing a formal meta-analysis, they opted for a traditional narrative review, which introduces some subjectivity, but enables them to present their arguments and assessments of the evidence, and to leave readers better informed to draw their opinions based on their study. They summarized the lengthy review of studies into a set of tables including the minimum wage change variation, group studied, data used, results, and main criticisms. This survey revealed that about two-thirds of the papers suggest that the minimum wage has a negative impact on employment while fewer than 10 of the 100 studies demonstrate that it has a positive effect. In addition, they focused on 33 studies they saw as providing more credible evidence, and 28 pointing to negative employment effects. Thus, they have concluded that an increase in the minimum wage adversely affects employment.
Based on their review, Neumark and Wascher suggest that recent studies targeting the U.S. can be divided into four types. The first type uses state-level panel data and defines teenagers or young adults as the target of analysis. The second type is case studies of specific industries. The third is time-series studies of teenagers using aggregate data. And the fourth focuses on less-skilled or low-wage workers, those who are most directly affected by the minimum wage (regardless of age). The survey finds that analyses using state-level panel data over longer periods tend to affirm a negative effect of the minimum wage on employment and that many of the papers concluding that the minimum wage has a positive or zero effect on employment are based on short-term panel data or case study analyses focusing on specific industries and states. These latter papers are considered to have the following problems. First, the effect of the minimum wage on employment cannot be observed over a short period since firms do not adjust their inputs immediately. An analysis using short-term panel data cannot capture any potential long-term adjustments to minimum wage increases and therefore is likely to underestimate the adverse effect on employment. Studies focusing on specific industries, like that of Card and Krueger (1994), which featured an analysis of employment at fast food restaurants, cannot take into consideration the substitution in consumption among different narrow industries that may shift employment among them, and thus fail to clarify the overall effect of the minimum wage on low-skilled workers.
Based on these results, when examining whether or not to raise the minimum wage it is advisable to quantify its positive and negative effects, namely the improvement in income and the reduction in employment. It is important to identify trade-offs between jobs and equity. In this regard, it is vital to note that the objective of the minimum wage system is to reduce poverty or help low income families. As a consequence, it is essential to distinguish between the effect of the minimum wage on low-paid workers and its effect on low-income households. An examination of U.S. data on the relationship between family income and wages earned by workers in each family shows that many low-wage workers are in middle-income or high-income families. This implies that the minimum wage policy may ultimately not benefit low-income families. To determine whether minimum wages help low income families, then, it is important to investigate directly how minimum wages affect the distribution of family incomes. An analysis by Neumark and Wascher of changes in the number of low-income households resulting from minimum wage increases over the 1980s and 1990s has shown that the number of such households went up, contrary to the objective of the minimum wage policy. This means that an increased minimum wage may merely benefit the youth of middle-income or high-income families and it may reduce employment for poor households. There is a good deal of research that tends to find no distributional effects of minimum wages, either positive or negative, and less evidence consistent with Neumark and Wascher's finding of adverse distributional effects. On the other hand, there is virtually no existing research that finds positive distributional effects of minimum wages.
The assertion is therefore that the minimum wage policy is not effective in reducing poverty because of the improper orientation of its benefits, as well as the costs that it imposes. Moreover, other research, including that by Neumark and Wascher, finds that the Earned Income Tax Credit (EITC), a scheme of subsidizing earnings of workers in low-income families, is effective in increasing incomes of low-income families. Part of the reason for this is surely that the EITC targets low-income families, rather than simply targeting low-wage workers. As a result, the EITC is highly regarded by both parties in the U.S. Congress.
While pointing out that negative employment effects of the minimum wage have been observed in many countries, Professor Neumark stressed that the distributional effects could vary from country to country as they are heavily dependent on their own institutions such as welfare systems and labor market systems, as well as the distribution of income and the distribution of low-wage workers throughout the family income distribution in other countries. As a result, caution is required when applying the U.S. experience regarding distributional effects to other countries.
The minimum wage issue in Japan
Professor Neumark's report was followed by a presentation from Dr. Daiji Kawaguchi, associate professor at Hitotsubashi University, on the system and current circumstances surrounding minimum wages in Japan. In his presentation, Dr. Kawaguchi mentioned that Japan's minimum wages are effectively set by the Central Minimum Wages Council of the Ministry of Health, Labour and Welfare, that there is a massive disparity between the trend in the minimum wage level and that of the average wage, and that researchers have considered the possibility that minimum wages actually constitute no major constraint on the setting of wage levels in Japan. However, comparisons in wage distribution by gender suggest that female wage distribution is heavily distorted by the presence of the minimum wage. Dr. Kawaguchi reported that a considerable number of female workers are paid at nearly the minimum wage level. Moreover, the gap between the average wage and the minimum wage varies significantly among prefectures. For instance, the gap is smaller in Aomori prefecture than it is in Tokyo. Given this, he added that the minimum wage is very likely to act as a constraint in the provincial female labor market.
In the following roundtable discussion, Dr. Yukiko Abe, associate professor at Hokkaido University produced materials to demonstrate the possibility that the minimum wage could compress wage distribution in rural areas, since it serves to support the wage level in such regions.
After these reports, the discussion was opened. A participant challenged Professor Neumark's view that thinking about the labor market in the framework of the competitive model is probably adequate by stating that it is subject to massive frictions in some job types and in some regions and that imperfect competition could occur.
In response to a question about how inter-regional migration of labor is perceived in existing studies, Professor Neumark replied that the disemployment effect of the minimum wage might be underestimated if inter-regional labor migration and commuting issues were taken into account.
Responding to a comment that the minimum wage should be raised for reasons of fairness or more equal distribution of income, Professor Neumark pointed out that arguments in favor of seeking poverty reduction as a goal must be strictly distinguished from the argument about whether a particular policy achieves that goal. He noted that economists can identify the distributional effects of minimum wages and of other policies, but that it is up to policy-makers as to what distribution of incomes they are trying to achieve.
Another argument offered in support of the minimum wage is that it raises productivity of workers. In response, Professor Neumark noted that most economists would support the goal of raising the skills of low-skilled workers to increase their productivity and earnings, although accomplishing this is difficult. However, the kind of productivity increase most likely to result from a higher minimum wage does not achieve this goal. Although it is true in the standard economic model, as the labor input is reduced the marginal product of labor rises, in this case the productivity increase simply reflects lower employment so that, for example, each worker works with more capital. But it does not reflect higher skills of workers. He agreed with other participants' views that to achieve a more equal distribution of income, if a minimum wage policy is implemented it should be coupled with other measures to increase skills, such as training programs. In this way the increased skills might reduce the negative employment effects of a higher minimum wage.
With respect to a citation about subsidies to boost the productivity of small- and medium-sized enterprises severely affected by higher minimum wages, Professor Neumark explained that there has been a debate in the U.S. on tax relief for small businesses as a part of the minimum wage increase package.
Concluding the discussion, Dr. Kawaguchi mentioned that minimum wage studies were still underdeveloped in Japan and pointed out that there was urgent need to accumulate not only analyses using prefectural panel data but also research using microdata, like the many studies referred to by Professor Neumark, to better understand the impact of the minimum wage on income distribution and employment.
Fellow titles and links in the text are as of the date of publication.
For questions or comments regarding RIETI Report, please contact the editors.
RIETI Report is published monthly.