RIETI Report October 2009

Desirable antipoverty policy

With less than a month in office, the Democratic Party of Japan's manifesto behind their philosophy of "Putting People's Lives First" is already the subject of public discussion. In this month's RIETI Report, KAWAGUCHI Daiji, RIETI Faculty Fellow and Associate Professor in Economics, Hitotsubashi University, looks at the DPJ's proposal to establish a regulatory "national minimum wage" and discusses whether such action is an effective antipoverty policy.

Drawing on empirical findings from a 2009 study based on extensive government-level data, Professor Kawaguchi takes a look at the current situation and policy of regulatory minimum wages across the nation, discussing how the labor market might react to the DPJs proposal and whether or not such policy would have the desired effect. He also suggests how the implementation of earned income tax credits as an antipoverty policy might work for Japan as it has done in the U.S. and UK.

This month's featured article

Desirable antipoverty policy

KAWAGUCHI DaijiFaculty Fellow, RIETI
Associate Professor, Faculty of Economics, Hitotsubashi University

Is the Minimum Wage Effective Antipoverty Policy?


In its policy manifesto, the Democratic Party of Japan (DPJ) has pledged to raise minimum wage levels. Specifically, the DPJ has defined the minimum wage level as " living expense sufficient to support a worker and his/her family" and promised to: establish a regulatory "national minimum wage" (presumably at 800 yen per hour) applicable to all workers; aim to raise the national average minimum wage to 1000 yen per hour; and implement fiscal and financial measures to facilitate the compliance of small and medium-size enterprises (SMEs) with the regulatory minimum wage requirement.

Current state and policy debate

Currently, regulatory minimum wages for Tokyo, Aomori, and Okinawa stand at 766 yen, 630 yen, and 627 yen per hour respectively, as effective from October 8, 2008. The regulatory minimum wages are set on a region-by-region basis in a way to reflect inter-regional disparities in actual wage levels. The adequacy of the regulatory minimum wage levels has become subject to much public debate with many people finding them unreasonably low.

The Minimum Wage Act, as amended on July 1, 2008, has effectively acknowledged the need to rectify the current state of affairs, where individuals working a given number of hours at the minimum wage rate are earning less than the income of those receiving livelihood protection benefits. Specifically, the law calls for revising the regulatory minimum wage rates to levels justifiable compared to the "livelihood protection standards," threshold income levels below which households are eligible for livelihood benefits. The legal framework for raising the regulatory minimum wage levels has already been thus established in Japan.

However, raising the minimum wage as a means to diminish poverty has traditionally been subject to much controversy. Setting regulatory minimum wage rates is nothing but a price control and many economists including George Stigler have shown a strong allergic reaction to the idea of influencing the allocation of resources by manipulating prices. Stigler (1946) argued that setting a floor for wage rates - i.e., the price of labor - is undesirable for the following two major reasons: 1) many of those working at the lowest wage are members of middle- and middle-upper income families (typically, housewives and university students working part-time), and 2) raising a minimum wage would increase the number of SME bankruptcies, resulting in a decrease in employment.

On the other hand, it has been argued that in the case of Japan where regulatory minimum wage rates are far too low relative to market wage rates, changes in the regulatory minimum would have no impact on actual wage levels. But this view is biased to a considerable degree, over-representing the situation in Tokyo. In fact, the regulatory minimum wage rates in Okinawa and Aomori are rather high relative to the actual wage levels in these prefectures. Thus, both Okinawa and Aomori prefectures have many people working at or around the regulatory minimum wage levels.

Using microdata collected for the government's Basic Survey on Wage Structure (Wage Census) in years from 1994 to 2003, Kambayashi, Kawaguchi and Yamada (2009) examined the relationship between the regulatory minimum wage and the distribution of wages. They found that the regulatory minimum wage in Tokyo was set at a level outside the range of wage distribution both in 1994 and 2003, meaning that even a part-timer at a fast food restaurant is paid an hourly wage rate above the regulatory minimum of 766 yen. Thus, the above argument about the regulatory minimum wage rate having no impact on actual wage levels holds true for Tokyo. However, in the case of Aomori Prefecture, the number of male workers (including both full-time employees and part-timers) paid at the regulatory minimum wage in 2003 is not negligible. Likewise, an extremely large number of female workers in the prefecture were working at or around the regulatory minimum wage in 1994 and it has been found that in 2003 the regulatory minimum wage affected wage decisions in a significant proportion of cases.

This study of wage distribution found that the regulatory minimum wage is functioning as a policy. Then, the next question is whether it is desirable as an antipoverty policy. Drawing upon empirical findings in a joint study by Kawaguchi and Mori (2009), I would like to present my analysis to answer this question.

Who are prone to become minimum wage earners?

In this study, we examined the percentage of those working at or below the regulatory minimum wage to total workers (hereinafter, "minimum-wage workers ratio") using data collected in the government's Employment Status Survey, more specifically, micro data on about 1 million individuals aged 15 or older taken from a sample of about 440,000 households for the years 1982, 1987, 1992, 1997, and 2002. (Non-full-time employees working an irregular work schedule, estimated to account for about 12% of total workers, are outside the survey sampling frame, while self-employed and persons doing piecework at home are excluded from the sample.

Due to constraints in data collection, figures derived from the data are subject to a margin of error. However, as a general tendency, it has been found that the minimum-wage workers ratio is relatively high for: those whose highest education attained is either junior high or high school; female workers; those in the young and old age categories; workers in relatively rural areas; those engaged in the wholesale and retail trade and those in accommodation and hospitality businesses. Furthermore, in almost all categories, the minimum-wage workers ratio increased between 1982 and 2002, which is perceived to be attributable to the rise in the regulatory minimum wage levels over the 20-year period, rather than indicating an increase in poverty in Japan.

Who are more likely to be minimum-wage workers?

Although policy challenges faced by Japan are very similar to those faced by European countries, factors underlying these challenges are somewhat different.

In Japan, only about 10% to 14% of minimum wage workers are the heads of poor households with an annual income of 2 million yen or less, and nearly half of minimum wage workers are members of middle- to high-income households with an annual income of 5 million yen or above. From this finding, we can see that those who are to benefit from an increase in the regulatory minimum wages are not necessarily members of poor households. It is more likely the case that these beneficiaries turn out to be non-head members of middle- or higher-income families, such as children and spouses of the heads of households. This suggests the possibility that raising the regulatory minimum wage rates may not be appropriate as a policy targeted at the poor.

From the Employment Status Survey data, we also learned that minimum wage workers tend to have low educational attainment and that middle- or older-aged women account for a significant portion of the minimum wage workers.

Regulatory minimum wages and employment

Now, even if raising the regulatory minimum wages is not well-targeted at the poor, we still cannot entirely deny the possibility of it being an effective anti-poverty policy if higher minimum wage requirements do not result in increased joblessness. Let's suppose a situation where a regulatory minimum wage is raised to a level equal to or above the market wage rate. Would this lead to losses in employment?

Theoretically, such a minimum wage hike would not necessarily result in losses in employment and there is even the possibility that it could lead to an increase or maintenance of employment. The key determinant here is whether or not workers have bargaining power in the labor market.

For instance, in a perfectly competitive labor market, where multiple employers compete for labor, employers would not be able to employ workers if their offered wage rates were lower than the market rate. In such a situation, raising the regulatory minimum wage to or above the market wage would result in employment losses.

On the other hand, in a non-competitive labor market, where employers have relatively strong bargaining power and workers' options are limited, employers would be able to drive wage rates down by reducing employment. For instance, suppose a labor market with a monopsonistic labor market. When in need of more workers, a situation to which an employer in a competitive labor market would respond by offering higher wages to secure the necessary workforce, the monopsonistic employer might be able to hold down the overall labor cost by temporarily reducing employment and thus driving down the market wage. However, when a regulatory minimum wage is introduced in this monopsonistic labor market, the market wage cannot be driven down below the regulatory wage floor and, hence, the employer's incentive to reduce employment as a means of holding down the labor cost does not exist. In this situation, it is theoretically possible that a hike in the regulatory minimum wage may lead to an increase in employment in what may seem to be a strange phenomenon.

Neumark and Wascher (2008) in the United States warned against drawing a conclusion on the effects of a regulatory minimum wage on the labor market as a whole from a case study of a specific industry or state. Based on an analysis of longitudinal panel data with an appropriate model, they concluded that regulatory minimum wages tend to have a negative impact on the aggregate employment. On the other hand, Card and Krueger (1994) argue that a rise in regulatory minimum wages does not necessarily lead to employment losses. Like this, with various studies made and still continuing, no consensus has been reached in the United States.

Likewise, no definite conclusion has been obtained in Japan. However, in their analysis using prefectural panel data for the period 1997-2002, Kambayashi et al. (2009) found that an increase in regulatory minimum wages leads to a decrease in employment of middle-aged women. Meanwhile, using the data collected in the Japanese Panel Survey of Consumers (JPSC), Kawaguchi & Yamada (2008) showed that a hike in regulatory minimum wages leads to a statistically meaningful decrease in the labor participation rate of affected individuals in the following year.

We have conducted an expanded survey, collecting data from a larger sample (more than one million individuals) over a longer period of time compared to those in the preceding studies. The regression analysis of the sample data has found that a rise in regulatory minimum wages tends to:
1) have a negative impact on the employment of men in the young age group;
2) lead to a decrease in married women in the middle age group; and
3) have no impact on the employment of men and women in the old age group.
These findings indicate that raising regulatory minimum wages is not necessarily a desirable antipoverty policy.

What is a desirable antipoverty policy?

First and foremost, the government should accurately define and identify poor households and transfer income to those households in a pinpointed manner. The question is how and in what form such income transfer should be done. The existing livelihood protection is problematic because it does not provide proper incentives for work. Under the current program, households with income below the poverty threshold are provided with livelihood protection benefits of an amount sufficient to make up the difference between their earned income and the threshold income. Thus, so long as their earned income falls below the poverty threshold, these households are left with the same amount of money even if their earned income decreases.

Meanwhile, carefully designed earned income tax credits could serve as effective wage subsidies. And it would become possible to stimulate work incentives by raising wage rates. In fact, cases of successful implementation of earned income tax credits as an antipoverty policy have been reported from the United States and the United Kingdom. However, this policy could end up benefiting employers, rather than workers, by helping them to save on the cost of labor because an increase in labor supply puts downward pressure on wages. Due attention must be paid to this risk.



In the United Kingdom, the regulatory minimum wage levels have been raised but the unemployment rate remains relatively low. How would you interpret this? Also, from an economics viewpoint, what significance do you see in implementing relevant fiscal and financial measures in combination with raising regulatory minimum wages as has been proposed by the DPJ in its manifesto?


An international comparison by Neumark and Wascher has shown that the youth unemployment rate tends to be high in countries with a high ratio of minimum wage to average wage. However, what is difficult about international comparison is that countries differ in various aspects, for instance, in terms of the strictness of regulations and how wages are set in this particular case. Indeed, Neumark and Wascher noted that the relationship between minimum wage levels and employment would change depending on cross-country institutional differences. But, generally speaking, they found that the youth employment rate tends to be low in countries where the Kaiz index, or the ratio of minimum wage to average wage, is high.

It has been reported that the DPJ intends to implement, in combination with its plan to raise regulatory minimum wages, fiscal and financial measures worth about 200 billion yen to help SMEs adapt to the stricter minimum wage requirements. Provided that we consider labor market distortions resulting from raising the regulatory minimum wages as given, such a policy would be justifiable. But it is desirable to keep the increase in the regulatory minimum wages to a minimum and avoid implementing fiscal and financial measures in a way to distort the financial market. In order to finance such measures for SMEs, the government needs to take away funds from other programs, causing distortions in financial resource allocation. I am afraid that it is quite difficult to achieve optimization by simultaneously redesigning the labor and capital markets.


I believe that the United Kingdom raised the regulatory minimum wage levels with an aim to provide work incentives for members of households on welfare. Can we expect the same effect if the DPJ raises the regulatory minimum wages as it promised in its manifesto?


Raising the regulatory minimum wages as a means of stimulating work incentives is a viable possibility. But earned income tax credits can also increase the wage rate after tax credits. Therefore, from the viewpoint of avoiding the unwanted side effect of employment losses, I think providing earned income tax credits is a more desirable way of giving work incentives. As to the impact of a labor supply increase on the wage rate, if the labor demand curve is relatively flat, that is, in a situation where a decrease in the wage rate leads to a relatively large increase in employment, the wage rate would remain almost unaffected by the increased labor supply. There is no telling exactly how things will turn out as it depends on the individual case. However, regarding the slope of the labor demand curve, employers in industries with a large number of minimum wage workers, such as convenience stores and family restaurants, tend to determine the employment level in a very cost-sensitive manner. Therefore, it is to some extent predictable that even a slight decrease in the wage rate will significantly increase employment. The negative effect of increased labor supply, i.e., a decrease in the wage rate, may not be as big as feared.

This report was edited by RIETI editorial staff based on RIETI BBL seminar No.606 held on September 9, 2009.

Related article

Is Minimum Wage an Effective Anti-Poverty Policy in Japan?
Faculty Fellow, RIETI
Associate Professor, Faculty of Economics, Hitotsubashi University

Minimum Wage and Employment
Minimum Wage and Employment
RIETI International Seminar: November 16, 2007

Minimum Wage and its Effect on Employment
RIETI Report No. 089: December 14, 2007

Event Information

For a complete list of past and upcoming event information.



BBL Seminars

Fellow titles and links in the text are as of the date of publication.

For questions or comments regarding RIETI Report, please contact the editor.

*If the "Send by mailer" button does not work, please copy the address into your email "send to" field and connect the prefix and the suffix of the address with an "@", sending it normally.

RIETI Report is published bi-weekly.