Reforming Labor Market Institutions to Promote Elderly Employment (Summary)

As societies around the world age, economies are struggling to adapt. One way to keep economies healthy is by increasing labor, which also means keeping older workers in the labor force for longer. But what challenges do these workers face? What kind of legislation is in place to help older workers as they start winding down their careers? And do the policies meant to prevent age discrimination actually end up disenfranchising other groups, such as women or youth? International experience and evidence and possible implications for Japan and much more were discussed during the RIETI/IZA World of Labor Policy Symposium, "Reforming Labor Market Institutions to Promote Elderly Employment."

• FUJITA Masahisa (President and CRO, RIETI / Professor, Konan University / Adjunct Professor, Institute of Economic Research, Kyoto University)

According to an estimate by the Organisation for Economic Co-operation and Development (OECD), the aged dependency ratio (ratio of the number of people aged 65 and over to the number of people aged 15 to 64) will exceed 60% in such nations as Japan, Spain, Italy, and Greece. This means that every six aged persons will be supported by 10 working-age persons, moving ever closer to a " kataguruma (shoulder-riding) society" where a single working-age individual must each shoulder the full burden of supporting an elderly person.

Today, at this symposium, we will hear expert insights on the effects that policies to promote the employment of the elderly in the United States, Europe, and Japan have had on job opportunities for them and other age groups in these countries. After these presentations, we will discuss what Japan can learn from other nations' experiences when considering its future policies to promote elderly employment. I hope today's symposium will be a good opportunity to disseminate the research results on the promotion of elderly employment conducted at RIETI and IZA and to provide materials for policy formation.

• Alessio J. G. BROWN (Director of Strategy and Research Management, IZA)

The Institute for the Study of Labor (IZA) is an independent research institute focusing on the analysis of global labor markets and is also the largest research network in economics. On the basis of our research and the resulting evidence, IZA experts also provide evidence-based scientific policy advice, which is the central aim of IZA World of Labor.

Scientific guidance is essential for economic policy being successful and promoting welfare. Evidence-based policy making adopts solid evidence rather than ideology into policy making to promote good public policies. Evidence-based scientific policy advice consists of providing viable options and highlighting their implications such that policy makers can make evidence-based, but ultimately their own, decisions.

To provide appropriate advice, not only must the advice be independent, but also it must be based on prior independent, internationally competitive, and peer-reviewed research. It requires adherence to strict scientific ethics rules and standards as well as disclosure of conflicts of interest. Data access is also essential, though policy programs need to account for data collection before implementation. The need for independent scientific evaluation is still too often ignored at the time of implementation of policies. Thereby, evidence-based policy making is forced to prove itself repeatedly.

The relationship between science and policy making faces an additional central challenge: academic science generally is not targeted immediately at policy advice. Consequently, while empirical evidence on many issues is available, often it is not relevant or usable immediately to policy makers. This is where the IZA World of Labor ( http://wol.iza.org ) fills an important gap: as an independent service provider for evidence-based policy making, it "translates" the existing evidence and makes it transparent. IZA World of Labor condenses the available state-of-the-art knowledge on key labor market issues in a practical format and derives functional policy recommendations and insights of value to society from the existing evidence.

By making our knowledge freely accessible to a global audience including journalists, economics students, and the interested public, we aim to contribute to promoting evidence-based policy making.

• David NEUMARK (Professor of Economics, University of California, Irvine / Research Fellow, IZA)

All developed countries struggle with aging societies, especially with getting seniors to work more and support pensions. But to what extent is age discrimination a problem?

Older unemployed workers take a lot longer to find jobs, and there are prevalent negative stereotypes about older workers. Other evidence also suggests that age discrimination plays a role in U.S. labor markets. The United States has passed laws to stop it, which likely boosted employment rates of older individuals by about 4%, probably by preventing the workers from being fired. In effect, those have been anti-termination laws, not hiring laws.

The law is less effective with regard to hiring because it is largely enforced in the courts. First, penalties for firms are also decided based upon the economic damages that are done, and these damages can be low in hiring cases. In contrast, firing a 55-year-old is quite costly due to the need to replace health insurance and lost pension accruals. Second, the strongest discrimination suits are class action suits, but it is hard to identify an affected class in hiring cases. This is potentially problematic because substantial increases in the employment of older workers probably have to come about through hiring into "post-career" jobs.

I will now discuss two studies on age discrimination. The first measures whether stronger laws are more effective in assisting older workers increase labor supply. The United States has changed Social Security, bumping up the age at which full benefits are received and providing lower benefits for earlier retirement. But does age discrimination diminish the effect of that change in policy? This can be measured in the United States due to variation across states in age discrimination laws. Stronger laws do increase the responses of older workers to changes in Social Security.

The second study is about the Great Recession in the United States, when older workers' unemployment durations shot up dramatically. Age discrimination claims filed during that time spiked and remained high. In this case, in the states with stronger laws, unemployment rates and durations went up relatively more. It seems that the laws tend to fail during a severe recession, making it more costly to hire an older worker perhaps because with uncertain labor demand, firms have to worry about having to terminate an older worker and being sued for age discrimination.

Overall, the evidence suggests that stronger age discrimination laws could enhance the ability of older workers to retain and find jobs, and make other policy reforms intended to lengthen work lives more effective.

• Juan F. JIMENO (Head of the Research Division, Bank of Spain / Research Fellow, IZA)

Europe's demographic situation is not as severe as that of Japan, but it already has the highest unemployment and highest youth unemployment. Nearly no one older than age 65 works in Europe. Aging is the most intense in Japan, but its elderly employment rate is the highest. Regarding labor market policies for promoting employment of older workers, Europe has more to learn from Japan than vice versa.

What evidence is there to propose changes in labor market policies? One study reflects on labor market flows in different countries during the financial crisis. The most interesting part was the countries where unemployment increased significantly. In Spain, increases in job destruction for older workers were much lower compared to young workers. Less-educated young women were the worst off of all.

The data can be analyzed to find whether differences are across countries or population groups, and if they are related to labor market institutions. Very simple regressions are performed to understand how the effects of labor market institutions on job creation and destruction differ across population groups.

If employment labor legislation is very stringent, there is less job destruction. With less destruction, young workers are less protected than older workers. Temporary contracts make it easier to dismiss young workers than older workers. The differences are not large, but still show how different labor market institutions across European countries may affect job creation and destruction.

These simple regressions provide some idea of the effects of legislation on job destruction of different population groups. Similar regressions were performed for unemployment benefits, as well as wages and union density, and the level at which collective bargaining takes place. The same results were found from unemployment to employment and job creation, depending on the institutions in place.

The relationship between wages and productivity might cause labor market policies to make those flows particularly intense. Most labor markets have seniority-based pay systems, but employment protection protects those with more seniority, meaning that older workers typically have more bargaining power than younger workers. Thus, wages tend to increase with seniority. Since productivity declines with age, that means firms have a serious imbalance between wages and productivity. That should give firms incentive to hire younger workers, but employment protection is too stringent. Young workers are hired largely on temporary contracts, and are much cheaper to dismiss. The current effects of employment protection on youth unemployment are particularly high.

Employment opportunities and benefits must be neutral. A firm should internalize the societal costs of dismissal, such as with employee training and unemployment benefits. Another important factor in dismissing workers could be the stigma that a firm will draw. Dismissing workers was not a problem before because enough social insurance was available, but mandatory retirement is not the way.

The fact that incentives to invest in skills are different makes older workers less attractive to firms. But for transferring specific human capital to younger cohorts, all workers are useful; there is no better method of transfer.

• KONDO Ayako (Associate Professor of Economics, Yokohama National University)

When looking at trends in the labor participation rate and the employment-to-population rate of men in their late 50s to 60s in Japan, one recent feature is the rise in the employment-to-population rate in the early 60s age group. This is the result of a policy change, and, although the labor participation rate has also increased, the rise in the employment rate has been even larger.

As for women, both the labor participation rate and the employment rate are lower than that of men in all age groups, with a significant proportion of women working either part-time or in their family business. However, the employment rate is on an upward trend in all women age groups. Another notable point is that the gap between their employment-to-population rate and the labor participation rate is smaller than that of men, which means that the unemployment rate is lower.

When comparing the employment rates of people in their 60s or older, Japan's figure is higher than that of other nations, with the rate for men in their early 60s in particular standing out considerably higher. This phenomenon is related to the Elderly Employment Stabilization Law (EESL). The employment rate of men aged 65 or older is also high, with one out of four in this group still working. Women's employment rate is relatively high too. Despite the comparatively low employment rate among younger women in Japan, the rate is almost the same as that of the United States in older age groups, and higher than that in European nations except Sweden.

Why is the employment rate so high? There are two main reasons. First, the pensions for self-employed workers are not very generous. Many say that they continue working out of necessity because it is impossible to live only on the national pension. Second, a relatively large number of people also choose employment for reasons other than money, such as health benefits, sense of fulfillment, and social participation.

In the 2000s, the EESL was revised following a reform of the pension system (which raised the eligibility age for the employees' pension). Both changes had the biggest impact on men in their early 60s who were in full-time employment until the mandatory retirement age of 60.

In addition to the raising of the pension age, a change was implemented in the in-work pension scheme in April 2005. Previously, if a pensionable person was employed in a job that would entitle them to join the employee pension scheme and received remuneration, their pension was cut by 20%. The change ended this fixed-rate reduction, and it is thought that this also encouraged more elderly people to take up employment.

Another significant change was the revision of the EESL in 2006. Because the pension age was raised in 2001, the EESL was revised to fill the gap between the pension age and the mandatory retirement age. This revision took effect in April 2006, five years after the eligibility age started to rise. This revision can be viewed as a formalization of the already common practice of spontaneous continued post-retirement employment into an explicit compulsory requirement. These changes had a particularly significant impact on male employees who were working in large-sized companies.

There is no clear evidence yet as to whether these measures to protect elderly employment have a negative impact on other age groups. Although based on insufficient data, research suggests that elderly employment protection does not affect the employment of new graduates. However, according to my own research, the possibility that such elderly re-employees are replacing middle-aged or older female part-time workers cannot be denied.

When considering effective policies, it is essential to analyze the potential impact on the supply of female labor. With the ongoing population decrease, it is important to increase the female labor supply in order to secure a sufficient labor force, and this aspect deserves continued attention. It is also necessary to accumulate appropriate data to verify the impact on corporate productivity.

• TSURU Kotaro (Chair of Employment Working Group, The Regulatory Reform Council, Cabinet Office / Program Director and Faculty Fellow, RIETI / Professor, Graduate School of Business and Commerce, Keio University)

"Limited-type regular workers" are barely discussed here in Japan. Understanding the Japanese employment system and the relationship between wages and tenure is crucial. Teamwork is a defining feature, as well as horizontal information sharing.

However, the Japanese system has a blind side: unlimited scope of duties. Temporary workers are obliged to accept the same conditions as permanent employees, such as location transfers and overtime. Limited-type regular workers have clearly-defined job duties but for lower wages. More importantly, there are no legal restraints in employing such workers.

Our human resources departments are very different. An unlimited scope of duties has profound implications for work-life balance in Japan. Permanent workers are difficult to dismiss, thus temporary workers are preferred; their contracts are tenuous, but the duties are the same. Their overtime hours are extremely demanding, and women have faced considerable obstacles in maintaining their professional careers.

Some Japanese still view employee dismissal as strict and do what they can to have workers voluntarily quit. But wages rise with tenure, not productivity; so employees have an incentive to stay, even under negative pressure and harassment. Mandatory retirement has also changed, with all employees wishing to work after age 60 being free to continue. Japanese companies face many difficulties.

• KAWAGUCHI Daiji (Faculty Fellow, RIETI/ Professor of Economics, Hitotsubashi University / Research Fellow, IZA)

The elderly employment rate is quite high in Japan, which may give a false impression that there are no problems with elderly employment. In fact, problems relating to elderly workers in Japan almost always center upon the sustainability of pension accounts. The Japanese government has raised the pension eligibility age to 65 for the national pension, and, by 2025, other pension schemes will be in line. Some policy makers argue that the gap between the retirement age and pension eligibility age would create problems and have proposed policy interventions to prevent them.

Some argue that government intervention is not necessary because the rising pension eligibility age encourages elderly workers to seek employment to secure their incomes. On the other hand, while Japan has legislation preventing age discrimination, there are many loopholes that make the law virtually ineffective; for example, targeting new graduates for recruitment is still legal. How should we assess the current scheme?

Some argue that legal intervention to increase elderly employment might reduce opportunities for other workers such as youth and women; therefore, the elderly employment stabilization law may stifle opportunities for disenfranchised groups.

I would like to invite the panelists' opinions on these issues.

NEUMARK: Presumably, Japanese firms employ mandatory retirement because it works for them. The issue isn't that people stop working at age 60 or 65, but rather that it is what happens after.

TSURU: In Japan, it is essential to consider why wages decline so significantly after retirement. It is a very striking difference, in our seniority-based wage system.

JIMENO: In Europe, seniority-based wages are the norm, but the situation is completely different. Europe does not share the same sort of assumed wage-profile as in Japan, and there is no mandatory retirement. In fact, Europe has the opposite problem: workers retire too early and older worker employment rates are too low.

KONDO: Right now, with workers accepting huge wage cuts at age 60, they can receive enormous subsidies from unemployment insurance. It is not ideal, but it is a good compromise. Whether it is necessary to make the wage profile flatter and to raise mandatory retirement age, or implement age discrimination legislation, or otherwise keep things as they are, is a matter of debate.

TSURU: Introducing a mid-career, limited-type regular worker system could be a solution. They could learn many different jobs and master various skills. Accordingly, their wages should depend on the job, not ability. Japan has to change anyway, so it should shift to a system with a better work-life balance. The Japanese employment system worked well before, but it is now in decline. As the system adjusts, non-regular workers hold a very unstable position in terms of job security, which has become a social problem.

NEUMARK: Our discussion has focused on the phenomenon that workers get a sort of extension of mandatory retirement, and make a lot less as a result. However, one possibility is that there is quite a bit of age discrimination in hiring. Essentially, the firm makes a very low take-it-or-leave-it offer, which the workers take because they have no other option.

We also should not lose sight of the fact that increasing employment of the elderly is not the only option for increasing the number of workers. The situation of women in Japan is quite abysmal relative to other developed countries. The other option to look at is immigration.

JIMENO: Europe learned that there must be many ways to deal with problems. Countries using different margins of adjustment did much better in the financial crisis. Germany adjusted hours, productivity, and so on, and it did not experience a spike in unemployment. Spain only adjusted temporary employment, and unemployment rose significantly. For elderly employment, some workers with very high skills and value to their firms could be rehired with the same or higher wage. By providing unemployment benefits for low-productivity workers, firms would internalize the social costs of mandatory retirement if workers cannot find a new job. That sort of additional instrument could benefit the situation.

NEUMARK: What could be done more to learn about the potential impact of the Elderly Employment Stabilization Law (EESL) on other groups of workers, such as women? If the employment rate for middle-aged women rose from 40% to 70%, that is a significant amount of labor.

KONDO: If policy aims to increase the labor force, then female employment has a much greater margin than that of the elderly because elderly employment rates already are relatively high. Also, although the EESL currently does not have much negative effect for workers other than middle-aged female part-timers, if the government instead mandated firms to increase the mandatory retirement age, then the substitution effect might work in a completely different way, and side effects will spring up somewhere else.

JIMENO: It is important to increase both female and elderly participation rates. If benefits are not reduced, then they must be kept relative to productivity, which implies a doubling of the employment rate; that is impossible. To minimize the cutting of benefits that would make the welfare state sustainable, employment of women and older workers must be increased.

NEUMARK: In the face of the high dependency ratio, increasing both elderly workers' employment rates and those of female workers are important.

JIMENO: In many European countries, before the crisis, employment of older workers was increasing and unemployment of youth was decreasing. During the crisis, the situation was reversed: both employment rates of older worker and youth unemployment increased.

TSURU: Japanese companies determine beforehand the number of new graduates that they take on every year. They prefer hiring new graduates, so the effect of rehiring elderly workers is already small. Companies must work on hiring different workers. The roles that elderly workers will play is different.

Q1. What would be the impact of introducing robots or artificial intelligence on employing the youth or elderly?

NEUMARK:
People have worried for ages that mechanization and technology would destroy jobs, but that hasn't come to fruition. It can affect the wage distribution, but probably does not destroy jobs on net.

JIMENO:
This time could be different due to a deficit of demand generated by high savings and low investment. Labor income distribution could become polarized.

KONDO:
It would become an issue of distribution. Innovation would improve productivity, so people can work less with more output. If distribution is not an issue, it benefits everyone. Wages would have to improve as well.

Q2. This is regarding RIETI President Masahisa Fujita's opening remark, where he said that significant institutional reforms are needed to attain an active labor force participation of the elderly. Are there any proposals for reforms?

TSURU:
It is simply an issue of human resources management. Everyone still thinks that the current employment model and wage profile are good, especially at large companies. They aren't sustainable, but there is too much momentum behind them. Twenty years ago, companies used temporary workers, but that has spiraled out of control. Japanese people themselves must change and cope.

NEUMARK:
What could the government actually change with legislation? If rigidity prevents firms from reforming, the rigidity must be identified and we need to ask what the government can do about it. A single firm can't make the change--it's a coordination problem.

Q3. Can you explain more about limited-type regular workers?

TSURU:
The same labor and wages should be realized first to promote mobility in the labor market, otherwise disparity could further expand. Japan has yet to see the growth of a market for non-regular employment. How someone is treated within an organization should be examined as a matter of course. If there is a reasonable explanation of the differences, it is not necessarily bad, but disparities among workers must be minimized.