NBER Reporter: Spring 2000

Labor Markets and Firm Benefit Policies in Japan and the United States

The NBER and The Japan Center for Economic Research

The NBER and the Japan Center for Economic Research (JCER) held a joint conference on "Labor Markets and Firm Benefit Policies in Japan and the United States" in Hawaii on January 20-23. The following papers were discussed:

Andrew A. Samwick, NBER and Dartmouth College, and David A. Wise, NBER and Harvard University, "Option Value Estimation with Health and Retirement Study Data"
Discussant: Yukiko Abe, Asia University

Makoto Kawamura and Seiritsu Ogura, Hosei University; and Tamotsu Kadoda, JCER, "Reengineering the Japanese Social Security System: Treating Everyone Equally in Health Insurance and Saving Money"
Discussant: David M. Cutler, NBER and Harvard University

Yukiko Abe, "Fringe Benefit Provisions for Female Part-Timers in Japan"
Discussant: David A. Wise

David M. Cutler, "Supplementing Public Insurance Coverage with Private Coverage: Implications for Medical Care Systems"

Discussant: Tadashi Yamada, University of Tsukuba

Tadashi Yamada, "The Demand for Health Checkups and Uncertainty"
Discussant: Mark B. McClellan, NBER and Stanford University

Seiritsu Ogura and Takehiko Hagino, JCER, "Why Do the Japanese Spend So Much on Drugs?"
Discussant: Mark B. McClellan

Matthew J. Eichner, NBER and Columbia University; Mark B. McClellan; and David A. Wise, "Individual Expenditures and Medical Saving Accounts: Can They Work?"
Discussant: Seiritsu Ogura

Toshiaki Tachibanaki, Kyoto University, "The Role of Firms in Welfare Provisions"
Discussant: Andrew A. Samwick

Takao Kato, Colgate University, "The Recent Transformation of Participatory Employment Practices in Japan"
Discussant: Hiroyuki Chuma, Hitotsubashi University

Yuji Genda, Gakushuin University, "Who Really Lost Jobs in Japan? Youth Employment in the Aging Japanese Society"
Discussant: Takao Kato

Hiroyuki Chuma, "Determinants of the Shadow Value of Concurrent Information-Sharing in Japanese Machine Toolmakers"
Discussant: Richard B. Freeman, NBER and Harvard University

Richard B. Freeman, "Changing the Guard: The Rise of the United States to Peak Capitalist Economy"
Discussant: Toshiaki Tachibanaki

Yoshifumi Nakata, Doshisha University, "Total Labor Cost and Employment Level Adjustment under the Shukko System"
Discussant: Takao Kato

Fumio Ohtake, Osaka University, "Unions, Cost of Job Loss, and Nonattendance"
Discussant: Richard B. Freeman

Do financial incentives designed to delay retirement actually work? Using Health and Retirement Study data, Samwick and Wise find that both pension wealth and pension incentives have statistically significant effects on the probability of retirement. These effects are more robust when retirement is defined only as a job separation rather than as a complete transition out of the labor force. Samwick and Wise also investigate possible interactions between the effects of wealth, health, and health insurance on retirement.

Kawamura, Kadoda, and Ogura suggest that Japan's elderly health care system and its National Health Insurance (NHI) programs together have created a crisis in Japanese public health insurance. The two systems transfer funds from corporate employees and general tax revenues to pay health benefits, but beneficiaries themselves pay relatively little. As Japan's population ages, the number of health care beneficiaries continues to grow, putting enormous pressures on Japan's national budget and increasingly alienating corporate employees, who now pay almost twice the costs of their own health care. The authors then ask what would happen if the elderly health care system were eliminated, and if NHI collected enough revenue via consumption taxes to pay its own costs. Their results suggest that the number of welfare cases would roughly double, to about two million households, but that a drastic reduction in cross-subsidization would be achieved.

Abe documents patterns of participation in social insurance programs by female part-time workers in Japan and asks how the structure of the social insurance system affects that participation. Because participation in social insurance programs -- such as public pensions, health insurance, and Employment Insurance -- is required only for workers who put in enough hours or earn enough, many part-time workers are not required to participate. Currently, the financial incentives for participation are weaker for married women and older workers than for others. Abe finds that from 1990 to 1995, for workers who satisfied the hours and earnings conditions for social insurance insurance programs, the married women and older workers were no less likely to participate than others. On the other hand, among workers who did not meet these same conditions, married women and older workers were significantly less likely to participate. Female part-time workers did increase their participation in social insurance programs from 1990 to 1995, not because of changes in the labor supply but because of a general increase in participation, Abe concludes.

Cutler considers the rationale and economic effects of allowing supplemental medical insurance to exist. Countries may allow three types of insurance supplements: insurance for uncovered services, which most countries allow; insurance to pay for cost sharing under the standard government insurance plan, which some countries allow but others do not; and insurance that allows people to jump to the front of queues, again allowed in some countries but prohibited in others. In the United States, insurance to pay for cost sharing under Medicare is common. Cutler shows that such insurance raises Medicare spending by up to one-third.

Good health enhances market earnings by increasing the number of days that employees work and by increasing their nonmarket productivity in terms of time spent on household production. And, health checkups are one way to secure good health. Yamada, using sample data from Japan's 1995 Comprehensive Survey of Living Conditions of the People on Health and Welfare, finds that a number of socioeconomic and demographic factors determine the demand for health checkups among workers aged 30 to 60. These factors include age, gender, earnings, type of health insurance coverage, firm size, occupation, and objective evaluation of health condition. Furthermore, Yamada shows that health checkups do reduce the probability of becoming ill and of becoming a hospital in-patient. His evidence strongly supports the theory that medical checkups are highly cost-effective as preventive medical care.

In Japan, prescription drug expenditures account for a huge proportion of total health care costs. Among five major countries, Japan and France each rank at the top in absolute terms for per-patient drug costs, spending more than three times as much as Great Britain, twice as much as the United States, and one and a half times as much as Germany. Ogura and Hagino establish that this is the result of a systematic distortion in drug price controls. The government's attempts to control drug prices have failed, as drug companies have been allowed to substitute old drugs with more expensive new drugs. Ogura and Hagino suggest that by removing distortions that encourage health care providers to select more expensive drugs, drug expenditures in Japan could be reduced by 20 percent or more.

Low levels of household savings in the United States essentially make catastrophic health insurance plans infeasible since most households would find it difficult to pay even modestly large medical expenditures. Medical savings accounts (MSAs) offer one way to accumulate funds from which medical expenses could be paid under a catastrophic health insurance plan, making individuals responsible for determining whether the benefit of the treatment is worth the price of care. Eichner, McClellan, and Wise consider whether the pattern of individual medical expenditures over time presents an important limit on the feasibility of MSAs. Based on the illustrative plan they simulate, the authors find that most employees would approach retirement with a substantial proportion of MSA contributions remaining in the account. Thus they conclude that the persistence of medical expenditures does not present an overriding obstacle to the adoption of MSA plans.

Tachibanaki presents five theoretical justifications that explain why firms contribute to welfare provisions: 1) agency theory, 2) tax advantage, 3) employees' preference, 4) cost savings, and 5) improved industrial relations. He then shows that the importance of these five justifications has declined recently in Japan. Tachibanaki proposes that two arrangements can substitute for contributions to welfare: an increase in wages, provided that the payroll tax incidence in Japan is known; and introducing a progressive expenditure tax.

Kato shows that Japanese firms have continued to use participatory employment practices, even during the overall economic slowdown in the 1990s and the recent financial crisis at the end of the decade. However, he points to a few early signs of trouble for these practices. First, while the number of full-time union officials has been falling substantially as a result of continued labor force downsizing, the amount of time and effort needed to put participatory employment practices in place has not changed. This often results in an uncompensated increase in workload for union officials, which could lead to their being less prepared and less committed to the interest of the rank and file. Second, top management may find its participatory system detrimental to timely and efficient management and hence try to streamline it. Overloaded union officials could offer less resistance to this type of management initiative. Third, the current system tends to produce a gap in the quantity and quality of information acquired from management between top union officials and their general membership. Such a gap eventually may result in the breakdown of the participatory system.

Young workers likely face more difficulty than middle-aged or senior workers in finding full-time jobs in an aging Japanese society. Yet job opportunities for middle-aged and senior employees are being maintained and protected, while no clear answers or political strategies have been found for solving the decline in youth employment in Japan. Genda, using cohort analysis, a job creation and destruction framework, and labor flow functions, shows that decreasing the chance for youths to acquire skills and restricting dismissals severely by law results in the displacement of youths in Japan's labor market.

Japan has led the world in machine tool production since 1982. In examining how the Japanese machine tool industry has achieved this record, Chuma maintains that human relations practices have been especially important. He finds that a concurrent comprehensive and egalitarian information-sharing system exists between production workers and design and development engineers. This type of system was established in Japan in the 1980s, when machines had become so complex that close collaboration among various professionals and designers was essential. His questionnaire data confirm the statistical analysis which predicts that this information-sharing system is likely to occur more frequently in machine toolmakers that retain both highly skilled assembly workers and machinists, produce lathes, and have more than 100 full-time workers.

In the 1990s, many analysts viewed the United States as the peak capitalist economy, whose institutions other countries would be wise to emulate. By contrast, in the 1970s and 1980s, many analysts viewed Japan as the peak capitalist economy. Freeman develops criteria for judging which, if any, extant economy might be legitimately labeled the peak economy. He examines the features of the U.S. variant of capitalism and compares them to those of the Japanese variant. Freeman then evaluates the employment, wage, and productivity growth record of the United States and Japan to determine whether the shift in the peak label from Japan to the United States is justified. He concludes that the case for the United States as peak capitalist economy rests almost entirely on the country's 1990s job market performance. Only as long as full employment lasts will that label be legitimate.

People have long believed that Japanese firms, particularly large ones, have upheld their employment commitments to workers despite any unfavorable economic conditions. But now, this practice has been criticized as the major cause of the prolonged Heisei recession. Nakata compares the employment adjustment among different industries and among firms in those industries and uses per-capita total labor cost in place of average monthly salary as the relative factor price variable reflecting the growing share of nonwage total labor costs. He also uses both single firm data and consolidated firm data to investigate the impact of Shukko, a common Japanese employment practice of dispatching workers to their related firms. Nakata finds that large scale employment adjustment is not uncommon among major Japanese firms. Employment adjustment is both firm and industry specific: in other words, the patterns of employment adjustment vary by industry and by firm within an industry. When redundancy is considered, total labor cost is relevant as the firm labor cost variable. Finally, when the Shukko effect on employment stability is tested, the results are consistent with the author's Shukko hypothesis.

Ohtake hypothesizes that the presence or absence of a union influences the effectiveness of internal and external threats in reducing nonattendance. He argues that, because unions reduce the potency of threats of job loss (by making it more difficult for employers to dismiss workers), nonattendance responds less to such threats in union firms than in nonunion firms. An analysis of data from the 1985 and 1993 General Surveys on Working Hours and Conditions in Japan supports this hypothesis.


National Bureau of Economic Research, 1050 Massachusetts Ave., Cambridge, MA 02138; 617-868-3900; email:

Contact Us