This Selected Issues paper examines implications for long-term bond yields in case of Japan. The analysis finds that so far, upward pressure on interest rates from high public debt has been offset by domestic factors, including a stable investor base with a preference for safe assets. As these effects could decline with population aging, yields could rise unless reforms are implemented to stimulate growth and reduce the public debt-to-GDP ratio. In such a scenario, long-term Japanese government bond rates would remain relatively low and stable. The paper also analyzes to what extent rising health care spending poses a fiscal risk to Japan’s economy.

Abstract

This Selected Issues paper examines implications for long-term bond yields in case of Japan. The analysis finds that so far, upward pressure on interest rates from high public debt has been offset by domestic factors, including a stable investor base with a preference for safe assets. As these effects could decline with population aging, yields could rise unless reforms are implemented to stimulate growth and reduce the public debt-to-GDP ratio. In such a scenario, long-term Japanese government bond rates would remain relatively low and stable. The paper also analyzes to what extent rising health care spending poses a fiscal risk to Japan’s economy.

The Path to Higher Growth: Does Revamping Japan’s Dual Labor Market Matter?1

This paper discusses causes and economic consequences of Japan’s dual labor market, as well as options for reform. The paper argues that Japan’s excessive labor duality reduces Total Factor Productivity (TFP), due to a negative impact on non-regular workers effort and on firms’ incentives to train them. Based on an empirical analysis of the determinants of labor market duality for a panel of OECD countries, the paper argues that reforms aimed at reducing the difference in the degree of employment protection between regular and non-regular workers could significantly reduce duality in Japan, thus stimulating TFP and growth.

A. Introduction and Main Findings

1. Labor market duality forcefully emerged in Japan in the last two decades, as declining growth challenged the traditional lifetime employment model. After serving the country well in high-growth decades, Japan’s traditional lifetime employment system was challenged by declining growth in the wake of the asset bubble collapse in the early 1990s, which led to Japan’s so called “lost decade.” The prolonged recession, together with labor law reforms—which made it easier to hire non-regular workers—gave firms incentives to explore alternative human resource practices. As a result, a growing share of non-regular workers were removed from the traditional lifetime employment model, which nonetheless continues to be widespread even in present days among regular workers.

2. This paper focuses on trends, causes and economic consequences of Japan’s dual labor market, and suggests options for reform. The paper argues that excessive duality can reduce total factor productivity (TFP) due to its negative impact on non-regular workers effort and on firms’ incentives to train them. Excessive duality can also reduce social cohesion and lead to the perception that economic growth is not inclusive, thus potentially eroding support for needed structural reforms. On the basis of cross-country panel estimation, the paper argues that one option to reduce duality in Japan could be the introduction of a Single Open Ended Contract (SOEC) for all new hires, in which employment protection increases with tenure. This reform could be accompanied by a more general shift toward the so-called “flexicurity” model, in which the focus is on providing lifelong employment opportunities and supporting workers during periods of temporary unemployment, rather than on protecting specific jobs. For the proposed package of reforms to work, a change in “soft institutions”—aimed at improving the work-life balance—would also help.

B. Japan’s Labor Market Duality: Trends and Economic Costs

3. The share of non-regular workers in Japan, which was below 20 percent before the burst of the bubble in the early 1990s, has now reached 35 percent. Compared to regular workers, non-regular workers have a much lower level of job security, are paid lower wages and receive significantly less social insurance coverage. About 70 percent of non-regular workers are women, accounting for 55 percent of female workers. Low-value added service sectors are highly reliant on non-regular employment. Japan’s share of temporary workers—the only proxy of labor market duality available for international comparisons—is above the OECD average (text chart).

uA04fig01

Share of Temporary Workers (2011)

(In percent)

Citation: IMF Staff Country Reports 2013, 254; 10.5089/9781475563795.002.A004

Source: OECD.
uA04fig02

Annual Change

Thousand persons

Citation: IMF Staff Country Reports 2013, 254; 10.5089/9781475563795.002.A004

Sources: Ministry of Internal Affairs and Communications and IMF staff calculations (Feb. 1985-2001, 2002-2012 Jan.-Mar. average).

4. Recent studies and international experience suggest that the macroeconomic costs of excessive labor market duality are likely to be substantial in Japan. Japan’s large labor market duality likely reduces productivity and growth, by lowering firms’ incentives to train non-regular workers, and through reduced workers’ incentive to exert effort due to low job satisfaction (Fukao and others, 2007; Kawaguchi and others, 2006). Cross-country empirical studies also suggest that excessive labor market duality can reduce TFP (Dolado, Ortigueira, and Stucchi, 2011; Damiani, Pompei, and Ricci, 2011). Excessive duality in Japan could also potentially lead to the perception that economic growth is not inclusive, thus eroding support for structural reforms needed to increase potential growth. In light of this, next section will look at empirical determinants of labor market duality, with the aim of identifying policy options to reform Japan’s labor market.

C. Determinants of Labor Market Duality

5. Various factors have been highlighted in the literature as contributing to the surge of non-regular employment in Japan. Asano, Ito, and Kawaguchi (2011) estimate that demographic changes and shifts in the industrial structure explain about one quarter of the increase in non-regular employment since the mid-1980s. A more detailed list of factors include:

  • The increase in female labor-force participation and the consequent demand for flexible work schedules (Gaston and Kishi, 2007; Asano, Ito, and Kawaguchi, 2011).

  • A shift from manufacturing to services, because the service sector requires more flexible staffing to accommodate demand fluctuations (Asano, Ito, and Kawaguchi, 2011).

  • Economic globalization. As documented by Machikita and Sato (2011) using manufacturing firm-level data, outward FDIs and outsourcing are important determinants of the increase of non-regular work in Japan.

  • The high level of employment protection of regular workers. Japan’s laws are not particularly restrictive in the case of fair dismissals, which are allowed with a 30-day notice period on the basis of incompetence of the employee, violation of disciplinary rules and for economic reasons, and do not oblige companies to provide severance pay. However, as stressed for example by Tsuru (2012), the legal doctrine on unfair dismissal is amongst the strictest in the OECD and orders of reinstatement with back pay are frequent. Anecdotal evidence suggests that in Japan courts apply stricter standards for dismissals of regular workers compared to non-regular ones, and that firms try to avoid dismissals of regular workers by reaching voluntary agreements, which include severance packages (Asano, Ito, and Kawaguchi, 2011; The Nikkei Weekly, 2012).

6. In this paper, we assess the impact of various determinants of labor market duality by estimating a structural model for a panel of OECD countries. The dependent variable is the share of temporary workers in the labor force, which is regressed on variables which we expect to impact the size of labor market duality, such as institutional and demographic factors and relevant macroeconomic control variables. Our regressors, which include proxies of the determinants listed in paragraph 5, are (i) OECD’s indexes of the strictness of employment protection legislation for both regular and temporary workers; (ii) FDI outflows as a ratio to GDP; (iii) the share of employment in services; (iv) the share of female labor participation; (v) union density; (vi) the unemployment rate; (vii) inflation; (viii) the output gap; (ix) government spending on vocational training as a ratio to GDP. We also control for country fixed effects and we include a lag of the dependent variable to allow for autocorrelation.

7. Our results suggest that a higher level of employment protection of regular workers increases labor market duality, while a higher level of protection of temporary workers reduces it. Table 1 shows that the sign of the coefficient on the employment protection index for regular employees (EPR) is positive and statistically significant, meaning that high employment protection for this category of workers increases our proxy of labor market duality (the share of temporary workers). On the other hand, the negative and significant sign on the employment protection index for temporary employees (EPT) implies that increased protection of temporary workers tends to reduce labor market duality. Alternative model specifications and robustness checks, not reported here for brevity, confirm these findings.

Table 1.

Determinants of Labor Market Duality in a Panel of OECD Countries: Regression Results 1/

article image
Source: IMF staff calculations.

T-statistics are reported in parenthesis. * denotes significance at 10% level, ** significance at 5% level, and *** significance at 1% level. Countries included in the panel are Austria, Belgium, Canada, Denmark, Finland, France, Germany, Ireland, Italy, Japan, Netherlands, Norway, Portugal, Spain, Sweden, Switzerland, and United Kingdom. Annual data from OECD and the World Bank for 1985-2010 (or less, depending on availability) are used. A Hausman test rejected the null hypothesis that a random effects model would be statistically different from a fixed effects one.

Fixed effects estimation with default standard errors.

Fixed effects estimation with clustered standard errors.

Arellano-Bond estimation with one lag.

8. Several other institutional, demographic and macroeconomic factors have an impact on the degree of labor market duality. FDI outflows and female labor participation have the expected positive sign—that is to increase temporary employment—and the latter is also significant. The coefficient on the share of employment in services is negative and significant, thus contradicting the findings by Asano, Ito, and Kawaguchi (2011) for Japan. The unemployment rate has a positive impact on duality, suggesting that higher levels of unemployment make it easier for firms to impose temporary contracts on workers. Union density has a negative and significant impact on duality, suggesting that stronger union influence makes it difficult to impose temporary contracts. Inflation has a negative and significant impact. One explanation for this result is that workers could be less motivated to seek employment in better-paid regular jobs when inflation is low. The negative (albeit not significant) coefficient on government spending for vocational training suggests that this kind of policy can reduce duality.

9. Reducing the difference in employment protection between regular and non-regular workers is key to reduce Japan’s labor market duality. In order to estimate how changes in employment protection would affect the overall share of non-regular workers we have followed a two-step procedure. First, we have used the coefficients estimated in the benchmark model (Table 1) to assess the impact of changes in EPT and EPR on the share of temporary workers. We have then estimated how a change in the latter would translate into a change in the overall share of non-regular workers on the basis of a bilateral OLS regression carried on using data for Japan for 1984–2011.

10. The estimated impact of changes in labor employment protection on the share of non-regular workers in Japan is sizeable. Our results show that if, for example, EPR was reduced from Japan’s level of 1.9 to Denmark’s level of 1.6, and the share of non-regular workers in Japan could be reduced from the 2012 level of 35.1 percent to close to 31 percent (text table). The estimated reduction could be even larger for more radical reforms. For example, if EPR was reduced from its current level to the UK’s level of 1.1, the share of non-regular workers could drop below 30 percent. These estimates are based only on ceteris paribus first-round effects. If the reduction in labor market duality results in higher growth, this would reduce unemployment and help exiting deflation. Given the coefficients on unemployment and inflation in the panel regression (Table 1), there could be a second round effect of the reforms which could further reduce duality. An important caveat of this analysis is that the share of temporary workers is not a perfect proxy of labor market duality, because it does not include all non-regular workers. The two-step procedure that we have followed could therefore bias our results upward—in the sense of overestimating the reduction in the share of non-regulars—if the bilateral regression that we carried out for Japan captures correlation rather than causality.

Estimated Impact of Changes in EPT and EPR on Share on Regular Workers

article image
Source: IMF staff calculations.

D. Options for Labor Market Reform in Japan

11. In light of our empirical results, one option that could help reduce labor market duality in Japan could be replacing regular and non-regular contracts currently offered to new hires with a Single Open Ended Contract (SOEC). Under the SOEC employment protection would increase gradually and severance pay would rise with tenure. Introducing a SOEC would therefore drastically reduce firms’ marginal costs of converting non-regular to regular positions. The SOEC would imply lower job security at initial stages compared to current regular employment, but higher job security compared to current non-regular employment. According to the empirical results discussed above, a move in this direction would reduce duality. Introduction of a SOEC could be complemented by other measures aimed at reducing duality while also helping build consensus for the reform package. Another useful reform would be giving workers the right to choose between part-time and full-time work (after a certain tenure period) while maintaining the same hourly wage and legal rights. This would reduce incentives for employers to favor full-time workers to part-time workers in terms of training, because any full-time worker might at any point in time decide to switch to part-time (and vice versa).

12. A possible first step to implement the recommendations given in this paper could be a wider use of limited regular (“gentei seishain”) contracts. Employees who are classified as limited regular workers still enjoy regular workers status and benefits, but with limitations on one or more of the following: (i) job content; (ii) working hours; and (iii) mandatory relocations. Discussions in governmental committees and policy seminars have recently focused on the possibility of expanding the use of this kind of contract, while at the same time clarifying the legal framework for dismissals of limited regular workers, thus reducing uncertainties for firms. A wider use of such contracts, which offer more flexible work arrangements for regular workers while at the same time preserving incentives to invest in workers and their productivity, would be consistent with the recommendations given in this paper.

13. A move toward the so-called “flexicurity” model could also be part of the reform package. Reduced guarantees of life-time employment for regular workers could be complemented by measures aimed at encouraging more job mobility and supporting workers during periods of temporary unemployment. This would mark a shift away from the current Japanese labor market model—characterized by excessive duality and lifetime employment for regular workers—to one more similar to the Danish “flexicurity” model, in which the focus is on protecting workers rather than jobs. One important measure in this regard would be increasing unemployment insurance benefits (text chart).

uA04fig03

Strictness of Employment Protection and Unemployment Benefit Generosity

Citation: IMF Staff Country Reports 2013, 254; 10.5089/9781475563795.002.A004

Source: OECD.1/ Strictness of Employment Protection is measured as EPL for Regular Contracts (2007), Unemployment Benefit Generosity is measured as Net Replacement rate over 5 years of unemployment (2007).
uA04fig04

Real Wage and Productivity Growth

(1991-2011, in percent)

Citation: IMF Staff Country Reports 2013, 254; 10.5089/9781475563795.002.A004

Sources: OECD; IMF staff calculations.

14. Policies aimed at raising wages would also facilitate acceptance of reduced employment protection for regular workers. Real wage and productivity growth trends in the last 20 years suggest that there is scope for wage increases in Japan (see text figures). Against this backdrop, the authorities have recently been using moral suasion and tax incentives to encourage companies to increase wages. Given the importance of increasing wages to exit deflation, the implementation of policies aimed at encouraging wage growth is appropriate.

uA04fig05

Japan

(In percent)

Citation: IMF Staff Country Reports 2013, 254; 10.5089/9781475563795.002.A004

Source: OECD.
uA04fig06

Workers Working More than 49 Hours per Week

(Percent of total employment, 2004-05)

Citation: IMF Staff Country Reports 2013, 254; 10.5089/9781475563795.002.A004

Sources: “Working Time Around the World” ILO and Routledge 2007.

15. For the proposed package of reforms to work, a change in “soft institutions” would also help. In particular, in order to make the reduction in employment protection of regular workers socially and politically acceptable, reforms to improve the work-life balance, such as increased accessibility to annual leave, the right to refuse involuntary relocations, and a reduction of overtime work could be pursued. Although an international comparison shows that Japanese workers work less than those in the U.S. and Italy (text charts), these data likely underestimate the use of overtime for regular workers in Japan because they include part-time workers. Working until very late in the night (substantially above 49 hours per week) among regular workers is very common in Japan and anecdotal evidence suggests that the problem is more serious than in most other advanced countries. One of the reasons for this is that the premium that employers have to pay for work outside statutory hours is comparatively low at 25 percent of the regular hourly wage. Overtime pay also does not count toward the earnings on the basis of which bonuses are calculated, which account for 20–30 percent of worker’s annual earnings in big firms. Measures aimed at eliminating these biases in favor of intensive overtime use would help, but a voluntary change in working practices by firms (“soft institutions”) would also be required.

uA04fig07

Average Annual Hours Actually Worked

(Hours per worker, including part-time workers)

Citation: IMF Staff Country Reports 2013, 254; 10.5089/9781475563795.002.A004

Source: OECD.

E. Conclusions

16. This paper argues that Japan’s dual labor market is likely to entail substantial economic costs. In particular, microeconomic studies and international experiences suggest that excessive labor market duality can reduce TFP due to a negative impact on non-regular workers effort and on firms’ incentives to train them. Excessive duality can also have negative consequences on social cohesion in Japan, thus eroding support for needed structural reforms, which could increase potential growth.

17. On the basis of cross-country empirical evidence on the determinants of labor market duality, the paper proposes some reform options for Japan. Our empirical estimations suggest that reducing the difference in employment protection between regular and non-regular workers would substantially reduce labor market duality in Japan. One reform consistent with these findings is the introduction of an SOEC for all workers, under which employment protection would increase gradually and severance pay would rise with tenure. A SOEC would imply lower job security compared to current regular employment, but higher job security compared to current non-regular employment. A first step in this direction could be encouraging a wider use of limited regular contracts, which would allow for more employment flexibility while preserving incentives to train workers. Reforms to employment contracts would be helped by adopting complementary measures, such as a shifting toward the so-called “flexicurity” model, which supports workers instead of protecting specific jobs.

References

  • Asano, Hirokatsu, T. Ito, and D. Kawaguchi, 2011, “Why Has the Fraction of Contingent Workers Increased? A Case Study of Japan.” Available online at http://www.rieti.go.jp/jp/publications/summary/11030030.html

    • Search Google Scholar
    • Export Citation
  • Damiani, Mirella, F. Pompei, and A. Ricci, 2011, “Temporary Job Protection and Productivity Growth in EU Economies,” Quaderni del Dipartimento di Economia, Finanza e Statistica 87/2011, Università di Perugia, Dipartimento Economia, Finanza e Statistica.

    • Search Google Scholar
    • Export Citation
  • Dolado, Juan Jóse, S. Ortigueira, and R. Stucchi, 2011, “Does Dual Employment Protection Affect TFP? Evidence from Spanish Manufacturing Firms,” Economics Working Papers 1137 (Madrid: Universidad Carlos III).

    • Search Google Scholar
    • Export Citation
  • Fukao, Kyoji; R. Kambayashi, D. Kawaguchi; H. U. Kwong, Y. G. Kim, and I. Yokoyama, 2007, “Deferred Compensation: Evidence from Employer-Employee Matched Data from JapanHi-Stat Discussion Paper Series, No. 187 (Tokyo: Hitotsubashi University).

    • Search Google Scholar
    • Export Citation
  • Gaston, Noel, and T. Kishi, 2007, “Part-Time Workers Doing Full-Time Work in Japan,” Journal of the Japanese and International Economies, Vol. 21, 4, December 2007, pp. 43554.

    • Search Google Scholar
    • Export Citation
  • Kawaguchi, Daiji, R. Kambayashi, Y. G. Kim, H. U. Kwong, S. Shimizutani, K. Fukao, T. Makino, and I. Yokoyama, 2006, “Are Wage-Tenure Profiles Steeper than Productivity-Tenure Profiles? Evidence from Japanese Establishment Data from the Census of Manufacturers and the Basic Survey Wage Structure,” Hi-Stat Discussion Paper Series No. 189 (Tokyo: Hitotsubashi University).

    • Search Google Scholar
    • Export Citation
  • Machikita, Tomohiro, and H. Sato, 2011, “Temporary Jobs and Globalization: Evidence from Japan.” Available online at www.rieti.go.jp/jp/publications/dp/11e029.pdf.

    • Search Google Scholar
    • Export Citation
  • The Nikkei Weekly, 2012, “Fired IBM Japan Employees Cry Foul.” Available online at http://e.nikkei.com/e/ac/tnw/Nni20121210IS1IBMJA.htm.

    • Search Google Scholar
    • Export Citation
  • Tsuru, Kotaro, 2012, “Japan Should Introduce a Severance Payment Scheme for Settlement of Dismissal Disputes”. Available online at http://www.rieti.go.jp/en/papers/contribution/tsuru/19.html?stylesheet=print.

    • Search Google Scholar
    • Export Citation
1

Prepared by Chie Aoyagi and Giovanni Ganelli (both OAP).

Japan: Selected Issues
Author: International Monetary Fund. Asia and Pacific Dept