Abstract
This Selected Issues paper discusses the impact of global financial turmoil on Japan. It describes how close integration in global financial markets has deepened and expanded the channels for spillovers. The paper discusses options for increasing the consumption tax to finance the public pension system. It explains that while parametric reforms have reduced fiscal pressures, the basic pension is becoming a less adequate safety net for the retired. The paper also attempts to explain the declining share of labor income.
III. Why Are Japanese Wages So Sluggish?1
A. Introduction
1. Real wages have stagnated in Japan over the past decade—the real hourly wage has increased by only 1 percent despite solid labor productivity growth. Contrary to what might have been expected, wage growth has failed to pick up during the post-2002 economic recovery and the gap between real hourly wages and productivity has widened to about 20 percentage points (Figure III. 1A)2.
2. These developments should be seen in the context of a longer-term decline in the labor income share in advanced economies. The labor income share, 3 which tracks the evolution of wages relative to productivity, has been falling in many advanced countries, primarily as a result of globalization and technological change. In the past, Japan’s labor income share was higher than in other advanced economies, but has recently dropped toward the G-7 average (Figure III. 1B).
Advanced Economies: Compensation, Productivity, and Labor Share
Citation: IMF Staff Country Reports 2008, 254; 10.5089/9781451820713.002.A003
Advanced Economies: Compensation, Productivity, and Labor Share
Citation: IMF Staff Country Reports 2008, 254; 10.5089/9781451820713.002.A003
Advanced Economies: Compensation, Productivity, and Labor Share
Citation: IMF Staff Country Reports 2008, 254; 10.5089/9781451820713.002.A003
3. Stagnating real wages have become a major policy issue in Japan:
As in other advanced economies, there is a general concern that workers are not getting their fair share of the benefits from technological progress and globalization (see, for example, Financial Times, 2008).
Moreover, overall wage stagnation has been accompanied by a widening of inequalities between high-skilled and low-skilled workers on the one hand; and (well-paid) regular and (poorly-paid) temporary workers on the other.
Weak wage growth has held down private consumption, which in turn has made Japan’s growth highly dependent on foreign demand. From a conjunctural perspective, this dependence makes the Japanese economy vulnerable to a global downturn.
B. Japanese Wages—The Main Facts
4. Researchers have attributed Japan’s sluggish wage growth to a variety of factors:
The impact of foreign competition and technology. The growing integration of large emerging-market economies into the global economy and advances in communication and IT technology have facilitated relocation of production and off-shoring of some production activities to low-cost areas. This process has had two effects on advanced economies including Japan:
➢ International competition has pushed down relative prices of manufactured goods, reducing export revenues. In Japan, the real value added per worker in manufacturing increased by over 40 percent over the past decade. However, the nominal increase in productivity was much smaller, only about 15 percent.4
➢ Firms operating internationally have become more sensitive to cross-country wage differentials and have limited wage growth in their home countries. Indeed, employee’s compensation in Japanese manufacturing trailed nominal productivity gains by about 8 percentage points during 1996-2006.
Deregulation measures. Liberalization measures adopted in Japan in the second half of the 1990s (especially in 1996 and 1999) expanded the list of industries that can hire “non-regular” workers.5 These nonregular workers are much easier to dismiss, have limited social insurance, and have typically earned about 40 percent less than regular workers. The increasing share of low-wage workers, especially in services sectors, has helped to compress aggregate wage growth.
Population aging. The Japanese working-age population has started to shrink—the share of persons between 15-64 years of age has declined from about 70 percent of total population during the 1990s to about 64 percent at present. Given the seniority-based wage system, population aging could have reduced wages for a couple of reasons:
➢ The high-wage baby-boomers are being replaced by low-wage young workers (however, this report does not find any direct evidence supporting this popular view).
➢ Retirees are often re-hired as part-time workers at lower wage rates, as firms would like to maintain skilled labor, while the retirees seek to supplement their pension incomes.
5. Since the impact of globalization and technological progress on wage shares in advanced economies has recently been explored in several studies (notably, Feenstra, 2007, and Jaumotte, Tytell, 2007), the following sub-sections focus on the other two factors that may have had a disproportionate effect on wage growth in Japan—the increasing incidence of non-regular employment and population aging. All the elements will be brought together in an econometric framework in Section III.
Deregulation: the increasing incidence of nonregular employment
Is Japan’s experience with increasing non-regular employment unique?
6. Nonregular workers make up a significant portion of total employment in Japan. The share of nonregular workers has grown from 20 percent of total employment in 1990 to 34 percent in 2007, creating a significant duality between employment types in the labor market. An accurate comparison of the incidence of non-regular employment across countries is not possible due to definitional differences. However, OECD data on part-time employees—typically a large fraction of nonregular employment—indicate that Japan has the third largest share of part-time employment among the sample of 19 advanced economies (Figure III.2A).6 That said, the overall increase in part-time employment over the past couple of decades has been larger in some European countries than in Japan.
Advanced Economies: Labor Market Regulations and Incidence of Part-Time Employment
Citation: IMF Staff Country Reports 2008, 254; 10.5089/9781451820713.002.A003
Advanced Economies: Labor Market Regulations and Incidence of Part-Time Employment
Citation: IMF Staff Country Reports 2008, 254; 10.5089/9781451820713.002.A003
Advanced Economies: Labor Market Regulations and Incidence of Part-Time Employment
Citation: IMF Staff Country Reports 2008, 254; 10.5089/9781451820713.002.A003
7. Hiring of temporary employees has been stimulated by the large differential between employment protection and benefits for regular and temporary workers. The average strictness of employment protection legislation (EPL) in Japan is close to the OECD average but Japan has one of the largest differentials between the EPL indexes for regular and temporary employment in the OECD area—the protection of regular workers is fairly strict, while temporary employment is relatively lightly regulated (Figure III.2B). Moreover, over the past couple of decades, most of the decline in Japan’s EPL can be attributed to a less strict regulation of temporary employment, while the EPL for regular jobs has remained largely unchanged.7 Again, labor market liberalization measures taken elsewhere, especially in Europe, has been more radical than in Japan, which potentially explains the larger increase in part-time employment in these countries.
8. It should be noted that a high percentage of nonregular employees are comfortable with their flexible working status. In particular, the labor market reforms helped previously disadvantaged groups such as women with children (or retirees) to enter (or remain in) the labor force. Indeed, the labor participation of women has increased from 60 percent to 66 percent of the working-age female population since the early 1990s. Switching to regular contracts is mostly sought by young cohorts (MHLW, 2003).
Who hires non-regular workers?
9. Most of the increase in non-regular employment has occurred in services, that is, in the sectors, which do not face direct international competition. In many services sectors, This pattern generally applies also to other advanced economies (OECD, 2004). non-regular employees now make up 50 percent or more of total workforce. Despite concerns about the effects of international competition, the number of non-regular, low-wage, workers has increased only modestly in manufacturing and their share has remained low at 20 percent of manufacturing employment (Figure III.3A).
Japan: Sectoral Trends in Non-Regular Employment, Wages, Productivity, And Labor Income Shares
Citation: IMF Staff Country Reports 2008, 254; 10.5089/9781451820713.002.A003
Japan: Sectoral Trends in Non-Regular Employment, Wages, Productivity, And Labor Income Shares
Citation: IMF Staff Country Reports 2008, 254; 10.5089/9781451820713.002.A003
Japan: Sectoral Trends in Non-Regular Employment, Wages, Productivity, And Labor Income Shares
Citation: IMF Staff Country Reports 2008, 254; 10.5089/9781451820713.002.A003
What have been the wage outcomes?
10. Wage growth has been significantly smaller in industries with a larger proportion of nonregular workers. Looking at sectoral data, a greater increase in the share of non-regular employment has typically been associated with lower wage growth (Figure III.3B). In many service sectors, average wages fell significantly over the past five years.
11. Manufacturing companies granted their employees larger wage increases than other, more domestically oriented, sectors. However, wage increases in manufacturing have been well below productivity gains. This would seem consistent with the hypothesis about international wage competition but, clearly, domestic factors have also been at play:manufacturers have not been pressed to offer higher wages as the domestically-oriented sectors have been offering much less attractive compensation packages.
12. Notwithstanding the higher proportion of non-regular workers and lower average wages, the labor income share in non-financial services declined only modestly over the past decade and, surprisingly, has even increased in recent years(Figure III.3C). These developments have reflected low labor productivity growth of Japan’s service sectors (Figure III.3D) 8. Put differently, although services sectors have been hiring extensively at low wages, profitability of these sectors has not improved as a result of hiring non-regular employees. Instead, most of the decline in aggregate labor income share can be attributed to the manufacturing sector.
13. Compared with other advanced economies, growth within the Japanese economy has been unbalanced. The cross-country data suggest that Japan’s non-financial services sector (accounting for about 2/3 of total employment) has achieved one of the smallest productivity gains among the sample of 15 large advanced economies over the past decade—in sharp contrast with significant productivity growth in manufacturing that places Japan above the average of its peers.
14. The unbalanced nature of Japan’s growth has likely contributed to the lackluster average wage growth. Since wages in manufacturing could not increase as much as productivity due to globalization and technological change, and wages in the non-financial services sector have been constrained by low productivity growth, aggregate wages have trailed productivity gains. Indeed, cross-country evidence suggests that countries with sizeable differentials between manufacturing and services productivity growth have accumulated the largest gaps between wages and productivity gains in recent years, despite falling unemployment.
Advanced Economies: Wage Gaps and Productivity Differentials
(2002-06; in percent)
Citation: IMF Staff Country Reports 2008, 254; 10.5089/9781451820713.002.A003
Sources: OECD, Annual National Accounts; and IMF staff calculations.Notes: Wage-productivity gap is defined as the percent difference between real wage and productivity growth per full-time employee during 2002-06. Productivity differential equals the difference between manufacturing and nonfinancial services productivity growth over the same period.Advanced Economies: Wage Gaps and Productivity Differentials
(2002-06; in percent)
Citation: IMF Staff Country Reports 2008, 254; 10.5089/9781451820713.002.A003
Sources: OECD, Annual National Accounts; and IMF staff calculations.Notes: Wage-productivity gap is defined as the percent difference between real wage and productivity growth per full-time employee during 2002-06. Productivity differential equals the difference between manufacturing and nonfinancial services productivity growth over the same period.Advanced Economies: Wage Gaps and Productivity Differentials
(2002-06; in percent)
Citation: IMF Staff Country Reports 2008, 254; 10.5089/9781451820713.002.A003
Sources: OECD, Annual National Accounts; and IMF staff calculations.Notes: Wage-productivity gap is defined as the percent difference between real wage and productivity growth per full-time employee during 2002-06. Productivity differential equals the difference between manufacturing and nonfinancial services productivity growth over the same period.Dualities between large and small firms
15. The duality in productivity performance between manufacturing and services is closely linked to dualities between small and large enterprises. Large firms,9 especially in manufacturing, have generally managed to increase productivity and offset rising input costs with greater success than small firms (Figure III.4A). However, the small firms account for most of total employment—about 70 percent.10 Their low productivity has been a drag on economy-wide wages because the aggregate productivity gains have been concentrated among large manufacturers—that is, precisely those firms that are likely to limit wage increases due to the globalization of labor and product markets.
16. There are further similarities between the dualities of manufacturing vs. services and large vs. small firms. Small firms do not seem to have become more profitable as a result of hiring non-regular employees—the labor income shares in small firms have remained roughly unchanged. Most of the decline in aggregate wage share has occurred in the segment of large corporations (Figure III.4B).11
Japan: Productivity and Labor Share by Company Size
Citation: IMF Staff Country Reports 2008, 254; 10.5089/9781451820713.002.A003
Japan: Productivity and Labor Share by Company Size
Citation: IMF Staff Country Reports 2008, 254; 10.5089/9781451820713.002.A003
Japan: Productivity and Labor Share by Company Size
Citation: IMF Staff Country Reports 2008, 254; 10.5089/9781451820713.002.A003
Population aging
17. In recent years, the Japanese working-age population has started to shrink. The share of population over 65 years old has increased from 15 to about 21 percent over the past decade, the largest increase among advanced economies.
18. Re-employment of retired workers has put some downward pressure on wages. The practice of offering part-time employment to retirees has become common as firms try to keep skilled employees while shedding labor costs. As a result, the relative wage of elderly workers has declined (Figure III.5A).
Japan: The Impact of Population Aging on Average Wages
Citation: IMF Staff Country Reports 2008, 254; 10.5089/9781451820713.002.A003
Japan: The Impact of Population Aging on Average Wages
Citation: IMF Staff Country Reports 2008, 254; 10.5089/9781451820713.002.A003
Japan: The Impact of Population Aging on Average Wages
Citation: IMF Staff Country Reports 2008, 254; 10.5089/9781451820713.002.A003
19. More generally, the shape of the earnings-age profile has changed considerably during this decade. The peak earnings age has shifted forward from the 50–54 year age group to the 45–49 year group. Since the earnings of workers aged 40–44 has also increased relative to the average, the wage system has effectively become less seniority-based.12 Besides the elderly workers, another group whose earnings fell relative to the average wage is the cohort of the currently 30–34 year old employees, again mostly due to the expanding part-time employment. In this regard, changes in the earnings-age profiles highlight the equity concerns as different age groups have been influenced by globalization and deregulation differently.
20. For comparison, the earnings-age profiles in Japan and the United States continue to display some notable differences (Figure III.5B). In the United States, peak earnings are reached earlier and the peak earnings period last even longer than in Japan. Moreover, older workers are paid a relatively higher wage in the United States than in Japan.
21. Given the existing (albeit weaker) seniority-based wage system, the retirement of the high-wage baby-boomers and their replacement with young, low-wage, employees has also been cited as another cause of sluggish aggregate wage growth. However, the negative composition effect from population aging per se has likely not contributed to sluggish wages in Japan. Calculations of aggregate wages under the assumption of constant earnings-age profiles suggests that demographic shifts may have had quite the opposite effect (Figure III.5C).13 In other words, the movement of age cohorts “along the earnings-age curve” (Figure III. 5A) cannot easily explain why aggregate wages have been sluggish in recent years. The intuition for this result is that while the share of employees behind their earnings peak (over 60 years) has risen, the share of low-wage young workers (under 35 years) has fallen as well—thus limiting the mechanical impact of demographics on the aggregate average wage.
22. That said, one cannot rule out the possibility of the negative composition effect from demographics in the coming years. Flattening of the simulated wage paths in recent years (see again Figure III.5C) suggests that the demographic structure could be at the “tipping point” at which it is about to start exerting a downward pressure on average wages. However, additional calculations using the U.N. population projections for 2010 and 2015 (not reported here) also do not predict any negative composition effects from expected demographic changes.
23. Indeed, aggregate wage sluggishness can be accounted for by the changing shape of the earnings-age profiles (Figure III.5D), which have in turn been linked to factors such as greater openness, international wage competition, and legal changes that encouraged non-regular employment.
C. Japanese Wage Developments in the International Perspective—Econometric Analysis
Modeling labor income share
24. The previous section has illustrated how a variety of factors such as globalization or labor and product market dualities may explain stagnating wages in Japan. This section analyzes Japanese wage developments in a more general framework by estimating a model that links labor income shares in a group of advanced economies to a variety of factors, including technological change, labor and product market policies, and other controls such as demographics or business cycle.
25. The explanatory variables of the model are grouped as follows:14
Technological change. As suggested in earlier research (for example, IMF, 2007a), greater use of information and communication technology can, in the short term, reduce labor share by reducing demand for low-skilled workers. But labor share could also rebound once the skill adjustment to ICT investment is completed. This effect is captured in the model by assuming a quadratic relationship between technology and labor share.
Globalization and competition. Increasing openness can make corporations more sensitive to international wage differentials (thus lowering labor share)—OECD, 2007; it can also make domestic markets more competitive, reducing any excess profits (with an ambiguous impact on labor share).
Labor and product market institutions. Employment protection and product market regulation both limit competition in markets, creating “rents.” Changes in labor and product market policies and institutions are therefore likely to affect the labor market share. Moreover, high unemployment replacement rates can create disincentives to work (OECD, 2004b), reducing the labor income share.
Dualities in labor and product markets.
➢ Incidence of part-time employment. Nonregular workers have lower bargaining power; a higher share of part-time workers in total employment could therefore be associated with lower labor income shares. Moreover, more flexible work arrangements may facilitate entry of new workers into the labor force. Since these workers are likely of low productivity, the marginal product of labor could fall, pushing down the aggregate wage.
➢ Differential between productivity of manufacturing and nonfinancial services. As discussed above, any large differences between “tradable” and “nontradable” productivities could reduce the wage share in advanced economies as the labor market gets more globalized.
➢ Differential in ICT investment between manufacturing and non-financial services. This variable is included to reflect the low use of ICT in Japan’s services sectors.
Other controls include demographics, output gap and a measure of real exchange rate overvaluation.
26. As is apparent from the discussion above, the panel regressions include—besides “deep parameters” such as openness, institutions or demographics—also several “outcome” measures, specifically the incidence of part-time employment and productivity differentials. These outcomes should in principle be linked to their underlying determinants—for example, the productivity differential between manufacturing and services could be modeled as a function of sector-specific labor and product market regulations, market contestability and other factors. However, such a full breakdown into deep parameters is not feasible because many of the existing structural and institutional measures are of qualitative nature, change infrequently, and the relationship between the deep parameters and outcomes is likely non-linear. This gives the inclusion of outcome measures in regressions some merit—with the qualification that, in particular, any changes in the labor market share due to labor and product market dualities should be interpreted using country-specific information in a broader context of technological and regulatory change, increasing openness and so forth.
Estimation results
27. Table III.1 presents estimation results from a panel data model estimated over a sample of 15 advanced economies during 1980-2005. The estimation results suggest that:
Labor Share in Advanced Economies: Panel Regression Estimates
Labor Share in Advanced Economies: Panel Regression Estimates
| Group | Variable | (1) | (2) | (3) |
|---|---|---|---|---|
| Technology | Share of ICT in non-residential capital | -0.690 | -0.351 | -0.370 |
| stock | (4.14)*** | (1.77)* | (2.06)* | |
| Share of ICT capital squared | 0.019 | 0.012 | 0.012 | |
| (4.59)*** | (2.71)** | (2.96)** | ||
| Globalization | Trade openness | -0.077 | -0.096 | -0.107 |
| (3.82)*** | (7.25)*** | (3.75)*** | ||
| Dualities | Share of part-time employment | -0.206 | -0.238 | -0.228 |
| (1.99)* | (2.79)** | (2.89)** | ||
| Differential between labor productivity | -2.866 | -2.235 | -2.082 | |
| in manufacturing and services (in logs) | (2.31)** | (1.80)* | (1.40) | |
| Differential between ICT capital | -1.080 | -1.612 | -1.477 | |
| in manufacturing and services (in logs) | (2.07)* | (3.59)*** | (3.06)*** | |
| Labor and product | Unemployment benefits | -0.233 | -0.239 | -0.245 |
| market policies | (7.93)*** | (7.26)*** | (7.79)*** | |
| Product market regulation | 0.503 | 0.613 | 0.547 | |
| (2.10)* | (1.47) | (1.17) | ||
| Employment protection legislation | -1.513 | -1.994 | -2.049 | |
| (2.44)** | (2.61)** | (3.10)*** | ||
| Cycle | Output gap | -0.152 | -0.123 | -0.118 |
| (4.58)*** | (2.71)** | (2.60)** | ||
| Other controls | Population 15-64 as percent | … | … | 0.025 |
| of total population | (0.12) | |||
| Deviation of REER from trend | … | … | -0.016 | |
| (0.70) | ||||
| Constant | 80.19 | 77.218 | 76.31 | |
| (41.39)*** | (31.25)*** | (6.14)*** | ||
| Fixed effects | Yes | Yes | Yes | |
| Time effects | … | Yes | Yes | |
| Summary statistics | Sample | 1980-2005 | 1980-2005 | 1980-2005 |
| Observations | 312 | 312 | 312 | |
| Adjusted R-squared | 0.88 | 0.90 | 0.90 |
Labor Share in Advanced Economies: Panel Regression Estimates
| Group | Variable | (1) | (2) | (3) |
|---|---|---|---|---|
| Technology | Share of ICT in non-residential capital | -0.690 | -0.351 | -0.370 |
| stock | (4.14)*** | (1.77)* | (2.06)* | |
| Share of ICT capital squared | 0.019 | 0.012 | 0.012 | |
| (4.59)*** | (2.71)** | (2.96)** | ||
| Globalization | Trade openness | -0.077 | -0.096 | -0.107 |
| (3.82)*** | (7.25)*** | (3.75)*** | ||
| Dualities | Share of part-time employment | -0.206 | -0.238 | -0.228 |
| (1.99)* | (2.79)** | (2.89)** | ||
| Differential between labor productivity | -2.866 | -2.235 | -2.082 | |
| in manufacturing and services (in logs) | (2.31)** | (1.80)* | (1.40) | |
| Differential between ICT capital | -1.080 | -1.612 | -1.477 | |
| in manufacturing and services (in logs) | (2.07)* | (3.59)*** | (3.06)*** | |
| Labor and product | Unemployment benefits | -0.233 | -0.239 | -0.245 |
| market policies | (7.93)*** | (7.26)*** | (7.79)*** | |
| Product market regulation | 0.503 | 0.613 | 0.547 | |
| (2.10)* | (1.47) | (1.17) | ||
| Employment protection legislation | -1.513 | -1.994 | -2.049 | |
| (2.44)** | (2.61)** | (3.10)*** | ||
| Cycle | Output gap | -0.152 | -0.123 | -0.118 |
| (4.58)*** | (2.71)** | (2.60)** | ||
| Other controls | Population 15-64 as percent | … | … | 0.025 |
| of total population | (0.12) | |||
| Deviation of REER from trend | … | … | -0.016 | |
| (0.70) | ||||
| Constant | 80.19 | 77.218 | 76.31 | |
| (41.39)*** | (31.25)*** | (6.14)*** | ||
| Fixed effects | Yes | Yes | Yes | |
| Time effects | … | Yes | Yes | |
| Summary statistics | Sample | 1980-2005 | 1980-2005 | 1980-2005 |
| Observations | 312 | 312 | 312 | |
| Adjusted R-squared | 0.88 | 0.90 | 0.90 |
Technology and globalization have been important determinants of the labor income share. Technology indeed seems to enter the model nonlinearly, reducing the wage share at low levels of technology use but later raising it once some of the low-skill occupations are eliminated. Trade openness tends to reduce the labor share with a fairly large elasticity—the labor income share falls in advanced countries by about 1 percentage point for a 10 percentage point increase in the ratio of export and imports to GDP.15
Labor and product market dualities have significant explanatory power in the regressions. The coefficient on the share of part-time employment is large and highly statistically significant—for every 5 percentage points increase in the share of part-time employment, the labor income share falls by 1 percentage point. The differential between manufacturing and services productivity also enters with a large coefficient—in Japan’s case, it contributed about 1.5 percentage point to the decline in labor share.
As expected, high unemployment benefits and strict labor market regulations reduce the labor income share, in line with previous research.
Labor share tends to decline during economic expansions but on the whole, the elasticity is small. Demographics and a crude measure of exchange rate overvaluation do not enter significantly.
Decomposition of changes in the labor income share
28. The estimated coefficients can be used to decompose changes in the labor income share to the contribution of various explanatory variables. In Japan, the labor share fell by 4.5 percentage points during 1990-2005.16 This decline be split into: a -1.8 percentage point contribution of technology, a -0.6 percentage point contribution of openness, and a -3.3 percentage point contribution of dualities (Figure III.6A).17 Within the dualities group, the largest contribution is from the rising share of part-time employment (-1.4 percentage point) and productivity differential between manufacturing and services (-1.5 percentage point). The direct impact of changes in labor and product market regulations is estimated positive, at 0.3 percentage points, mostly because unemployment benefits relative to the median wage declined over the period. However, caution is needed in interpreting this result because liberalization measures have reduced the labor share through indirect channels, most notably by contributing to an increase in non-regular employment.
Advanced Economies: Decomposition of Changes in Labor Income Shares
Citation: IMF Staff Country Reports 2008, 254; 10.5089/9781451820713.002.A003
Advanced Economies: Decomposition of Changes in Labor Income Shares
Citation: IMF Staff Country Reports 2008, 254; 10.5089/9781451820713.002.A003
Advanced Economies: Decomposition of Changes in Labor Income Shares
Citation: IMF Staff Country Reports 2008, 254; 10.5089/9781451820713.002.A003
29. In some European countries, dualities have also contributed significantly to falling labor share. For example in Germany and Italy, a rapid increase in part-time employment reduced wage shares by more than one percentage point, while the differential between manufacturing and services productivity gains contributed another ½-1 percentage point to the labor share decline.
30. The experience of the United Kingdom and United States provides an interesting contrast. The overall decline in labor share was only about ½ of Japan’s during 1990-2005 (Figure III.6B). In these two Anglo-Saxon economies, technology tended to boost the labor share as ICT investment has surpassed the threshold level: the share of ICT capital in total non-residential capital stock is higher by about 2/3 in the United Kingdom. and the United States compared with Japan. Also, dualities seem to play a relatively smaller role, reflecting a limited increase in part-time employment.18
D. Summary and Policy Recommendations
31. Wages have been trailing productivity gains in most advanced economies over the past couple of decades. This paper confirms the importance of technological change, globalization, and labor market policies in explaining these developments. In case of Japan, the analysis highlights the role of the factors that have contributed to the build-up of labor and product market dualities—between regular and non-regular workers, large and small companies, or manufacturing and services.
32. Japan’s economic dualities may explain why wage growth has been more disappointing in Japan than in some other advanced economies. The large difference in the employment protection and benefits between regular and nonregular workers stimulates the firms with low profitability (often SMEs) to offset rising input costs by offering mostly lower-paid temporary positions. As a result, profitable companies have a reduced incentive to agree on attractive compensation packages. Besides the large difference in employment protection between regular and nonregular workers, the unbalanced nature of Japanese growth together with greater openness may also have contributed as manufacturers have limited wage increases, while services sectors did not generate labor productivity growth sufficient to support higher wages. Cross-country evidence suggests that countries with large differentials between manufacturing and services productivity growth have experienced the largest declines in the aggregate wage share in recent years, despite low unemployment.
33. From the short-term perspective, very recent amendments to the part-time employment legislation have provided some boost to full-time employment, while average nominal wages have picked up. However, many of the factors underlying wage sluggishness in Japan are deep-seated and policy changes are needed to put wages firmly on an upward trajectory:
Reforms aimed at increasing productivity in services (or, from another point of view, at small enterprises) and reducing the gap in employment protection and benefits between regular and nonregular workers could help boost per-capita incomes. However, policymakers may face here a trade-off between higher productivity and employment level, especially in regions with limited short-term growth prospects.
The general direction of future labor market reforms should be toward lower employment protection. Rather than increasing employment protection for temporary workers, flexibility should be increased for the permanent contracts (OECD, 2008). It is notable that economies with flexible labor markets such as the United Kingdom, the United States, and New Zealand have observed smaller declines, or even increases, in labor income shares during the recent cycle. Moreover, the more equal treatment of regular and nonregular workers would help ease equity concerns.
Certain financial reforms would also be useful, either by facilitating emergence of new high-productivity enterprises, or by encouraging restructuring of existing inefficient businesses. Measures aimed at encouraging greater FDI could also force restructuring, while potentially helping local SMEs gain access to foreign markets, helping to boost productivity (IMF, 2007b).
Annex I. Update on the Privatization of Japan’s Postal Savings and Insurance:Key Risks and Challenges1
A. October 1st, 2007: The First Step towards Privatization
1. As a first step towards privatization, the government has reorganized Japan Post into four separate businesses under a single holding company. On October 1, 2007, Japan Post Holdings (JPH) was created with full ownership in four subsidiaries: Japan Post Service, Japan Post Network (JPN), Japan Post Bank (JPB) and Japan Post Insurance (JPI). Under the 10-year privatization plan, the government will require Japan Post Holdings to divest its shares in JPB and JPI by 2017, while at the same reducing the government’s stake in JPH to one-third. Plans are also underway to list shares in JPB and JPI on the Tokyo Stock Exchange starting as early as 2010.
2. The legal separation of the postal network from the savings and insurance arms has opened the door for private institutions to compete on a more level playing field with JPB and JPI. With its 24,000 branches nationwide, the postal network has long provided an important distribution advantage for its former savings and insurance units. While the postal network will continue as a retail window for JPB and JPI (from now on against the payment of fees), the network will also be allowed to distribute financial services originated and developed by other private financial institutions.
3. In addition, JPB and JPI will now face similar regulatory and supervisory treatment as other private institutions. In October 2007, JPB and JPI were brought under the Banking Law and the Insurance Business Law, respectively. As such, they are now regulated and supervised by the FSA under the same set of rules, risk-controls and disclosure requirements as other financial institutions. In addition, new deposits and policies at the JPB and JPI no longer carry an explicit government guarantee, but instead will be covered under the public insurance schemes with the same limits as for private institutions.2
B. Key Challenges and Risks
4. JPB poses a challenge to commercial banks, particularly at the regional level. JPB’s deposits have come down by almost a third since 2000, as households have shifted away from deposits into higher-yielding financial products. Despite this reduction, JPB’s deposit base remains enormous, roughly two-third the size of city banks combined, and almost equal to that of all regional banks. While its market share in the urban areas is relatively small (in Tokyo it is less than 10 percent), it still maintains a large presence in the rural areas. In response, regional banks have taken steps to join forces, in some areas and in some instances seek out business alliances with JPB.
Postal Savings: Outstanding Deposits
(In trillions of yen)
Citation: IMF Staff Country Reports 2008, 254; 10.5089/9781451820713.002.A003
Source: CEIC.Postal Savings: Outstanding Deposits
(In trillions of yen)
Citation: IMF Staff Country Reports 2008, 254; 10.5089/9781451820713.002.A003
Source: CEIC.Postal Savings: Outstanding Deposits
(In trillions of yen)
Citation: IMF Staff Country Reports 2008, 254; 10.5089/9781451820713.002.A003
Source: CEIC.5. Enhancing the profitability and ensuring the soundness of JPB will be important for its future viability as a private entity. JPB’s net interest margin is only ¾ of a percent—less than half that of private banks—with a return on assets of only ½ percent. The meager returns are due mainly to its large concentration in low-risk assets such as JGBs. In order to raise its profitability (and prepare for future listing), JPB is aiming to expand into new business lines and take on more risk. Already, it has been permitted to enter into credit card businesses and intermediate housing loans. However, JPB does not have the adequate risk management systems for conducting direct mortgage lending and instead initially sells mortgages and limited number of financial products originated by other private financial institutions. It is important that JPB have the systems in place to manage properly these new risks.
Total Deposits, end 2007
(In trillions of yen)
Citation: IMF Staff Country Reports 2008, 254; 10.5089/9781451820713.002.A003
Source: CEIC.Total Deposits, end 2007
(In trillions of yen)
Citation: IMF Staff Country Reports 2008, 254; 10.5089/9781451820713.002.A003
Source: CEIC.Total Deposits, end 2007
(In trillions of yen)
Citation: IMF Staff Country Reports 2008, 254; 10.5089/9781451820713.002.A003
Source: CEIC.6. The privatization process will also need to manage the risks to financial markets from potential portfolio shifts by JPB and JPI. Reflecting their historical role in securing funding for public lenders and infrastructure projects through the Fiscal Investment and Loan Program (FILP), JPB and JPI have amassed huge portfolios in risk-free assets, mainly JGBs and Fiscal Loan Funds deposits. Given the size of their balance sheets, any shift in investment strategy by the JPB and JPI, such as into stocks and foreign securities, could have a significant impact on asset prices, as well as on capital flows.
Asset Allocation as Share of Total Assets
Citation: IMF Staff Country Reports 2008, 254; 10.5089/9781451820713.002.A003
Sources: CEIC and staff estimates.Asset Allocation as Share of Total Assets
Citation: IMF Staff Country Reports 2008, 254; 10.5089/9781451820713.002.A003
Sources: CEIC and staff estimates.Asset Allocation as Share of Total Assets
Citation: IMF Staff Country Reports 2008, 254; 10.5089/9781451820713.002.A003
Sources: CEIC and staff estimates.
Asset Allocation as a Share of Total Assets
Citation: IMF Staff Country Reports 2008, 254; 10.5089/9781451820713.002.A003
Sources: CEIC and the Life Insurance Association of Japan.Asset Allocation as a Share of Total Assets
Citation: IMF Staff Country Reports 2008, 254; 10.5089/9781451820713.002.A003
Sources: CEIC and the Life Insurance Association of Japan.Asset Allocation as a Share of Total Assets
Citation: IMF Staff Country Reports 2008, 254; 10.5089/9781451820713.002.A003
Sources: CEIC and the Life Insurance Association of Japan.C. What is Being Done to Deal with these Challenges?
7. During the transition, the government is maintaining tight controls over the business activities of JPB and JPI. The expansion of JPB and JPI into new business activities and the management of their assets is still subject to government approval. Fair competitive relationships and business conditions of the institutions shall be considered when the government makes decisions on their business expansions, upon hearing an opinion from the Postal Services Privatization Committee (PSPC). In December, the government allowed the JPB to engage in a limited number of operations, including syndicated loans, swaps and interest rate futures, and reverse repos. In addition, the government also allowed the JPB to engage in credit card businesses, and intermediary businesses of housing loans and life insurance products including variable annuities in April. However, it maintained the ¥10 million per person cap on postal savings deposits to limit JPB’s expansion.
8. The tight rules on asset management also suggest that any portfolio shifts will likely be gradual. Time deposits at JPB and policies at JPI that are contracted before October 2007 continue to benefit from a full government guarantee, but in return, the corresponding funds are required to be invested in safe assets and managed conservatively. Only deposits and policies contracted post October 2007 will be allowed to be managed with greater flexibility. As a result, JPB and JPI now operate two sets of books—a “pre-privatization” and “post-privatization” book. With the “post-privatization book” likely to grow slowly, the diversification of assets at JPB and JPI away from safe assets is expected to be very gradual, limiting the potential impact on financial markets and capital flows.
References
Bank of Japan, 2005, The State of the Japanese Economy: From the Perspective of Employment and Income, Tokyo, January.
Bassanini, Andrea and Duval, Romain, 2006, “Employment Patterns in OECD Countries: Reassessing the Role of Policies and Institutions,” OECD Economics Department Working Paper No. 486.
Bentolila, Samuel and Saint-Paul, Gilles, 2003, “Explaining Movements in the Labor Share,” Contributions to Macroeconomics, Berkeley Electronic Press.
Blanchard, J. Olivier, Nordhaus, D. William, and Phelps, S. Edmund, 1997, “The Medium Run,” Brookings Papers on Economic Activity, 1997(2), pp. 89–158.
Cabinet Office (CAO), 2006, Annual Report on the Japanese Economy and Public Finance, Chapter 3.
European Commission, 2007, Employment in Europe 2007: The Labour Income Share in the European Union, Brussels.
Feenstra, C. Robert, 2007, “Globalization and Its Impact on Labour,” The Vienna Institute for International Economic Studies, Working Paper.
Financial Times, 2008, “Japan Exporters Keep Check on Pay Rises,” by Mariko, Sanchanta, , March 12.
Guscina, Anastasia, 2006, “Effects of Globalization on Labor’s Share in National Income,” IMF Working Paper 06/294 (Washington: International Monetary Fund).
International Monetary Fund, 2007, World Economic Outlook, April 2007: The Globalization of Labor, Chapter 5 (Washington).
International Monetary Fund, 2007b, “Japan: Boosting Productivity in Services—Priorities for Deregulation,” Chapter V in Japan: Selected Issues, July.
Jaumotte, Florence and Tytell, Irina, 2007, “How Has The Globalization of Labor Affected the Labor Income Share in Advanced Countries?” IMF Working Paper 07/298 (Washington: International Monetary Fund).
Ministry of Health, Labor, and Welfare (MHLW), 2003, General Survey on Actual Conditions of Diversification in Employment Styles.
Organisation for Economic Co-operation and Development (OECD), 2002, “Taking the Measure of Temporary Employment,” Chapter 3 in OECD Employment Outlook 2002.
Organisation for Economic Co-operation and Development (OECD), 2004a, “Employment Protection Regulation and Labour Market Performance,” Chapter 2 in OECD Employment Outlook 2004.
Organisation for Economic Co-operation and Development (OECD), 2004b, “Wage-setting Institutions and Outcomes,” Chapter 3 in OECD Employment Outlook 2004, Paris.
Organisation for Economic Co-operation and Development (OECD), 2007, “OECD Workers in the Global Economy: Increasingly Vulnerable?” Chapter 3 in OECD Employment Outlook 2007.
Organisation for Economic Co-operation and Development (OECD), 2008, OECD Economic Surveys: Japan.
This paper was prepared by Martin Sommer.
The accumulated gap between nominal hourly compensation and labor productivity is much smaller—about 10 percentage points. This is due to a divergence in price trends: the CPI index fell by -0.5 percent during 1996-2006, while the GDP deflator used for the calculation of real productivity fell by 10 percent over the same period, in part due to rapidly falling prices of capital goods.
In the chapter, labor income share is defined as the share of employees’ and self-employed persons’ compensation in GDP (for the aggregate economy) or gross value added (for sectors). The methodology of adjustment for self-employed persons’ income follows European Commission (2007).
The manufacturing deflator fell by 20 percent’much faster than the GDP deflator or CPI (footnote 2).
Nonregular employees include part-time workers, workers dispatched through temporary agencies, short-term hires and others (see Bank of Japan, 2005 for details).
This comparison potentially makes the Japan’s labor market dualities appear less pronounced because the largest increase in non-regular employment in Japan in the past couple of years was in the segment of shortterm hires and workers dispatched through temporary agencies. See OECD (2002) for a detailed discussion of various pitfalls in comparing temporary employment data across OECD member countries.
This pattern generally applies also to other advanced economies (OECD, 2004).
The low labor productivity of Japanese services is related to the barriers to entry, limited contestability of local markets, low use of information and communication technology, and other impediments. See IMF (2007b) for details.
Large firms are defined as corporations with capital exceeding Y100mn.
Small firms also contributed most to the employment increase in recent years.
Some caution is necessary when making productivity comparisons between large and small companies. The MoF Survey does not report hours worked and it is therefore not possible to control for the increase in part-time and other forms of non-regular employment. Moreover, the MoF Survey and national account samples are different.
In the late 1990s and 2000s, a number of companies introduced elements of performance-based compensation (Bank of Japan, 2005).
In order to assess the impact of changing demographics on aggregate wages, one can calculate the counterfactual average wage by holding the earnings-age profile unchanged at some year’s level (for instance, 2002), while accounting for the actual age structure of employees. Figure III.5C presents three versions of such calculation using vintage years 1999, 2002, and 2006.
The model estimated in this section builds upon the econometric framework of Jaumotte and Tytell (2007) but is expanded to capture factors that reflect dualities. The data sources for variables are OECD, Ameco, EUKlems, and IMF databases, and Bassanini and Duval (2006). Other recent studies of wage developments in advanced economies include Bentolila and Saint-Paul (2003), Guscina (2006), European Commission (2007), and IMF (2007a). Earlier research includes Blanchard, Nordhaus and Phelps (1997).
Bank of Japan (2005) presents interesting sectoral evidence that larger import penetration tends to be associated with a greater weight of skilled labor in total labor cost.
Year 1990 was selected as the starting point for this calculation because the early 1990s coincide with the increasing integration of large emerging-market economies into the global economy and regulatory reforms in many countries. Moreover, labor shares were adjusting downward during the 1980s in response to the previous unsustainable rise in the period of oil shocks.
A wave of corporate restructuring in Japan during the 1990s and 2000s has accelerated the reallocation of employment from manufacturing to services. While the restructuring process could on its own explain some of the wage sluggishness, its impact on the labor share (i.e., wages relative to productivity) is likely to have been positive as service sectors typically have higher income shares than manufacturing
Of course, rising wage inequalities have raised public concerns also in these countries, nothwithstanding the relatively solid growth of average wages (see, for example, IMF, 2007a).
Prepared by Romuald Semblat (APD)
However, both institutions will still benefit implicitly from having the government as their sole shareholder.
