Is Korean labour market dualism different?

The preceding sections have shown that labour market dualism is strong in Korea in the sense that a large share of the workforce are non-regular workers or are employed in SMEs and service sector firms, and these workers experience significantly worse pay and other employment conditions than regular workers in large firms. Furthermore, mobility into regular jobs appears to be difficult for many non-regular workers. In analysing the appropriate policy response, it is useful to compare these patterns with those observed in other OECD countries, to highlight Korea’s specificities and possible ways to handle them. Indeed, this sub-section finds both similarities and differences between the labour market duality found in Korea and that characterising other OECD countries.

Figure 2.4 in Chapter 2 shows that earnings inequality has risen particularly rapidly in Korea since the mid-1990s, at the same time that non-regular work was expanding. Indeed, Korea now has the highest incidence of low-paid work among full-time workers in the OECD area and the third highest level of overall earnings dispersion, just below the United States and Israel (Figure 3.4). These patterns need to be interpreted carefully since data sources and definitions are not fully harmonised across the countries shown. Nonetheless, these data suggest that growing labour market dualism - in the form of an increased incidence of non-regular work and rising pay penalties for non-regular, SME and service sector employment - probably has had a strong impact in raising earnings inequality in Korea. These developments have increased the need to further develop the income support system for low-income workers (Chapter 2).

Figure 3.4. The incidence of low-paid work and earnings dispersion*

• a) The incidence of low pay refers to the share of full-time workers earning less than two-thirds of median earnings. Earnings include both fixed pay and bonuses, except for Japan where it is limited to fixed pay. When only fixed pay is considered for Korea, the incidence of low pay declines to 22%, still substantially higher than the 14% rate in Japan.
• b) Earnings dispersion is measured by the ratio of 9th to 1st deciles limits of earnings for full-time workers.
• c) Data refer to 2008 for Iceland, 2009 for the Czech Republic and France, and to 2010 for Australia, Austria, Belgium, Denmark, Finland, Germany, Greece, Hungary, Ireland, Israel, Italy, Poland, Portugal, the Slovak Republic, Spain, Sweden and Switzerland.
• d) Unweighted average of the countries shown above.
• e) Information on data for Israel is available at: http:/dx.doi.org/10.1787/888932315602.

Source: OECD Earnings Distribution Database.

Increased dualism also appears to have had an unusually large impact on job stability in Korea. Panel A of Figure 3.5 compares the incidence of temporary employment in OECD countries. According to the internationally harmonised definition used by the OECD, temporary work accounts for 24% of total employment in Korea in 2011, making it the largest component of non-regular work.¹⁶ This is double the OECD-average incidence of temporary work. Only three OECD countries have higher incidences of temporary work than Korea: Chile, Spain and Poland. Temporary employment has declined modestly in Korea since reaching a peak of 27% in 2004.

One of the concerns about temporary employment is that it may expose workers to higher levels of employment insecurity. This could be an important concern in Korea because the worker turnover rate implied by the high incidence of low job tenure is the highest observed in the OECD area (Panel B of Figure 3.5). Worker turnover can play a useful role in reallocating workers from declining firms and sectors to more dynamic ones, but much of the very high level of turnover in Korea appears to reflect “churning” of workers through on-going jobs.¹⁷ These statistics suggest that the increase in temporary employment since the 1997 economic crisis has provided employers with considerable flexibility to adjust the size and composition of their workforces, but also that the associated adjustment costs may be excessively high and unfairly concentrated on non-regular workers. Another concern is that very high turnover rates in Korea reduce employers’ incentives to train their workers. As is shown in Table 3.6 (above), most continuing vocational training is funded by employers and this type of training is less often provided to non-regular workers than to regular workers.

Figure 3.5. Two indicators of numerical flexibility in the Korean labour market

• a) Following the internationally harmonised definition of the OECD, temporary employees are wage and salary workers whose job has a pre-determined termination date as opposed to permanent employees whose job is of unlimited duration. National definitions broadly conform to this generic definition, but may vary depending on national circumstances. Country-specific details are available at www.oecd.org/dataoecd/13/57/43103377.pdf. For Korea, temporary employment encompasses workers with a fixed-term contract, temporary agency workers and daily/on-call workers (excluding double-counting). This definition of temporary workers differs from the national definition used in Table 3.1.
• b) Data refer to 2006 for Australia, 2004 for Mexico and 2005 for the United States
• c) Weighted average of OECD countries whose data are available in 2011.
• d) Hirings in year t are estimated by the number of workers in year t with less than one year of tenure. Separations are then estimated as the difference between hirings and net employment change.
• e) Weighted average of OECD countries shown in Panel B.

Source: OECD Online Employment Database available at www.oecd.org/employment/database for temporary employment, and OECD calculations using the OECD Job Tenure Database for worker turnover.

The high turnover rates observed in the Korean labour market have a positive side for employers, who have considerable numerical flexibility in managing their workforce, but also for workers. The high turnover rates in Korea reflect, in part, the speed with which unemployed workers are able to find and move into new jobs. The efficiency of job matching is indicated by both the low unemployment rate in Korea and also by the low share of unemployed persons in Korea who become long-term unemployed. In 2011, the unemployment rate in Korea, at 3.4%, was less than half the 8% average rate for the OECD area. Among all unemployed Koreans, 7% had been unemployed for six months and over and under 1% for a year or more, while the corresponding averages for the OECD area were 48% and 34%. High turnover rates in Korea are reflected in a relatively low average job tenure (5.1 years in 2011, while job tenure in the OECD countries in the EU range from 8.0 years in Denmark to 12.0 years in Italy). Despite this fluidity, 18% of all employees in Korea have held their current job for ten years or longer.

Internationally harmonised statistics are also available for part-time employment, defined in terms of weekly working hours being below 30. According to this definition, part-time employment accounts for 14% of employment (OECD, 2012b), somewhat smaller than the share of temporary workers.¹⁸ Compared with full-time workers, part-time workers in Korea face a larger wage penalty than in most other OECD countries and also report considerably lower perceptions of job security (OECD, 2010a). This difference helps to explain why a large majority of part-time workers say they prefer a part-time job in most European and North American countries, in contrast with part-time workers in Korea (Table 3.4).

A final insight from international comparisons is that non-regular workers in Korea appear to be more at-risk of becoming trapped in poor quality jobs than their counterparts in other OECD countries. Panel A of Table 3.7 shows that temporary workers in Korea - defined according to the OECD harmonised definition - are less likely to move into permanent employment than their counterparts in 15 other OECD countries. This is true over both a one-year and a three-year horizon. This form of mobility is also quite low in Japan, whereas the “stepping-stone” pattern is much more prevalent in Europe. One likely reason for this contrast is that the majority of temporary workers in Europe are young and temporary jobs often provide a bridge between study and stable career jobs.¹⁹

Mobility patterns in Korea and Japan can be compared for a more comprehensive definition of non-regular work that encompasses e.g. part-time work and some additional forms of atypical work. Table 3.7, Panel B shows that the mobility rate of non-regular workers into regular employment is actually higher in Korea than in Japan, but how much higher is difficult to judge because the estimates vary a lot depending on how non-regular employment is identified in the KLIPS. When workers’ self-assessment of employment type is used, the overall mobility rate for all non-regular workers is similar to that for only temporary workers in Korea (as calculated in Panel A using the internationally harmonised definition), whereas it is sharply lower in Japan where the one-year transition probability falls from 18% to 8%. This appears to reflect the large number of Japanese women who work part-time and have a very low propensity to move into regular employment. Using instead the official definition of non-regular employment (as best it can be implemented in the KLIPS) results in much higher estimated mobility rates for Korea.²⁰

Table 3.7. International comparison of one-year and three-year mobility of temporary

and non-regular workers⁰

Percentages

• a) Self-employed persons are excluded from the samples used to calculate transition probabilities.
• b) Following the internationally harmonised definition of the OECD, temporary workers include workers with fixed-term contract and temporary agency workers. In Korea, this category also includes daily/on-call workers. This definition of temporary workers differs from the national definition used in T able 3.1.
• c) Consistent with national practice in Japan and Korea, non-regular work includes some types of non-standard jobs that need not be temporary, most notably part-time employment.
• d) Non-regular workers identified using the closest possible approximation of the official definition in the KLIPS dataset (as advised by the MOEL). Average values for waves 5-12 (2002-2009).
• e) Non-regular workers identified using the workers’ own assessment of their employment relationship as regular or non-regular. Average values for waves 2-12 (1999-2009).

Source: European countries: OECD calculations using the European Community Household Panel, waves 5-8 (OECD Employment

Outlook 2006). Japan: Calculations based on the Keio Household Panel Survey as reported in Shikata (2012), “Is Temporary Work

‘Dead End’ in Japan? Labour Market Regulation and Transition to Regular Employment”, Japan Labor Review, Vol. 9, No. 3.

Korea: OECD calculations using microdata from the KLIPS, waves 2-12 (1999-2009).