55082 World Bank Employment Policy Primer September 2009 No. 14 ADDRESSING THE EMPLOYMENT EFFECTS OF THE FINANCIAL CRISIS The Role of Wage Subsidies and Reduced Work Schedules As countries look forward to the resumption of growth financial crisis. And any results would also be difficult following the economic crisis there are concerns about the to generalize, since much depends on local conditions response of labor markets. Indeed, the human impact of and the structure of the labor market. Still, based on the recovery in terms of resumed poverty reduction and the past performance of these programs and what is increased well-being will be determined by the respon- known about the adjustment of labor markets during siveness of jobs--both their quantity and quality--to a downturn, five preliminary overarching implications increased economic activity. can be derived (this note later offers more details on The response of employment to growth will depend, the implications as they apply to each of the two types in part, on the efficacy of labor market policies, including of policy). those adopted in industrialized countries, to stabilize/ First, both wage subsidies and reduced work sched- promote employment and provide assistance to those ules can have an important role in helping employment who lose their jobs. In the first category the most com- recover, but their success will depend on the context. Both mon interventions include: (i) various forms of wage programs are likely to be more effective in countries subsidies; (ii) reduced work schedules; and (iii) access and/or sectors where the contraction in labor demand to credit. The second category involves mainly: (i) the is temporary, reflecting the economic cycle. In such extension of unemployment benefits, (ii) employment cases, wage subsidies and reduced work schedules can services (including training); and (iii) cash-transfers mitigate the effect of the economic slowdown on for- for low income workers and the poor often through mal employment and reduce the depreciation of skills public works. usually associated with long term unemployment or This note briefly reviews the experiences with wage informal work. However, in countries where structural subsidies and reduced work schedules in promoting changes in the economy are needed, these two programs employment and avoiding the depreciation of accumu- could delay necessary labor reallocations and thus com- lated skills and knowledge due to a temporary downturn. promise sustainable growth and employment creation These policies have been adopted by many high income over the medium term.1 countries as well as some middle income countries. It is to early o comment on their impact; to date, they 1 have not been rigorously evaluated in the context of the See, for example, Revenga and Rijkers (2009). Note prepared by David Robalino and Arup Banerji. It builds on previous notes on the financial crisis and labor markets prepared jointly by PRMPR and HDNSP, as well as inputs and comments provided by Ana Revenga, Ariel Fiszbein, Pierella Paci, Arvo Kudo and David Newhouse. Addressing the Employment Effects of the Financial Crisis: T h e R o l e o f Wa g e S u b s i d i e s a n d R e d u c e d Wo r k S c h e d u l e s Second, when adopting wage subsides and reduced Table 1: Wage Subsidy and Work-Sharing work-schedules, it is important to assess the tradeoffs with Programs in OECD Countries other active labor market programs that benefit the most vulnerable workers. Wage subsidies and reduced work Job subsidies, Reductions schedules mainly benefit formal sector workers, which recruitment in nonwage Short-time represent less than 50 of the labor force in most middle incentives labour costs work schemes and low income countries. Informal sector workers, agri- Australia X cultural workers in rural areas, and the self-employed-- Autria X none of whom have access to social insurance and also Belgium X X face higher earnings risks during a down-turn--are Canada X X X unlikely to benefit. Other programs such as public works CzechRepublic X X could serve some of these groups better, but fewer public Denmark X resources would be available to implement them. Finland X X Third, the design of the programs matters: France X X X Germany X X Employer-based subsidies can cost less and have a Greece X larger social impact if targeted to individuals with Hungary X X X lower pay. Their main role is to smooth the impact Ireland X of a temporary downturn. Italy X Targeted employee-based subsidies linked to other Japan X X X active labor market programs such as training and job counseling can be considered to improve the Korea X X employability of low-income unskilled workers over Luxemburg X X the medium-term. Mexico X X X For work-sharing arrangements, much depends on Netherlands X the negotiations between employers and employees. New Zeland X X Two important elements are the design of training Norway X programs and the setup of the wage compensation Poland X X X mechanism. Portugal X X Slovak Republic X X X Wage Subsidies Spain X X Sweden X X Wage subsidies come in different forms, pursuing Switzerland X different but related objectives. Subsidies can be given Turkey X X to employers to stimulate demand or to employees to United Kingdom X provide incentives for re-employment. They can take United States X the form of direct transfers, reductions in social secu- Source: OECD (2009). rity contributions, or income tax credits. They can be targeted to vulnerable workers or be provided across the board. And they can focus on those already employed or unemployed or move to lower productivity jobs in the only on new hires. Choices at all these levels determine informal sector--in both cases risking losing their skills. their impact on employment and their costs.2 In the second case, the focus is to increase exit rates from In the context of the crisis, wage subsidies have two the unemployment pool, particularly among vulnerable main functions: (i) dampening the pace of job destruction workers, thus reducing the destructive persistence of in the formal sector and avoiding the depreciation of skills; long term unemployment. and (ii) stimulating hiring and facilitating reemploy- ment. In the first case, the subsidies allow employers to keep workers who would otherwise fill the ranks of the 2 See Jae-Kap Lee (2005) for a typology. 2 Addressing the Employment Effects of the Financial Crisis: T h e R o l e o f Wa g e S u b s i d i e s a n d R e d u c e d Wo r k S c h e d u l e s Between 2008 and 2009, the majority of OECD coun- hired (or would not have been fired) in any circum- tries have adopted wage subsidies--mainly through reduc- stance, or induce substitutions between subsidized tions in social security contributions, and often targeted and non-subsidized workers.3 Others, however, at small enterprises or disadvantaged groups. France, for show that at the macro level net effects are usually instance, has reduced employer social security contribu- positive.4 Moreover, there is evidence of positive tions for firms with less than 10 employees hiring new (dynamic) effects on the ability of beneficiaries to low-wage workers in 2009. The reduction is the largest improve their labor market opportunities over the for workers hired at the minimum wage and declines as medium-term.5 the wage increases up to 1.6 times the minimum wage. During the crisis, reductions in social security contri- In Germany, there has been a reduction of the employee butions are likely to achieve their objective of main- and employer contributions to the unemployment insur- taining employment and preserving human capital. ance system. Spain has reduced employer social contri- Indeed, most evaluations agree that a lower wedge butions for the first two years of employment for unem- between the total cost of labor and take-home pay ployed people with children who transit to full-time can both reduce dismissals and stimulate hiring.6 permanent contracts. It has also implemented a reduc- The costs of effective wage subsidies, however, should tion in social security contributions for youth or disabled not be underestimated.7 Meaningful effects may workers who start up a business as self-employed. A require sizable reductions in the contribution rate different approach has been taken in United Kingdom, (e.g., a 4 percentage point reduction in the contribu- where companies will receive £2,500 for hiring workers tion rate to avoid a 2 percentage point reduction in who have been unemployed for more than six months. employment), which implies substantial budgetary Several middle income countries have also adopted transfers or an increase in the unfunded liabilities of wage subsidies. In Mexico, there has been a tempo- the social security system. rary reduction in employer social contributions and a In terms of design, subsidies targeted to vulnerable deferred payment of up to 50% of contributions. The workers (and linked to the stock of employees) are Slovak Republic has introduced a temporary exemption likely to be a better option than across the board sub- from health insurance payments for up to 24 months sidies.8 The later are not only more expensive but for persons entering self employment after a period of also subject to larger "dead-weight" losses, as they 3 months on the jobseekers register. Other countries, end-up subsidizing workers who would have stayed however, are not using the subsidies as a temporary in their jobs or would have been hired anyway. The measure. Turkey has legislated a permanent reduction focus could be instead on lower paid workers (usu- in employer contributions to disability, old-age and sur- ally low skilled and young workers), probably with a vivor pensions, as well as reimbursement of employer declining subsidy as wages increase (as in the exam- social security contributions for firms with 50 or more ple of France above). It is also important to link the workers and that have at least 3% of disabled workers in subsidies to changes in the stock of employment. their staff. It has also introduced a permanent reduction Wage subsidies that are simply linked to the wage in employer social contributions for the first five years could give firms more room for "gaming the system". of employment targeted to new women and youth (aged 18­29). Chile, on the other hand, has introduced an employment subsidy for hiring of young workers (age 3 See Tzanattos (1998) and Becherman et al. (2004) 18 to 24 years). Similarly, in Korea the government is 4 See Bell et al. (1999) and Richardson (1997) for a discussion. 5 See Gerfin and Lechner (2002) and Codax and Gobel (2005). providing wage subsidies to SMEs for new hires, interns, 6 See Chapter 7 in Robalino et al., (2009) and OECD (2003) for a as well as those that convert irregular into regular jobs. review. Based on past evaluations of wage subsidies, mainly 7 Studies suggest that, overall, a 10 percentage point reduction in in OECD countries, a few lessons are worth noting: the tax-wedge (the difference between the cost of labor and take- home pay) could increase employment between 1 and 5 percentage points (see Nikell, 2003; Kugler and Kugler, 2003; Rutkowski , 2007; There is mixed evidence over the effects of wage sub- and Betcherman and Pages,2007). sidies on employment. Some studies emphasize that 8 See Phelps (1994) for a discussion of the rational and OECD (2003) subsidies often go to workers who would have been for a discussion of the evidence. 3 Addressing the Employment Effects of the Financial Crisis: T h e R o l e o f Wa g e S u b s i d i e s a n d R e d u c e d Wo r k S c h e d u l e s In all cases, it is important to consider the tradeoff per cent of social security contributions (and 100 per cent between providing assistance to formal sector workers if reduced working hours are used for training). Similarly, and informal sector workers. Employer-based wage in Japan, the Employment Adjustment Subsidy is pay- subsidies benefit mainly the former and thus reduce able directly to employers and has been increased from available resource to protect the later. They can thus one-third to two-thirds of the of the employee's wage loss be regressive. in the case of large enterprises, and from two-thirds to Alternative options to consider are targeted employee- nine-tenths in the case of SMEs (there is a maximum of based subsidies linked to others active labor market $USD 80 per day). The subsidy is financed from employer programs such as training.9 The focus of these pro- contribution to employment insurance. grams would be on helping find jobs to the already Middle income countries that have implemented this unemployed--particularly vulnerable groups such type of program include Argentina, Colombia, Mexico, as youth and/or unskilled workers. This is the case Poland and Turkey. Argentina, quickly reactivated the of the Jovenes programs that have been imple- "Progama de Recuperacion Activa" that was designed mented successfully in several Latin American during the 2002 crisis: a monthly wage supplement countries (e.g., Argentina, Colombia, Dominican is provided to employees for up to 12 months with Republic).10 These programs combine on the job- an agreement not to dismiss workers and adjust work training, employment assistance, and training sti- schedules. Colombia and Mexico are using subsidized pends--which can be assimilated to wage subsidies. training to maintain workers within the formal sector, while enhancing their employability. In Poland, a work Reduced Work Schedules sharing program supported by workers and employers organizations was added in July 2009 financed from the Under this type of programs, workers are required to Guaranteed Employee Benefits Fund. It also establishes reduce working hours and accept a proportional reduc- training support with up to 90 per cent of the costs to tion in wages. The programs are thus accompanied by be paid from the Unemployment Benefit Fund (the rest interventions that involve access to partial unemploy- is financed by employer training funds). Similarly, in ment benefits, provision of paid holidays, or provision Turkey, the government has extended the length of time of short-term training stipends.11 that workers can benefit from the reduced hours fund Like wage subsidies, reduced work schedules can contain from three to six months and increased the benefit level unemployment and prevent skills depreciation, loss of firm- by 50 per cent. specific human capital, and informality. The arrangements As in the case of wage subsidies, there have been no effectively spread the burden of labor market adjustment impact evaluations of these programs in the context of across workers. Although earnings might decline, workers the crisis, but from past experiences a few lessons can remain gainfully employed and preserve access to social be derived: insurance benefits. The programs are also expected to bring benefits to employers in two ways.12 First, firms Overall, the effect of mandatory work sharing on are able to retain skilled workers (including those with employment is still under debate. There is evidence acquired firm-specific skills), thus avoiding the substan- in the case of France, for instance, that aggregate tial costs of recruiting and training new workers when employment was unaffected but that labor turnover demand recovers. Second, enhanced employee morale increased as firms shed workers who became more (vis-à-vis layoffs) can lead later to lower turnover costs. expensive.13 A review of programs in European Most OECD countries have introduced these programs as a result of the crisis. One of the largest is the Kurzarbeit 9 For evidence on the positive impact of this combined programs in Germany, which is expected to cost 5 billion between see Katz (1996) and Cockx et al. (1998). 2009 and 2010. The program just extended the maximum 10 See Robalino and Sanchez-Puerta (2009) for a review. 11 See Abraham and Houseman (1993). benefit period from 6 to 24 months for companies who 12 See Messenger (2009). sign on in 2009. The Federal Employment Agency covers 13 See Stevao and Sa 2008. The authors also show based on survey up to 67 per cent of wage losses for workers whose hours responses that the effects varies by gender, with women benefiting have been reduced, while also reimbursing employers 50 more coordination to a shorter work week 4 Addressing the Employment Effects of the Financial Crisis: T h e R o l e o f Wa g e S u b s i d i e s a n d R e d u c e d Wo r k S c h e d u l e s countries suggests that the results are mixed, with The efficacy of both programs, however, will depend impacts depend on the setting.14 on the economic context. The programs are likely During the crisis, much will depend on the types to be more effective in countries and/or sectors of arrangement negotiated between employers and where the contraction in labor demand is temporary employees. Thus, social dialogue, collective bargain- (cyclical). In countries where structural changes in ing and other forms of negotiations play an essential the economy are needed, these programs can delay role in determining the conditions for the imple- necessary labor reallocations. mentation of work sharing programs.15 When designing wage subsides and reduced work- Two important issues in terms of design involve train- schedules it is important to assess costs and the trad- ing and compensation mechanisms for lower wages. eoffs with other active labor market programs that A practice in industrialized countries worth con- benefit the most vulnerable workers. Wage subsidies sidering is encouraging employees who are on and reduced work schedules mainly benefit formal work-sharing schemes to participate in training sector workers, which represent less than 50 percent and retraining activities to upgrade their skills and of the labor force in most middle and low income improve their employability. In terms of the com- countries. pensation mechanisms, the unemployment insur- Finally, the design of the programs is important. ance system is usually used to top-up wages. When Employer-based wage subsidies can cost less and not available (the case of most middle and low have a larger social impact if they are targeted to income countries) reduced work schedules could be individuals with lower pay. Governments can also implemented linked to pension systems. In this case, consider employee-based subsidies targeted to vul- compensation for lower wages can be financed by nerable groups and linked to other active labor mar- borrowing from individuals' pension wealth.16 ket programs. The impact of work-sharing arrange- In all cases, the administrative costs for employers and ments, on the other hand, will depend on the the demands in terms of institutional capacity to man- negotiation process and the type of compensation age wage compensation should not be underestimated. mechanism. Ideally these programs should be linked Employers need to introduce substantial changes in to unemployment benefit schemes. human resources management practices. There are also costs in the short term due to changes in sched- References uling and work organization. From the government side, implementation can be difficult when the social Abraham, K. and S. Houseman, 1993. "Does Employ- security administration does not have well-devel- ment Protection Inhibit Labor Market Flexibili- oped contribution collection and record keeping ty? Lessons from Germany, France, and Belgium", systems that provide information about earnings of NBER working paper W4390. various workers in a firm. Betcherman Gordon and Carmen Pages. 2007. Estimat- Finally, as in the case of wage subsidies, attention ing the Impact of Labor Taxes on Employment needs to be paid to costs. Unless there is a well fund- and the Balances of the Social Insurance Funds in ed unemployment system, reduced work schedules Turkey. Synthesis Report. August 2007. World Bank. involve government subsidies that can be costly Betcherman Gordon, Karina Olivas, and Amit Dar. 2004. and regressive--since transfers are likely to benefit "Impact of Active Labor Market Programs: New Evi- mainly formal sector workers. dence from Evaluations with Particular Attention to Developing and Transition Countries." Social Pro- Conclusions tection Discussion Paper Series 0402. World Bank. Bell, B., R. Blundell and J. Van Reenen. 1999. "Get- Wage subsidies and reduced work schedules show ting the Unemployed Back to Work: The Role of some promise as measures that can help countries to increase the employment elasticity of growth dur- 14 Kapteyn et al. (2000). ing the recovery and avoid the depreciation of skills 15 See Messenger (2009). 16 See Robalino et al. (2009). associated with unemployment or informal work. 5 Addressing the Employment Effects of the Financial Crisis: T h e R o l e o f Wa g e S u b s i d i e s a n d R e d u c e d Wo r k S c h e d u l e s Targeted Wage Subsidies," The Institution for Fiscal Nickell Stephen 2003. "Unemployment and Labor Mar- Studies, Working Paper Series, No. W99/12. ket Rigidities: Europe versus America" Journal of Cocks, B. and C. Gobel . 2005. "Subsidized Employment Economic Perspectives. 11 (3): 55­74. for Young Long-Term Unemployed Workers: An OECD. 2003. "Employment Outlook 2003." OECD. Evaluation." Manuscript. Paris. Cocks, B., Van der Lindern, A. Karaa. 1998. "Active Labor Phelps, E . 1994. "Low-Wage Employment Subsidies Vs. Market Policies and Job Tenure," Oxford Economic the Welfare State." The American Economic Review, Papers 50: pp.685­708. Vol. 84, No. 2, Papers and Proceedings of the Hun- Estevão M. and Filipa Sá. 2008. "The 35 Week." Eco- dred and Sixth Annual Meeting of the American nomic Policy July. 417­463. Economic Association. 54­58. Gerfin M., M. Lechner and H. Steiger. 2002. "Does Revenga, Ana and Bob Rijkers (2009). "Policy Options Subsidized Temporary Employment Get the Unem- to Cope with the Labor Market Impacts of the Cur- ployed Back to Work?" An Econometric Analysis of rent Crisis." Mimeo., World Bank. Washington DC. two Different Schemes." Swiss Institute for Interna- Richardson, J. 1997. "Wage Subsidies for the Long-term tional Economics and Applied Economic Research Unemployed: A Search Theoretical Analysis," Centre (SIAW) University of St. Gallen. for Economic Performance Discussion Paper No. 347. Jae-Kap Lee. 2005. "Evaluation and Lessons from Wage Robalino David and Laura Sanchez-Puertas. (2009). Man- Subsidy Programs in OECD." Directorate of Employ- aging Labor Market Risks and Creating Better Jobs: ment, Labour and Social Affairs. OECD. Paris. Alternative designs for Income Protection and Active Holzmann Robert, K Iyer, and M. Vodopivec (forth- Labor Market Policies. Draft Manuscript. World Bank. coming). Severance Pay Programs Around the World: Washington DC. Rationale, Status and Reforms. In R. Holzmann and Robalino David, Milan Vodopivec and Andras Bodor. M. Vodopivec, eds, "Improving Severance Pay: An (2009). "Savings for Good and Bad Times: Intro- International Perspective." World Bank. Washing- ducing Unemployment Insurance to Developing ton DC. Countries. Social Protection Discussion Papers. Kapteyn Arie, Adriaan S. Kalwij, Asghar Zaidi. 2000. "The World Bank. Washington DC. Myth of Worksharing." Discussion Paper Series. IZA Rutkowski, Jan. 2007. "Labor Taxes and Employment in DP No. 188. Institute for the Study of Labor (IZA). ECA." Unpublished Manuscript. World Bank. Wash- Katz, L. 1996. "Wage Subsidies for the Disadvantage." ington DC.02/01/2009 17:56:00 NBER. Working Paper 5679, National Bureau of Economic Research. Kugler, A. & Kugler, M., 2003. The Labor Market Effects of Payroll Taxes in a Middle-Income Country: Evidence from Colombia. C.E.P.R. Discussion Papers, 4046. Messenger Jon. 2009. "Work Sharing: A Strategy to Pre- serve Jobs During the Global Jobs Crisis." Travail Policy Brief No. 1. International Labor Organiza- tion. Geneva. The findings, interpretations, and conclusions expressed herein are those of the author(s), and do not necessarily reflect the views of the International Bank for Reconstruction and Development / The World Bank and its affiliated organizations, or those of the Executive Directors of The World Bank or the governments they represent. 6