Coping with the job crisis and preparing for ageing.
After 15 years of almost continuous increases in employment, the labour market started to worsen in 2008 as the global recession gained momentum (Figure 3.1). While unemployment has risen significantly, its increase has been surprisingly small so far. This is mainly due to a surge in temporary layoffs, which by the end of 2009 covered approximately 3% of the labour force, and rapidly falling labour supply (see Chapter 1). With few signs of a significant turnaround in the economy, the labour market is projected to continue to worsen through 2010, with temporary layoffs gradually spilling over into unemployment, inactivity and labour market exit. A major challenge for the Finnish economy is to prevent a repetition of the results of the deep recession in the early 1990s, specifically the rise in long-term unemployment (Verho, 2008 and Chapter 1). These scarring effects still affect labour market outcomes: for example, employment rates for cohorts born earlier than 1960-64 (who were 30-35 years old at that time) have never recovered their pre-recession employment levels.
[FIGURE 3.1 OMITTED]
A number of similarities in the current labour market situation with the past recession may complicate recovery. Wage increases coming into the recession have again been substantial and real wages remain among the least flexible in the OECD. This time, real wage adjustment will have to take place under a fixed exchange rate, putting more burden on nominal wage adjustment. Another similarity is the high share of the working-age population that depends on income-replacement benefits. The dependency ratio of 26% in Finland was well above the OECD average of 19% in 2004 (Duell et al., 2009). In addition, replacement rates in the unemployment benefit system remain high, although in line with the other Nordic countries. Low activation rates due to late intervention on unemployment and a highly decentralised system of Employment Offices still characterise the Public Employment Service (PES). Retirement ages remain well below the OECD average, and lag Nordic neighbours even more.
This chapter discusses the challenges faced by labour market policies in preventing a repeat of the adverse employment outcomes of the previous recession and dealing with the challenges of an ageing workforce. It first discusses key characteristics and developments in Finnish labour markets before and during the crisis. It then looks at the impact of wage formation, labour market policies, unemployment benefits and social benefits. Finally, it analyses the effects of various retirement regimes on incentives to work and labour supply, and how these should be reformed to improve employment outcomes.
The recovery may be slowed by rigid labour markets
Labour market outcomes remain weaker than in the other Nordic countries
Although the labour market has recovered since the 1990s recession, employment and participation remain low (Figure 3.2). While overall employment and participation rates have risen to levels well above the OECD average, they remain below those prevailing before the 1990s recession and in the Nordic neighbours. Finland's relative underperformance in terms of employment rates vis-a-vis other Nordic countries is mostly due to low employment rates among younger individuals, women of child-rearing age and, especially, older age groups (Figure 3.2).
Overall labour utilisation in Finland, along with the other Nordic countries, is close to the OECD average (Figure 3.3). Compared to Denmark, Norway and Sweden, lower employment rates in Finland are compensated for by higher average hours worked. This reflects the low prevalence of part-time work in Finland, which to some extent is linked to low employment among groups that typically work part time (young, old and women). As discussed in the previous Survey (OECD, 2008a), the relatively low share of the service sector in the economy, inflexibility of childcare arrangements and the tax and social security system (including pensions) are likely explanations to the low prevalence of part-time work. Employment among women with very young children is also constrained by the relatively generous Home Child Care allowance and possibly the child supplement in the unemployment insurance.
[FIGURE 3.2 OMITTED]
Aligning wages better with economic conditions would help employment
Wages in Finland are among the most rigid in the OECD, which can hamper the recovery and contribute to further unemployment (Holden and Wulfsberg, 2007). The high rigidity partly reflects the tradition since 1970 of centralised wage negotiations between unions, employers' federations and the government. The settlements also often included agreements on income policies (Johansson, 2006). While centralised or co-ordinated wage bargaining has been shown to produce lower aggregate wage outcomes, decentralisation would allow real wage flexibility and promote employment among weaker groups in the labour market (OECD, 2006a).
Finland recently moved to a more decentralised wage bargaining system to improve relative wage flexibility. In the 2007/08 round, wage agreements were made at the industry level with very little government intervention, but as discussed in the previous Survey, the overall outcome was highly unsatisfactory, with large wage increases beyond those justified by economic conditions and no substantial improvements in local wage flexibility (Figure 3.4). While high wage increases have supported domestic demand, thereby helping to cushion the current slump, they are spilling over into higher prices and lost competitiveness with potentially adverse effects on employment (Chapter I). Local wage allowances rose as a share of the settlements, but as these were used in a fairly mechanical way they contributed little to aligning wages to local productivity conditions (Asplund, 2007).
[FIGURE 3.3 OMITTED]
The ongoing round is taking place in a similar institutional setting, but at the depth of a severe recession and against a background of rising unemployment and severe constraints on public finances. In such conditions, a properly functioning decentralised bargaining framework needs to deliver moderate wage increases. Yet, although the final results are still not known, there are risks that the moderate increases already negotiated in some export-related areas will not carry over to the government sector. This could in turn feed into additional compensatory demands from other sectors. In addition, there is little evidence that the settlements so far have increased relative wage flexibility.
The onus is therefore on the social partners to negotiate wages consistent with the weak cyclical situation and international competitiveness. If they fail, the authorities should consider alternatives to the current system. One option would be to increase coordination among and between employers and unions and agree on a reasonable aggregate wage outcome. This would foster more moderate overall settlements. Greater coordination would however also tend to hamper the intended movement towards greater local wage flexibility. This outcome might be avoided if the aggregate agreement did not impose further restrictions on industry level negotiations, and if bargaining at the industry or firm level took the opportunity to match wages better to their specific economic situations. This type of centrally co-ordinated and yet locally flexible wage negotiation framework has been successful in moderating wage increases in Sweden since 1997. (1)
[FIGURE 3.4 OMITTED]
Other measures to raise local wage flexibility should also be considered. One option is to reform the unemployment insurance system, which may have contributed to both high overall wage outcomes and reduced flexibility. As employers and the government are the main contributors to the unemployment insurance scheme (see Table 3.1), this may have given rise to insider-outsider mechanisms, where employed union members give less consideration to the social costs associated with too high wage outcomes. Therefore, the government and the social partners should consider setting up mechanisms that strengthen incentives to achieve low unemployment, for example through experience ratings. Furthermore, as is recommended below, lower replacement rates in the unemployment insurance and tightened access to early retirement could reduce reservation wages and hence constrain excessive wage increases.
As in other highly unionised countries, such as Belgium and Sweden, Finland has a compressed wage structure. (2) Centralised bargaining contributes to wage compression which is likely to lower employment levels among marginal groups in the Finnish labour market. While there is no robust evidence that high minimum wages contribute to unemployment in general, some studies find negative employment effects for young workers (OECD, 1998). Low-skilled young individuals may be especially at risk from high minimum wages, which may explain part of the relatively low youth employment in Finland (Figure 3.B). Lower minimum wages for young unskilled workers could enhance their employment opportunities.
[FIGURE 3.5 OMITTED]
Inactivity benefits are generous and with long duration
Replacement rates are among the highest in the OECD area for the long-term unemployed, reducing incentives to enter or return to work (Figure 3.6 and Box 3.1). The high replacement rates reflect generous and extended allowance periods and the interplay with special rules, such as the "unemployment pipeline" and other social security schemes. While high replacement rates can to some extent be compensated for by aggressive activation, as in Denmark, the Finnish combination of generous long-term replacement rates in combination with late activation is unusual among OECD countries.
The mix of high replacement rates and late interventions raises several concerns. First, there is strong evidence that high replacement rates hamper search intensity and in general tend to increase unemployment (Bassanini and Duval, 2006). The introduction of the increased Earnings-related Allowance (Box 3.1) has also been shown to decrease chances of re-employment (Uusitalo and Verho, 2007). Furthermore, there are strong arguments in favour of letting replacement rates fall over the unemployment spell to balance the individual's need for insurance versus the incentives to keep up search intensity (Fredriksson and Holmlund, 2006).
In order to maintain work incentives, the government should not only lower replacement rates over the full 500-day unemployment period, but also maintain or even increase the decline in the replacement rate with the length of the unemployment spell. (3) Although cuts in replacement rates should not be implemented during the depth of the recession, clearly signalling that such changes will take place when the economy recovers, would contribute to higher search intensity. (4) Additionally, the government needs to ensure that the Social Protection Reform Committees (SATA), proposed reforms to the housing benefit efficiently addresses the inactivity trap created by the Basic Income Support (BIS) combined with housing benefits (Box 3.1). The government should also consider ways to equalise replacement rates and activation for the non-means-tested Labour Market Support (LMS) and the BIS for those that have work capacity. (5) When implementing welfare reforms such as reforming the housing benefit system, the government needs to ensure sufficient targeted support to the poor (e.g. to support poor retirees and poor families with children). (6)
[FIGURE 3.6 OMITTED]
Box 3.1. Replacement rates from unemployment benefits and social benefits are high Unemployment benefits in Finland come from two alternative schemes; the Earningsrelated Allowance financed by voluntary unemployment funds, or a Basic Allowance for those that are not entitled to support from the earnings-related scheme. Membership rates in the unemployment funds are high, and thus for the majority of unemployed the earnings-related scheme remains most relevant. Both the Earnings-related and the Basic Allowances are available for 500 days (100 weeks). After 500 days, 180 additional days of Labour Market Support (LMS) without means-testing is available. After 680 days the unemployed have access to means-tested LMS or Basic Income Support (BIS, social assistance from the municipality) or a combination of these. The Earnings-related Allowance generates fairly high replacement rates compared to other OECD countries, although they are capped at 90% of the previous wage. For a wage earner on 67% of the average wage, net replacement rates in Finland average 79%, in line with the other Nordic countries (average of 81% in Denmark, Norway and Sweden) and well above the 70% average in the OECD area (Duell et al., 2009). Furthermore, there is a child supplement on the Earnings-related Allowance of a maximum of EUR 9.19 per day for three children. For employees made redundant and who are willing to participate in activation, replacement rates are further topped up the first 150 days through the increased Earnings-related Allowance, which raises replacement rates by an additional 6 percentage points. A further supplement for laid-off workers who participate in active labour market programmes was introduced on 1st of July 2009. The Basic Allowance is much less generous and provides only EUR 23.91 per day. Replacement rates are typically much lower once the unemployed individual moves into the LMS, but the beneficiary can also claim means-tested housing benefits and in some cases BIS, making the fall in replacement rates less steep (Figure 3.7). After 680 days or by refusing activation, the individual may move into means-tested BIS. While the BIS only replaces a fraction of the original wage, it covers 100% of housing expenditures up to a threshold. This yields a fairly high net replacement rate, sometimes significantly higher than the LMS. BIS also covers additional supplementary expenditures related to housing (electricity, insurance, etc.) and health care. For households and individuals who expect wages close to the minimum levels and who have high housing costs, replacement rates can be close to 100%. [FIGURE 3.7 OMITTED]
Activation policy should be more ambitious
Weak activation policies can delay recovery in employment and allow long-term unemployment to take root. Although activation policies have only a limited impact on employment during a recession, they can help maintain search capacity and employability until the labour market starts to improve (OECD, 2009a). In 2007, the costs of labour market policies in Finland amounted to 2.3% of GDP. This is roughly twice the OECD average, but comparable to Denmark, Norway and Sweden. However, the proportion of resources spent on active measures was only 38%, which is well below that in the three other Nordic countries (56%). Furthermore, increases in spending on the Public Employment Service (PES) and Active Labour Market Policies (ALMPs) during the recession have trailed other OECD countries (Figure 3.8, first panel).
Activation tends to take place late in the unemployment spell compared to other Nordics. Referral to an active labour market programme takes place after 100 weeks compared to a mandatory referral after 300 days (i.e. 60 weeks) in Sweden and 9 months (i.e., 40 weeks) in Denmark. The late activation has shown up in a fall in the number of participants in ALMPs since the recession started in 2008 (Figure 3.8, second panel). Finland's late activation also contributes to the weak performance in terms of benefit dependency, as illustrated by the exceptional deviation from the regression line in Figure 3.6. Some steps towards earlier activation have been taken recently by removing the upper age limit on the compulsory activation requirement and formulating individualised work plans and guaranteeing work or training for youth if the duration of the unemployment spell goes beyond 3 months.
[FIGURE 3.8 OMITTED]
Mandatory activation measures should also be implemented much earlier than after the current 500 days. As the SATA Committee has recommended, the government should consider mandatory interventions after 50 weeks or earlier, to be implemented as quickly as possible. This reform would require more personnel and resources in the PES. Existing profiling systems (see below) should be used early during the unemployment spell to identify individuals that have a high risk of ending up in long-term unemployment. These individuals should be subject to mandatory interventions as early as possible. Refusal of activation should always lead to sanctions. The government should put forward extra funding on a temporary basis to finance early interventions and ensure that these resources are quickly transformed into activation measures in order to dampen the decline in activation to unemployment ratios.
The overall efficiency of labour market matching has fallen over a long period of time (Figure 3.9). This tendency can to some extent be attributed to growing regional mismatches (see Chapter 4) and insufficient labour mobility (OECD, 2008a). There are also considerable differences in matching efficiency between regions, indicating that the efficiency in provided employment services may differ significantly. Estimates show that average unemployment rates could have been 2.4 percentage points lower if all labour market offices operated at the same efficiency as the best region in terms of matching between unemployed and vacancies (Hynninen et al., 2009). While these estimates suggest considerable potential for increases in efficiency, they do not signal that fiscal resources are the most important part of the explanation. In fact, resources in terms of staff expenditures per unemployed seem to be only weakly correlated with matching efficiency across regions (Figure 3.10).
To address inefficiencies in the PES system, more centralisation is needed. The fragmented nature of the institutional framework (Box 3.2) increases the risk that different local offices apply different procedures and criteria in handling cases. The decentralised setup also increases the risk of conflicts of interests within the Labour Committees. To achieve similarity in interventions and criteria across offices, clear instructions on interventions and sanctions should be available to all desk officers. The central government should make sure that the current benchmarking of labour market offices is used to harmonise intervention procedures and equalise the efficiency in service provision. The current system of performance-related pay should be stepped up. (7) As skill levels in the employment offices are typically quite low and a large share of their workforce is temporary, human resource issues need to be addressed (Duell et al., 2009).
[FIGURE 3.9 OMITTED]
Box 3.2. Many institutions influence the Public Employment Service (PES) The responsibility for employment policies and their financing is split among a large number of actors. Three line ministries (Employment and the Economy (MEE), Social Affairs and Health, and Education) are involved in policy making, together with the 348 municipalities, the Social Insurance Institution (KELA), the 34 unemployment funds, the unions and the employer organisations. In the PES, front line services are supplied by the 74 Employment and Economic Development offices (also known as T&E offices), which are financed by the MEE. There is no central body for the PES system at the national level, but the MEE implements employment policies through 15 regional T&E centres, which in turn manage the T&E offices. Decisions on eligibility and benefits are taken by local Labour Committees associated with local employment offices, where the PES, the municipalities and the social partners are represented. In 2004 the PES was reformed by setting up i) JobSeeking Centres in order to improve information sharing to job seekers, and ii) Labour Force Service Centres (LAFOS) to deal with more difficult to-place unemployed. The latter centres are jointly staffed by municipalities and the PES. In 2009, a further reform of the operations of the T&E offices was initiated, aiming at integrating labour and business services better.
[FIGURE 3.10 OMITTED]
Given that resources in the PES will be strained during the next few years, a wider use of the existing profiling system should be considered. Apart from liberating case officers from some duties, profiling may also be used to identify individuals in need of early activation measures and the use of a mechanical system to allocate individuals to ALMPs may seem fairer than the alternatives. Profiling systems may also be used more extensively by e.g. allocating individuals to different types of ALMPs (Frohlich et al., 2004).
Regional mobility should also be enhanced. As recommended in the previous Survey (OECD, 2008a), the allowance for work-related second residences should be phased out as it is likely to only benefit those that are wealthy enough to maintain two residences. Further tightening legal requirements for geographical mobility in job search would help and the recent extension of the required commuting radius is a move in the right direction. High replacement rates and the high cost of housing in areas with relatively stronger labour markets remain obstacles to mobility however.
Financial responsibilities for benefits are fragmented
Financial responsibilities for unemployment and related benefits are shared among many actors, which may blunt incentives. The central government pays for the infrastructure and activation programmes (Box 3.2). The costs of the LMS, housing subsidies and the BIS are split between the central government and municipalities. Recent evidence from reforms e.g. in the Netherlands in 2004, suggest that the number of beneficiaries could be lowered through replacing the shared cost formula by a block grant system where the local government bears the full responsibility for delivery (Duell et al., 2009).
Although the unemployment funds carry the financial responsibility for earnings-related benefits, their financing is opaque. Typically these funds are associated with unions. Employees' contributions and membership payments make up only 17% of the funds' total revenues, which has contributed to high membership rates. The rest of the costs are covered by government and employers' contributions (Table 3.1). There is evidence from some OECD countries that unemployment funds outside strict government control may be more generous to their members than prevailing rules allow (Duell et al., 2009). While increasing the member and employee contributions significantly could dampen perverse incentives, evidence from Sweden since 2007 suggests raising rates may mean that a large fraction of the insured exit the insurance system. Therefore, the government should instead consider whether a nationalisation of the (already largely government-financed) fund system would not be the best way to deal with these problems.
Keeping people employed longer would help meet costs of ageing
The programme of the present government has identified a rapidly ageing population and early exit from the labour market as the most serious challenges to growth and sustainable public finances going forward (Prime Minister's Office, 2007). The current crisis can magnify these problems if exits from the labour market increase and the effective retirement age falls. Although the pension reforms of 2005 (Box 3.3) helped increase the retirement age through abolishing the unemployment pension and raising the minimum retirement age, more needs to be clone to return to sustainability. As the disability pension, the unemployment "pipeline" and the old-age pension system are interconnected, any reforms should address those systems jointly.
The labour market for older workers has so far held up relatively well through the downturn. Unemployment rates have increased, but not differently from prime workers (Figure 3.2, second panel). Employment rates have declined, but the expected retirement age actually increased in 2009. There is however substantial uncertainty related to the temporary layoffs, where some information suggests that older workers are overrepresented in this category.
Box 3.3. The Finnish old-age pension system The Finnish old-age pension system that was reformed in 2005 has two major components: an income-tested basic pension (national pension), and a number of statutory earnings-related schemes, all being constructed in similar ways. The importance of the national pension has decreased over time and less than one in ten retirees receive only a basic pension, although roughly half of retirees received some part of their retirement income from this system in 2007. The earnings-related system can be characterised as a defined-benefit (DB) system with substantial buffer funds. It is financed by contributions paid by employers and employees. From 2005, pension accrual is 1.5% of pensionable earnings at ages 18-52, 1.9% at ages 53-62 and 4.5% at ages 63-67. Pension rights are also accrued while studying, on parental leave and while unemployed. The retirement age is flexible between the ages 63 and 68, with an early old-age pension option from 62 (with a significant penalty on retirement income) and a possibility to defer old-age pension after 68. However, there is no actuarial adjustment of pension rights during the retirement window, which reduces incentives for further work. For the median male earner, the net replacement rate is 62%, which is lower than the OECD average of 72% (OECD, 2009b). From 2010 onwards, new earnings-related pensions will be reduced in line with increased expected longevity. This adjustment will mean that retirement incomes in relation to wages will fall, unless future retirees respond by staying longer in the labour market. The earnings-related schemes are handled by insurance companies, company pensions and industry-wide pension funds. The Finnish Centre for Pensions (ETK) co-ordinate data and payments and the Ministry of Social Affairs and Health and the Financial Supervisory Authority supervise the funds. The national pension is administered by the Social Insurance Institution (KELA) and supervised by Parliament. The public sector earnedincome scheme has its own providers. In Finland, pension provision reflects agreements between the labour market organisations that are later codified into law. Thus unions, employers and the government all influence legislation. As described in Chapter 2, the financial assets and the transactions of the pension funds are in the general government accounts for Eurostat's statistical purposes. However, this and other similar systems are classified as private sector in to the OECD's pension classification scheme (OECD, 2009b). Moreover, from a strict legal point of view, the funds are private entities. In this Survey however, they are treated as part of the public sector. Contributions into the system will be seen as social security contributions and the ultimate fiscal responsibility will be assumed to rest with the government. The fact that the Finnish pension system can almost entirely be described as a defined-benefit system affects the distribution of risk across generations. A shock, for example, in terms of a permanent drop in asset values, will not affect accrued benefits, but will require increases in future contributions. This effectively lets current young and future generations bear a larger share of the risk. The exception is longevity risk, where the longevity adjustment allocates demographic risk to the generation receiving the pension. Compared to most OECD countries, third-pillar defined-contribution systems play a minor role. A larger reliance on defined-contribution schemes would be useful from a risk-sharing perspective, even though the government should be careful not to pursue such objectives with distorting subsidies.
The expected retirement age has been increasing, but may fall due to the recession
Labour supply in Finland is reduced by the relatively low retirement age. The average age of retirement remained roughly constant at 58.4 years between 2003 and 2008. According to an indicator constructed by the Finnish Centre for Pensions that reflects the age composition of the population, the expected retirement age for an individual at 50 years of age increased by 0.8 years between 2003 and 2009 to 61.7 years, and to 59.8 years for a 25-year old. (8) While this is a significant increase, it partly reflects a normalisation as the effects of the crisis in the early 1990s have faded from the labour force with recession-hit cohorts moving into old-age retirement (Gronqvist and Kinnunen, 2009). Furthermore, the increase has taken place during a period when labour markets were improving strongly (Figure 3.11). In fact, the fall in the unemployment rate can explain most of the increase in the expected retirement age since 1996, although the significant increase observed in 2009 when the labour market deteriorated is encouraging. The historical relationship between unemployment and the expected retirement age suggests that the expected retirement age may fall back during the next few years due to the recession.
The 2005 pension reform is expected to increase labour supply by raising the effective retirement age. It is too early to evaluate its full effects though (Box 3.3), as retirement decisions are still influenced by incentives and rules under the old system. For example, the previously available unemployment pension has not yet been fully phased out, and it continues to give a strong contribution to the number of retirees in Finland. An increasing share of new pensioners is retiring through the old-age pension however, and the share exiting through the unemployment and disability pensions has decreased significantly (Table 3.2). Some estimates show that the 2005 pension reform would increase the old-age retirement age by 8 months (Hakola and Maattanen, 2007). Nonetheless, the average retirement age would still remain well below the OECD average, and compared to the other Nordics it would be 4 years lower for men and 2 years for women. (9) The government's target is to increase the average retirement age by 3 years by 2030 (Ministry of Finance, 2008), but it is unclear to what extent this target incorporates effects of already implemented measures.
[FIGURE 3.11 OMITTED]
Retirement through the unemployment pipeline and disability pension needs to be restricted
The actual retirement age in Finland is lower than official statistics suggest, due to various early retirement schemes. Older workers tend to effectively withdraw from the labour market through the "unemployment pipeline" from 57 years of age and onwards. This pipeline provides unemployment benefits for an additional 3 years on top of the 500 standard days and allows individuals to retire to old-age pension at 62 without the penalty typically associated with early retirement. One indicator of incentives to continue to work instead of retiring is the change in net pension wealth accrued through working one additional year (additional benefits minus additional contributions), which can be regarded as the implicit tax on continued work (Duval, 2003). If the implicit tax is zero then the system is said to be "actuarially neutral". For an unemployed individual in the "pipeline" who is offered a job at the average wage, the implicit tax on working an additional hour on accrued benefits is close to 80%, while a job offer at 60% of the average wage would incur an implicit tax of close to 95% (Figure 3.12). The weak incentives to engage in job search under these conditions is illustrated by the fact that the number of unemployment beneficiaries among older individuals are much higher than the number of unemployed according to the Labour Force Survey (LFS) data. This means that many individuals receive unemployment benefits but do not consider themselves job seekers.
[FIGURE 3.12 OMITTED]
Entry into the unemployment pipeline has been tightened on several occasions and the impact of these partial reforms has been substantial. In 2005 the qualifying age for the pipeline was raised from 55 to 57, meaning that the last cohort to enter the old system was 55 years old in 2004 and turned 60 in 2009. The fall in long-term unemployment in the 55-59 age group has been dramatic since 2005 and much more pronounced than for the (unaffected) 50-54 age group (Figure 3.13). (10) Long-term unemployment among the 60-64 years old has increased substantially relative to the 50-54 age group during the same period, as the unemployment pension is being phased out and replaced by access to the unemployment pipeline.
[FIGURE 3.13 OMITTED]
The recession will increase pressure on the unemployment pipeline. Employers are more likely to permanently or temporarily lay off older workers who have access to the unemployment pipeline and thus to a secure income. Although long-term unemployment only has increased marginally so far, it may double within the coming two years if historical relationships between the headline and long-term unemployment still hold. (11) In order to restrict early withdrawal from the labour market, the government should follow up the success of previous reforms of the unemployment pipeline with a full-scale abolition, as recommended in the previous Survey, especially as there are looming labour shortages ahead.
Labour supply is also influenced by a large share of early retirements to disability pensions. Roughly 8.5% of Finland's working-age population receives disability benefits, a measurable decrease from the 10% prevailing in the early 1990s, but still among the highest in OECD. This rate has remained stable since 2001. While it is influenced by numerous factors (OECD, 2009a) generous replacement rates, durability and relatively lax assessment procedures (all measured by a compensation index) tend to be correlated with disability recipiency rates across the OECD area (Figure 3.14). With high replacement rates and less stringent assessment procedures in Finland relative to many other OECD countries (OECD, 2008b), it is not surprising that levels of disability pensions are high.
In most OECD countries it has proved difficult to affect the stock of disability recipients by increasing the outflow. In Finland, the outflow in 2006 was estimated to be 0.5% of the total stock, with roughly half of those exiting directly into unemployment (OECD, 2008b). The outflow is even lower for individuals who are older than 55 who face very high implicit tax rates if moving from the disability pension to work (Figure 3.12). The current focus on rehabilitation measures has also yielded little in terms of exit. Measures to lower the inflow into disability have proved more effective. Recent reforms in the Netherlands and Poland have shown significant success by tightening assessment procedures and restricting access to permanent benefits (OECD, 2009a). In Finland the initial medical examination and report is made by any qualified medical practitioner that the applicant chooses (typically a general practitioner) rather than by specialist insurance doctors (OECD, 2008b). This system makes it more likely that non-medical criteria are taken into consideration. The strong correlation between regional unemployment and disability pension suggests that local labour market considerations may influence disability pension decisions. Rehabilitation, assessment and financial responsibilities for disability benefits are also fragmented, potentially contributing to easy approvals. (12) With costs and responsibilities shared among many different agents, achieving a stricter assessment process and providing consistent rehabilitation quality and sheltered job offers become more difficult.
[FIGURE 3.14 OMITTED]
Rules should be changed such that disability pension is granted only on pure medical grounds. Shifting the initial medical evaluation away from patients' GPs towards insurance teams would help to achieve more consistency and increase focus on medical conditions. Vocational limitations should be minimised. Lower replacement rates in the disability pension should be considered in order to maintain incentives to remain in the labour force. Offering new entrants the choice between lower benefits and sheltered employment may be more efficient than current rehabilitation measures, but additional measures to increase labour demand for this group may be warranted. Furthermore, the government should reform the assessment and rehabilitation system so that one institution bears the full responsibility through the whole process. As municipalities bear the main costs for the related systems, such as BIS, they may be the most natural choice.
Old-age retirement ages remain low
Labour supply is also reduced by early old-age retirement. The 2005 pension reform increased the minimum and maximum retirement ages and introduced a mechanism for longevity adjustments, whereby retirement income will be lowered when expected longevity increases (see Box 3.3). Even though this reform significantly improved fiscal sustainability, the average retirement age into old-age pension has remained stuck at 63 years since 2003. While current retirement patterns are affected by the transition from the old system, the average retirement age is unlikely to rise, as incentives in the 2005 system are skewed towards retirement at 63. On one hand, there are relatively weak incentives for early old-age retirement at 62, as the implicit tax on working is low due to the heavy penalties applied to retiring early at 62 (Figure 3.15). On the other hand, the implicit tax on further work becomes much higher from 63 years onwards, as the 4.5% accrual rate is overwhelmed by the fact that no actuarial adjustment is applied during the 63-68 window of retirement (Borsch-Supan, 2005). Consequently, the number of individuals retiring at 63 is more than 5 times higher than at 62. (13) As few remain in the workforce after 63, new retirees reach much lower volumes at 64 and 65, whereafter there is virtually no one left to retire. The lowering of the statutory retirement age from 65 to 63 with the 2005 reform has resulted in a significant fall in the number of individuals staying in the labour force until 65 years of age.
Further pension reform is needed
Finland needs to ensure that working lives increase considerably in order to achieve sustainable public finances (see Chapter 2). To finance the current system, the Finnish Centre for Pensions estimates that contributions in relation to wages will need to increase by 5 percentage points over the next 15 years. Building on the 2005 pension reforms, further reforms should be pursued to raise the average retirement age and reduce costs. Reforms should be based on three broad types of measures:
* Incentives to continue working after the minimum retirement age should be strengthened. From an efficiency perspective, the marginal rate of taxation on labour for old workers should eventually approach zero, meaning that marginal implicit taxes on pensions should be negative.
* Deadweight losses related to excessive compensation during some phases of pension accrual should be minimised in order to lower costs.
* Early retirement paths should be discouraged through a combination of incentives and stronger gatekeeping, in order to make sure that less favourable accrual conditions do not spill over into early retirement, disability pension or unemployment. Hakola and Maatanen (2009) show that if the minimum retirement age was raised from 63 to 65, little would be achieved in terms of reducing labour market exit, as outflow would take place through the (unreformed) unemployment pipeline and disability pension system.
[FIGURE 3.15 OMITTED]
These goals could be reached by a set of reforms to the current system. Firstly, the minimum retirement age should be raised from 63 to 65 years of age and the upper retirement age abolished. Secondly, pension accrual up to 65 should be lowered to 1.5% and increased to 6% from 65 and onwards. Thirdly, actuarial adjustment should be applied during the full working life, including the period after the minimum retirement age. Such a reform would lower implicit taxes on further work dramatically and enhance incentives to continue to work after the age of 65 (Figure 3.15, "alternative system"). It should be accompanied by less stringent employment protection legislation (EPL) on dismissals after the minimum retirement age of 65 in order to raise employer's willingness to recruit and maintain workers older than 65, as older workers may be perceived to imply relatively large productivity risks. By lowering the accrual rate during ages 53-64 and considering lower accrual during non-working periods (unemployment, studying and parental leave), the deadweight costs of overcompensation are lowered, creating substantial savings for the earnings-related pension system equivalent to almost 2% of GDP (OECD, 2006b). To achieve these savings, substantial tightening of early retirement schemes would need to accompany these reforms as alternative paths of exit, such as the unemployment pipeline and disability pension, will become relatively more attractive. The minimum retirement age needs to be raised in order to counteract the high implicit taxes that would otherwise apply at ages 62-64. The costs of raising accrual and actuarial adjustment for older workers would be minimal as there are virtually no deadweight losses.
It is difficult to evaluate the overall effects of wide ranging reforms, especially if they encompass changes in the disability pension and the unemployment pipeline where implementation and gate keeping issues are key. Some initial assessments of the employment effects for the 65-69 years old age group can be made though. Given the more than 60 percentage point difference in the implicit tax rate for the 65-69 years old relative to the current system, participation rates should be able to increase by roughly 20 percentage points for this age group, amounting to an overall increase in participation of 1.2 percentage points. (14) Such an increase would move participation among 65-69 year olds in Finland above levels in Denmark, Sweden and Norway but remain significantly below Iceland. On top of that, an increase in the minimum retirement age, combined with tightening of entry into early retirement and a loosening of EPL for older workers could yield significant gains.
Pension reforms to boost the labour supply of older workers should be accompanied by measures to stimulate labour demand for older workers too. In 2006, the government introduced a payroll subsidy scheme aimed at employees older than 54 years with low wages in order to increase labour demand for this group. However, this reform seems to have had little success in raising employment levels so far (Huttonen et al., 2009). Although the payroll subsidy for older workers has not been effective so far and may need to be reconsidered, the government should continue to search for ways to raise labour demand for older workers. Employers also need to develop strategies to make better use of older employees, e.g. in terms of offering transition to part-time positions, especially in the light of a greying workforce.
Box 3.4. Summary of recommendations on wage formation, labour markets and pensions Wage formation * If the current round fails to achieve moderate wage outcomes in line with economic conditions, the industry-wide agreements may need to be replaced by arrangements with a higher degree of co-ordination to stop the decline in competitiveness. Wage flexibility also needs to increase to better reflect local productivity differentials. * The high minimum wages that are likely to contribute to low employment among unskilled youth should be re-considered and experience rating mechanisms strengthened. Work incentives * Unemployment benefits should be lowered and further tapered as the unemployment spell lengthens. Announce as early as possible but implement when recovery firms. * The government should consider equalising replacement rates and activation for the Labour Market Support (LMS) and the Basic Income Support (BIS) for those that have work capacity. * The recently proposed change to the housing benefit is welcome but the government needs to assure that it efficiently addresses inactivity traps and is supported by allowances better targeted at poor households. * Let municipalities shoulder the full responsibility of the LMS in order to be responsible for both the LMS and the BIS and finance them with block grants. Matching * Ensure earlier mandatory activation for the unemployed. Increase the funding for the Public Employment Service temporarily to deal with increasing caseloads and step up volumes of activation measures. * Make sure that available profiling tools are used by the local labour market boards to identify individuals with high risks for long-term unemployment to focus resources and activation measures to the most needy. * Strengthen central coordination in the Public Employment Service to harmonise intervention procedures across local labour market boards and improve efficiency. Extend performance-based pay in the Public Employment Service. * Phase out allowance for work-related second residence. Further tighten legal requirements for geographical mobility in job search. * Consider nationalising the unemployment insurance. The pension system * Abolish the unemployment pipeline immediately. * Tighten access to the disability pension and ensure that it is awarded on purely medical grounds. Consider lower replacement rates in the disability pension. Shift focus from rehabilitation to sheltered work. Consider giving municipalities the full responsibility for the disability pension. * The minimum retirement age in the old-age pension system should be raised from 63 to 65 years. The maximum retirement age should be increased or abolished. * Apply actuarial adjustment also during the period after the minimum retirement age to improve incentives to remain employed after the minimum retirement age. * Abolish the increased accrual rates applicable from 53 years of age and upwards. Consider lowering accrual during non-working periods (unemployment, study, parental leave). * Consider loosening EPL for workers above the minimum retirement age as well as other measures to support demand for older workers. * Consider measures to encourage a more significant defined-contribution third pillar, to improve intergenerational risk sharing.
Andersen, T. and M. Svarer (2009), "Business Cycle Dependent Unemployment Insurance", CEPR Discussion Paper, No. DP7334, London.
Asplund, R. (2007), "Finland: Decentralisation Tendencies Within a Collective Bargaining System", ETLA Discussion Papers, No. 1077, Helsinki.
Bassanini, A. and R. Duval (2006), "Employment Patterns in OECD Countries: Reassessing the Role of Policies and Institutions", OECD Economics Department Working Papers, No. 486, OECD, Paris.
Borsch-Supan, A. (2005), "The 2005 Pension Reform in Finland", Finnish Centre for Pensions Working Paper, 2005:1, Helsinki.
Duell, N., D. Grubb and S. Singh (2009), "Activation Policies in Finland", OECD Social, Employment and Migration Working Papers, No. 98, OECD, Paris.
Duval, R. (2003),"The Retiiement Effects of Old-Age Pension and Early Retirement Schemes in OECD Countries", OECD Economics Department Working Papers, No. 370, OECD, Paris.
Fredriksson, P. and B. Holmlund (2006), "Optimal Unemployment Insurance Design: Time limits, Monitoring, or Workfare?", International Tax and Public Finance, Vol. 13, pp. 565-58.5.
Frohlich, M. et al. (2004), "Statistically Assisted Programme Selection International Experience and Potential Benefits for Switzerland", IFAU Working Paper, 2004:1, Uppsala.
Gronqvist, C. and H. Kinnunen (2009), "Impact of Recession on Labour Supply: Experiences from the 1990s in Finland", BoF Online, 10/2009, Helsinki.
Hakola, T. and H. Maatanen (2007), "Vuoden 2005 Elakeuudistuksen Vaikutus Elakkeelle Siirtymiseen ja Elakkeisiin Arviointia Stokastisella Elinkaarimallilla", Finnish Center for Pensions, Helsinki.
Hakola, T. and H. Maatanen (2009), "Pension System, Unemployment Insurance and Employment at Older Ages in Finland", Prime Minister's Office Publications, 2009/2, Helsinki.
Holden, S. and F. Wulfsberg (2007), "Downward Nominal Wage Rigidity in the OECD", ECB Working Paper Series, No. 777, Frankfurt.
Huttunen, K., J. Pirttila and R. Uusitalo (2009), "The Employment Effects of Low-wage Subsidies", Labour Institute for Economic Research Discussion Papers, No. 254, Helsinki.
Hynninen, S. et al. (2009), "Matching Inefficiencies, Regional Disparities, and Unemployment", Labour, Vol. 23, Issue 3, pp. 481-506.
Johansson, [Angstrom]. (2006), "Wage Setting in Finland: Flexibility in Centralised Wage Agreements", OECD Economics Department Working Papers, No. 503, OECD, Paris.
Kannisto, J. (2004), "Effective Retirement Age in Earnings-related Pension Scheme in 1996-2003", Finnish Centre for Pensions Working Paper, No. 8, Finnish Centre for Pensions, Helsinki.
Ministry of Finance (2008), "Stability Programme Update for Finland 2008", Helsinki.
OECD (1998), OECD Employment Outlook, OECD, Paris.
OECD (2006a), OECD Employment Outlook, OECD, Paris.
OECD (2006b), OECD Economic Surveys: Finland, OECD, Paris.
OECD (2007), OECD Economic Surveys: Sweden, OECD, Paris.
OECD (2008a), OECD Economic Surveys: Finland, OECD, Paris.
OECD (2008b), Sickness, Disability and Work (Vol. 3): Denmark, Finland, Ireland and the Netherlands, OECD, Paris.
OECD (2009a), OECD Employment Outlook, OECD, Paris.
OECD (2009b), Pensions at a Glance, OECD, Paris.
Prime Minister's Office (2007), "Government Programme of Prime Minister Matti Vanhanen's second Cabinet", Helsinki.
Uusitalo, R. and J. Verho (2007), "The Effect of Unemployment Benefits on Re-employment Rates: Evidence from the Finnish UI-Benefit Reform", Labour Institute for Economic Research Working Paper, No. 229, Helsinki.
Verho, J. (2008), "Scars of Recession: Long-term Costs of the Finnish Economic Crisis", IFAU Working Paper, 2008:9, Uppsala.
(1.) Under the industrial agreement (Industriautalet) of 1997 in Sweden, the unions and the employers agree on overall average wage increases, while actual negotiations take place on the industry or even individual level. The unions and the employers also share a joint economic council (Industrins ekonomiska rad) to gauge economic development and suggest a range of reasonable aggregate wage increases. Furthermore, the National Mediation Office (Medlingsinstitutet) and the National Institute of Economic Research (Konjunkturinstitutet) provide annual information and recommendations on wage formation.
(2.) This holds for both the incidence of low pay and for the relative wage of the lowest wage decile compared to the highest (D9/D1) or the average (D5/D1) decile. Finland is among the few OECD countries that do not have a statutory minimum wage, as minimum wages are determined by collective agreements and later enshrined into legislation.
(3.) In Sweden, a number of reforms to the unemployment benefit system was made in 2006 and 2007 including abolishing the first 100-day top-up and lowering maximum benefits from 80% to 70% after 200 days and to 65% after 300 days. These reforms have been estimated to lower the structural unemployment rate by 0.5 percentage points (OECD, 2007).
(4.) There are strong reasons to believe that an optimal unemployment insurance should be countercyclical, i.e. providing higher and longer replacement during recessions (Andersen and Svarer, 2009). To construct an efficient and credible framework for implementing such a model would be difficult, however.
(5.) While the recently ended pilot scheme on privileged income, which let households keep up to 20% of income without affecting the level of the BIS, dealt with some issues related to inactivity traps, its effectiveness has not yet been evaluated.
(6.) In 2007, roughly 45% of the households below the poverty line (60% of average disposable income) did not receive housing benefits. 43% received child benefits and 30% received pension (Statistics Finland: Income Distribution Survey microdata from 2007).
(7.) Bonuses are capped below 7% of the average wage, while productivity differences among case officers are likely to be much higher. In discussions with government officials, it has been stated that "30% of case officers do 90% of all placements".
(8.) See Kannisto (2004) for further details.
(9.) See www.oecd.org/document/47/0,3343,en_2649_34747_39371887_1_1_1_1,00.html for further details.
(10.) The fall in long-term unemployment has been accompanied by corresponding increases in employment, while there is no evidence of larger inflow into disability pension for that age group during 2005-08.
(11.) Estimated over the sample 1991:1-2009:3 using a simple ordinary least squares of long-term unemployment on lagged total unemployment.
(12.) Rehabilitation, assessment and financial responsibilities for disability benefits are fragmented and could involve pension providers, KELA, the PES and municipalities (OECD, 2008b).
(13.) It should be noted that the unemployment tunnel most likely contributes to raising retirement at 62 years of age relative to 63, as individuals taking the tunnel can move into the old-age pension system at 62 without the penalty.
(14.) Based on estimates from Duval (2003). These estimates are likely to define an upper bound as they are estimated on the 60-64 age group.
Table 3.1. Revenues and expenditures of the unemployment funds In EUR million, 2008 Revenues EUR Mn % of total Membership payments 131 9 Government 560 37 Insurance contributions 822 54 Of which: Employers 699 46 Employees 123 8 Other revenues 19 1 Total revenues 1 531 100 Revenues Expenditures EUR Mn % of total Membership payments Benefits 1 445 95 Government Management 64 4 Insurance contributions Of which: Employers Employees Other revenues Other expenditures 19 1 Total revenues 1 524 100 Source: Data provided by the Ministry of Employment and the Economy, and OECD computations. Table 3.2. New retirees 2001 and 2008 Ordinary Early old-age Unemployment Disability old-age pension pension pension pension 2001 13 858 3 872 14 802 24 082 24% 7% 26% 42% 2008 28 949 3 400 12 613 27 638 40% 5% 17% 38% Ordinary Special pensions All new Part-time old-age pension for farmers retirees (1) pension (2) 2001 13 858 1 263 57 406 8 995 24% 2% 100% 16% 2008 28 949 1 136 72 668 7 032 40% 2% 100% 10% (1.) Excluding part-time pension. (2.) Share of all new retirees excluding part-time pension. Source: Finnish Centre for Pensions and OECD calculations.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Chapter 3|
|Publication:||OECD Economic Surveys - Finland|
|Date:||Apr 1, 2010|
|Previous Article:||Paving the way for sustainable public finances.|
|Next Article:||Rising inequalities challenge Finland's social model.|