Chapter 2. Getting the incentives right in an ageing society

The generous Danish welfare state relies on a high degree of labour force participation both for financing and in order to ensure social cohesion. This underlines the need for getting work incentives right and improve the employability of vulnerable groups of workers, in particular migrants. Many benefit recipients also face high marginal tax rates for returning to work, creating a barrier for inclusion. Likewise, as the population ages, the need for longer working lives becomes a central aim. In Denmark, much has been done to keep older workers in the labour market, but there is further scope for reducing barriers to work for this group, including through the design of the pension system. Cost pressures at social institutions could be addressed by better reaping the effects on municipal reform, more coordination between different service providers, and open the market for social services, for instance old age care, for private suppliers under a strict quality monitoring framework.

  

The Danish welfare system provides material well-being and stability for the whole population. The flexicurity model, which combines a high level of social protection with flexible labour market institutions, has proven its capacity to absorb shocks during the global crisis. Despite anaemic GPD growth since 2009, well-being in terms of both material conditions and quality of life has remained very high. Maintaining strong social institutions is crucial, notably in light of existing macroeconomic risks. The high level of indebtedness and the reliance on house price developments may reduce the capacity of households in responding to shocks and weigh on future demand (Chapter 1).

Social institutions in Denmark are confronted by two main challenges: the high number of social benefit recipients, with large groups in inactivity, and an ageing population putting pressure public spe nding on welfare services. Public social spending per unit of GDP – already high in Denmark – has grown in almost all OECD countries since the crisis, but the rise was more pronounced in Denmark following the recent boom/bust cycle of the economy (Figure 2.1, Panel A). More worrying, while employment and participation rates of 25-64 year are high by international standards they have been flat in recent years (Figure 2.1, Panel B). This is to some extent related to the increase of enrolment in education programmes following the crisis. While ageing in Denmark is not as marked as in some other OECD countries, demographic change will increase age-related spending (i.e. pension, healthcare, long-term care) and put pressure on the tax base.

Figure 2.1. Social spending and inactivity have increased over the past decade
picture

Source: OECD Social Expenditure database; OECD Employment Outlook database.

 http://dx.doi.org/10.1787/888933352578

Policy makers are already responding to these challenges. The authorities expect recent structural reforms to secure the sustainability of the welfare system, and maintain a high level of social protection, notably by increasing labour market participation among older workers. In particular, disincentives to work and thereby to contribute to the tax-financed system have been reduced. However, the expected future impact of the reforms, notably on effective retirement age, is outside past experience. Maintaining the current generous system without increasing the tax burden – already the highest in the OECD – therefore appears challenging.

The main messages of this chapter concern the return to work in the short and long run and the cost-efficiency of social institutions:

  • Recent reforms of unemployment benefits and social assistance benefits have reinforced economic incentives to work. This should be complemented by further measures improving the employability of disadvantaged workers (disabled, migrants, low skilled, older workers) according to individual needs.

  • In the longer run, lower than expected labour market participation of older workers poses a risk to fiscal sustainability and the robustness of social institutions. The pension system should be reformed for specific groups, who currently do not save enough for retirement. Work environment will have to be adapted to an ageing work force.

  • Efficiency gains in delivering social services could be achieved by eliminating barriers to competition and ensuring a more effective distribution of resources. Improving the coordination of providers of welfare services, in particular in the healthcare sector and putting more emphasis on health prevention and rehabilitation, can also reduce the cost of ageing.

Addressing barriers to employment for benefit recipients

Social institutions, defined as pension, health and long-term care, unemployment and social assistance schemes, account for a large share of public spending (Figure 2.2). The free access for all citizens to generous public services and the redistribution of income are ensured by a large tax base due to high employment rates and a high level of income taxes. This does, however, also make the system vulnerable to a weakening of the tax base, due to e.g. lower participation rates.

Figure 2.2. Social institutions account for a large share of public spending
Components of public spending, 2013
picture

1. “Others” includes defence, public order and safety, environmental protection, housing community amenities, recreation, culture and religion.

Source: OECD (2015a), Government at a Glance.

 http://dx.doi.org/10.1787/888933352746

The system is challenged by both population ageing and the high level of untapped labour resources. While the effect of ageing is less dramatic than in peer countries (e.g. Germany, the Netherlands, and Finland), the old age support ratio will continue to decline (the ratio between those aged 15-64 years and those aged 65 years and over, i.e. the number of person in working-age for every elderly person) from 3.5 to 2.3 by 2060, a level comparable to the OECD average, contributing to both reducing public revenues and increasing social spending. At the same time, the proportion of 16 to 64-year-olds on benefits reached 22% in 2014 and which weighs on public finances through lower growth rates, lower tax revenues, and higher social spending. Long-term unemployment rose by 1 percentage point since the crisis (from 2.9% of working age population in 2007 to 3.9% in 2013, but remains low by international and historic standards). The budgetary loss related to unemployment, calculated as the sum of benefit entitlements in the initial phase of unemployment (assuming entitlement to, and full take-up of, insurance and any means-tested benefits) and the direct taxes and social contributions (including by employers) payable in the lost job, is estimated to exceed the OECD average of 80% of GDP per head, reflecting the strong automatic stabilisers at work in Denmark (Immervoll, 2012).

In response to the increasing cost of social institutions and national goal of increasing employment rate from current 72% to 80% by 2020 (Denmark’s National Reform Programme, 2015), the Danish flexicurity model, based on flexible hiring and firing regulations, a generous social safety net and strong active labour market policies, has undergone important changes in the recent past (Box 2.1). Since 2010, the total number of benefit recipients (i.e. those receiving unemployment, cash, sickness and disability benefits and public pensions) has declined with fewer in early retirement, unemployment and sickness benefits (Figure 2.3, Panel A). However, the number of social assistance recipients relying on the social safety net, i.e. those who not eligible for other benefits, has increased (Figure 2.3, Panel B).

Box 2.1. Recent reforms of social institutions

Unemployment insurance

  • Originally, the 2010 reform of the unemployment insurance benefit (UIB) system reduced the maximum benefit duration from four to two years and increased the re-eligibility requirements for those expiring the full UI period. The 2015 revision of the reform allows an extension of the unemployment benefit from 2 years to a maximum of 3 years, if the benefit recipient has worked sufficiently, which earns him/her additional eligibility (i.e. for a day of work, two days of benefit entitlements are acquired). The possibility of an extra year of benefits is largely financed by a reduction in the unemployment benefit level for graduates. Waiting days (i.e. days without benefits) are introduced after periods of unemployment, but the waiting day can be avoided if the claimant meets a certain employment requirement. The 2015 reform will be fully implemented during 2017. Correspondingly, temporary measures introduced following the 2010 reform for those who exhausted entitlement to benefits will be phased out in 2017.

Social assistance

  • In 2013 cash benefits were reduced to the level of the education grant for those under 30. Recipients under 30 who are not ready for education or employment can receive an activity supplement. In addition, education and employment efforts are initiated sooner; youth without qualification (40% of those receiving social assistance) has to undertake education; those with education are offered activation measures, including longer vocational traineeships, earlier than previously and monitoring of job search is strengthened.

  • In 2016, a cap on total amount of social benefits a household can receive has been introduced (depending on the number of children and the marital status). Furthermore, new eligibility criteria require at least seven years of within country residence over past eight years, as well as 225 hours of work within a year. Benefit recipients that do not meet these conditions receive a lower cash benefits.

Disability and sickness benefits

  • In 2013 a reform introduced a number of changes. Persons over the age of 40 must participate in at least one so-called rehabilitation programme before being granted disability benefit. The rehabilitation programme involves coordinated health, social and employment services and aims at strengthening the individual’s work capacity. Applicants under the age of 40 are by default not granted disability benefit and must instead participate in one or more rehabilitation programmes. Those with no chance of improving work capacity, e.g. persons with serious brain damage, can be granted disability benefits without preceding rehabilitation programmes. Also, “flex jobs”, i.e. subsidised jobs offered to people with reduced working capacity, have been made more targeted to those with low work capacity, more flexible in terms of time arrangement and granted for 5 years maximum (except for persons over 40 years old, who may be allowed a permanent flex job).

  • A senior disability pension was introduced in January 2014 for those eligible for disability pension and with less than five years to retirement. However, such recipient has to have a minimum of 25 years of full time employment in the Danish labour market. Under the scheme, the benefit recipient is exempt from a prolonged so-called job clarification programme, i.e. an inter-disciplinary assessment of work capacity and its potential development. Partly due to the tight eligibility criteria (corresponding to the ones for regular disability pensions), the scheme is so far not widely used.

  • Since July 2014, the assessment of eligibility for a prolonged sickness leave takes place every 22 weeks (instead of 52 weeks). Those on sick leave, who cannot get their sickness benefits extended, and unable to find a job, have a possibility of entering into a the job clarifying programme. Under this scheme, the person on sick leave receives an income replacement benefit that is lower than the sickness benefit.

Pensions

  • Reforms of 2006 and 2011 introduced a gradually increasing statutory retirement age. In line with these reforms, the Parliament approved in November 2015 an increase of the statutory retirement age to 67 in 2022, and 68 years in 2030. From then on, the statutory retirement age will be – conditional on further Parliament’s approval – adjusted to life expectancy gains every five years (by a maximum of one year) to limit pension payments to 14.5 years on average. The duration of voluntary early retirement programmes is also being progressively shortened from five to three years with the minimum eligibility age increasing in line with the statutory retirement age. At the same time, earlier reforms have limited the incentive to opt into the early retirement scheme, especially among the young (Box 2.5).

Active labour market policies

The 2014 reform of public employment service strengthened screening and provision of individualised services to job seekers. As of January 2016, a new model for reimbursements of municipal expenses for a number of benefits (notably UIB, cash benefits and disability benefits) is in place. The reimbursement of municipalities will be determined by the duration of the benefit spell, rather than the type of benefit the person receives. This should encourage municipalities to speed up the return of benefits recipients into work or education.

Figure 2.3. Number of social assistance recipients has risen since the crisis
picture

Source: Danish Ministry of Finance; OECD SOCR database.

 http://dx.doi.org/10.1787/888933352585

Bringing more benefit recipients into work while upgrading their skills can increase social inclusivity and reduce inequality. The authorities should focus on low-skilled workers, migrants and disabled, who – as in most other OECD countries – encounter specific barriers to employment (Figure 2.4). By international comparison, employment gaps are relatively high in Denmark, in particular for immigrants, though this is partly a result of high employment rates among native born. In order to better understand the costs and benefits of employment for each group, one should identify those who are expected to pose the highest future net cost to the welfare system and set priorities accordingly, an assessment of the lifetime pattern of benefit recipients and its costs should be implemented as done for instance in New Zealand (Taylor Fry, 2013).

Figure 2.4. Scope for increasing employment is particularly high among immigrants
Gap in employment rates, in percentage points
picture

1. Low educated are those with less than upper secondary education.

Source: OECD Employment Outlook 2015; OECD (2010), Sickness, Disability and Work: Breaking the Barriers; Eurostat database; OECD Education at a Glance 2015.

Balancing protection and work incentives for job seekers and low income earners

A balance between high coverage, generosity of the system and work incentives, is difficult to find on pure economic grounds. The Danish flexicurity system is said to achieve this combination, but inactivity traps continue to be a matter of concern, and policies try to combine equity and efficiency.

Work incentives for job seekers and social assistance recipients have been raised by cutting duration and levels of the unemployment insurance benefits (UIB), capping the level of social assistance, and tightening eligibility criteria for both UIB and social assistance (see Box 2.1). These reforms have led to improvements in labour market performance: The Danish Economic Council (2014) estimates that unemployment rate dropped by 0.5 percentage points following the UIB reform and recent data show a decline in the number of young social assistance recipients in 2014, mainly as a result of the 2013 social assistance reform (Danish Ministry of Finance, 2015). Around 9% of the social assistance recipients aged between 25 and 29 years have left social assistance and between 1 600 and 2 600 25-29 year olds have found employment, corresponding to an increase in the employment share for this age group by 0.5-0.8 percentage points (Economic Council, 2015b). Also, the number of unemployed whose entitlement to unemployment benefits expired increased significantly (from around 5 000 in 2012 to more than 30 000 in 2013, but has since been reduced to 15 000 in 2015), in part due to phase-in of the 2010 UIB reform, with consequences on lower-tier benefits.

Box 2.2. The Danish unemployment insurance benefit system

A generous system with strict requirements

By international standards, the Danish UIB scheme is generous (with a 74% net replacement rate vis-à-vis 55% in the OECD for low income workers) and highly redistributive (with benefits capped to around 53% of the average wage). As a consequence, the income gain of taking up low-paid or part-time employment, or a job that pays less than the previous one while on the UIB benefits, is low, in particular for low wage earners.

At the same time, eligibility requirements are relatively strict (labour income of 212 000 kroners in the last three years, corresponding approximately to a year full time employment at the prevailing minimum wage set by collective agreements) and entitlement to unemployment benefits is conditional on the claimants’ job search effort. After three months, unemployment benefits recipients have to participate in activation programmes and to accept job offers regardless of their previous occupation and the wage offered. According to the new OECD indicator on eligibility criteria for unemployment benefits, the Danish system ranks among the strictest in terms of availability requirements, suitable work criteria, job search requirement and monitoring (Langenbucher, 2015).

A reduced coverage with a risk to public finances

The system is financed by members’ voluntary contribution and public subsidies. When the UIB scheme is in deficit, such as in 2013, 70% of the benefits were financed by taxes. Despite the improvement of the labour market, the share of tax financing has not declined over the past years. Compared with other OECD countries, unemployment insurance spending is close to the average but more volatile (Pareliussen, 2014). Moreover, shocks on unemployment have a higher impact on the Danish UIB spending than in an average OECD country, reflecting relatively generous benefits combined with fairly low unemployment.

Membership in UIB is wide-spread (around 80%) and has been quite stable over the past two decades despite reforms reducing the generosity of the system. However, for the young generation (i.e. below 30 years old) membership has declined and the coverage of the unemployed has fallen from more than 65% of the unemployed in 1995 to around 61% in 2015 partly due to the strengthening of the eligibility conditions and changes in educational patterns, which implies that people sign up for the UIB system later. The decline in coverage has been more pronounced for low-wage workers who have a higher risk of unemployment but also a lower income gain of participating in the scheme compared to social assistance benefits, in spite of high replacement rations for this group (Parsons et al., 2015). Indeed, the level of UIB for those with below-average previous earnings is close to the level of social assistance (from 89% to 103% depending on the family situation; Immervoll et al., 2015).

Impact of the 2010 and 2015 reform

The 2010 reform cut the benefit duration from 4 to 2 years and increased re-eligibility requirements for those exhausting the UIB period, in order to foster more rapid return to work, and thereby reducing UIB spending. In return, the flexibility of the labour code has been reduced with an introduction of a severance pay for the low-income earners. The impact of the reform on employment is set to be positive. The UIB Commission estimates an increase in employment of 15 000 persons.

The number of unemployed whose entitlement to benefits expired increased significantly in the phase-in period (from around 5 000 in 2012 to more than 30 000 in 2013, falling to 15 000 in 2015) with consequences on lower-tier benefits. In 2014, less than 10% of those who exhausted their benefits went into employment immediately, while around 20% received a form of social transfers and more than 70% received other benefits. However, after 3 months between 20 and 30% re-entered employment. With the unemployment benefit reform from October 2015 the number of persons who exhaust their benefits is expected to be reduced to around 7 000 persons by 2017, because of the possibility to extend the benefit duration to 3 years (see Box 2.1). The reform is expected to have a minor negative effect on structural employment.

There is room for strengthening work incentives in the UIB system further. Despite the 2010 and the 2015 reforms, the unemployment benefits remain relatively generous and economic disincentives to work, in particular for low paid workers, are strong (Box 2.2). Empirical analyses suggest incentives effect of changes in benefit generosity in countries with more generous unemployment compensation (i.e. Nordic countries) can be relatively large (Immervoll, 2012). Because exit from unemployment into employment spikes at the end of the entitlement period, decreasing benefit gradually with the length of the unemployment spell would accelerate return to employment and thereby limit the decline in earning potential of job seekers.

Such measure could have side effects. Because membership in the UIB system is voluntary, reducing its generosity further might discourage participation of workers, reducing overall coverage. Coverage has been good (close to 80%) and stable over the past decade, though it has declined slightly for young generations, primarily due to longer education and later labour market entry (Dagpengekommissionen, 2015). A decline of coverage would undermine the capacity to protect workers against income loss and conditions for effective job search. Furthermore, the impact of such a measure on inequality should be taken into consideration. A recent OECD study suggests that reducing maximum benefit duration and benefit levels increases income inequality, since it disproportionally affects those with more frequent and longer spells of unemployment (Causa et al., 2016, forthcoming).

The 2015 UIB reform eases of the rules for jobseekers with short or interrupted employment records to extend benefit rights, and could thereby improve incentives for those in non-regular jobs to contribute to the UIB scheme and for getting temporary jobs. All in all, the system remains complex, and has perhaps become even more so due to the 2015 reform. It is unclear what incentives exist for the self-employed and those on non-standard contracts. A comprehensive analysis of this issue is expected to be finished by in the first half of 2017.

Generous social assistance and means-tested benefits are often associated with high effective taxation of returning to work, and this link is pronounced in Denmark (Figure 2.5). Some OECD countries have a similar or higher level of minimum guaranteed income but lower effective taxation of returning to work for low-income earners. In Denmark, taking up a low-paid job entails an after tax income loss or only a minor gain for those receiving means-tested housing benefits, social assistance, and/or childcare benefits. Out of 150 000 benefit recipients in 2014, around 25 000 persons older than 30 years faced an implicit tax of at least 90% (Economic Council, 2015b). The implicit tax on returning to work is significantly above the OECD average, exceeding 100% for single parents taking up a job paid half of the average wage, and exceeding 85% when taking up full time job that pays two-thirds of the average wage.

Figure 2.5. Generous social assistance is combined with relatively high effective taxation
Minimum guaranteed income and average effective tax to work, 2014
picture

Note: Average Effective Tax Rates for a transition into full-time work paid 67% of the average wage for persons without entitlement to unemployment insurance but entitled to social assistance and cash minimum income benefits (including housing benefits) in percent of median household incomes.

Source: OECD, Tax-Benefit Models.

 http://dx.doi.org/10.1787/888933352594

In 2015, an increase in the maximum tax deductibility of the trade union fees and a minor increase in the earned income tax credit (“beskæftigelsesfradrag”) did not overall significantly increase income gains from work. On the other hand, the cap on total social benefits a household can receive, which was reintroduced in 2016, is expected to reduce the tax on returning to work for the concerned benefit recipients.

Raising income gain of work, while preserving the high level of benefit, is challenging. General in-work benefits exist already: when on social assistance, there is a threshold from which work income is deducted from social assistance and it was raised in 2014 (from DKK 15 to DKK 25 per hour worked within a maximum of 160 hours per month). Introducing stronger in-work benefits targeted to low-paid workers and withdrawing benefits more gradually for those who take up a low-paid job would make work more attractive by increasing the difference between the take-home pay and the alternative of relying on social benefits only. In-work benefits for all low-paid workers could be costly, given the compressed wage distribution and limited fiscal space. Thus, they should primarily benefit vulnerable groups. A monthly bonus for restricted groups of long-term unemployed who find a job as announced by the previous administration is an option, which, however, has not been taken up by the current government. Instead, the current government has reduced cash benefits for people having lived in Denmark for less than 7 of the latest 8 years and has introduced a cap on the amount of cash benefits a person or a family may receive.

Another option is to reduce social assistance while stepping up training and job search requirements. This was done in Germany, where the Hartz reforms significantly lowered benefits, introduced top-up subsistence payments for those in low-paid jobs, and improved job placement services, (Huefner and Klein, 2012). However, these have increased the poverty risk for the unemployed and a large share of benefit recipients who found a job continue to depend on top-up benefits (OECD, 2014a). Should the Danish social assistance be reformed in this direction, i.e. reduced further, measures supporting upwards income mobility, including upskilling programmes, would have to be reinforced to reduce poverty risk among vulnerable individuals with low employability and to avoid that they remain stuck in low-paid and precarious jobs.

Other levers to incentivise welfare recipients include strict job search requirements and eligibility conditions. Entitlement to social assistance is means-tested and conditional on the claimants’ effort to gain employment, which is closely monitored by social services. Also, from autumn 2016, welfare recipients have to work at least 225 hours over one year to receive the full benefit rate. This workfare policy is supposed to lower the reservation wage, increase job search incentives and thereby decrease inactivity. The measure should be carefully monitored and complemented by additional job counselling activities. A similar policy was in place from 2006 to 2011 and had a small positive impact on employment of recipients, in particular in municipalities offering extra support for job search. However, it also significantly decreased income for those affected by the reform and encouraged the social services to move recipients into other benefit schemes instead of applying the sanctions (Pedersen, 2013; Diop-Christensen, 2015).

Social assistance recipients often face multiple barriers to employment, including skills that do not match employers’ needs. This is particularly relevant for young persons who account for around 30% of social assistance recipients. Among them, only 7% are considered as ready to work and 40% have to pursue education. In this respect, effective labour market policies are central. As stressed in Jamet and Koen (2014), the efficiency of active labour market policies and upskilling programmes should be improved. This can be done by ensuring that municipalities face the right financial incentives to help the unemployed find a job and by increasing the quality control of training in the adult learning system. The 2014 reform of public employment services goes in the right direction (see Box 2.1). A new model for reimbursement by the central government of a share of municipalities’ expenses for income transfers has taken effect from January 2016. The level of reimbursement is now based on the length of the benefit period, rather than the type of benefit. This should also improve the performance of public employment services as it encourages municipalities to speed up the return of benefits recipients into work or education. Nevertheless, the capacity of municipalities to improve the employability of job seekers and upgrade skills of job seekers can be a challenge, and monitoring of training programmes needs improving. Encouraging information exchange among municipalities and the public employment office about successful experiences could help in this respect.

Promoting work and income security for disability benefit recipients

The share of the working age population receiving disability benefits remains above the OECD average, at 6% of the working age population in 2014. Around 30% of disability benefits recipients in that year were 60 years or older, while they accounted for only 24% of the total Danish population. Increases in the retirement age are likely to lead to rising number of benefit claims, not least for those in physically demanding occupations (see below).

Integration of the disabled on the labour market has not improved significantly since early 2000, with a substantial gap in employment rate between people with and without disability (see Figure 2.4, Panel B) suggesting work capacity of disabled can be better used. New activation programmes involving multidisciplinary rehabilitation teams (so called “resource process”) are being implemented. Entitlement to disability permanent benefit has been restricted, especially for persons below 40 years old, to better target recipients and to avoid an inactivity trap (see Box 2.1). While new entries in disability benefit scheme have started to decrease, the impact of the reform will be limited in the short run as it does not change conditions for those already granted disability and the flex-job schemes, requiring further adjustments:

  • The reform should be carefully monitored to ensure a permanent entitlement to disability benefit is granted only to those with a permanent inability to work, and that the number and quality of the rehabilitation programmes deliver objectives of the reform. The arbitrary age limit of 40 to participate in more than one rehabilitation programme should also be reconsidered to avoid disabled with work capacity being trapped in inactivity.

  • In 2014, around 40% of disability pensions were awarded on the basis of mental disease. The number of recognised mental illness has increased dramatically in Denmark (from less than 3 per 100 000 persons in employment in 1997 to more than 8 persons in 2011) and is much higher than in such countries as Sweden, Italy and France (Dansk Arbejdsgiverforening, 2015). At the same time, waiting time for diagnosis and treatment of persons with mental illness are long (Kvist, 2014 and OECD, 2013d). These developments warrant a deeper review of the origin and costs of mental health issues in Denmark.

  • Disability pensions are granted on a permanent basis, and it is possible that some benefit recipients might have recovered some work ability over time. Thus, a regular entitlement assessment of disability pensions should be implemented. Another obstacle to return to work is the fact that recipients might not gain much financially from re-entering the labour market: the replacement rate is close to 80% at the average wage level and 115% for a low wage earner (OECD, 2014c).

To improve inclusiveness, reforms of disability benefit systems should be complemented by measures to improve the employability of disabled people. The introduction of new rehabilitation programmes with integrated health and employment services is a step in the right direction. In this vein, prevention of diseases that create a progressive detachment of the labour market should be strengthened. Policies to increase awareness about the nature of disabilities, how they can be overcome and at what cost are also crucial to allow workers with partial work capacity to keep a foothold in employment.

Integration of immigrants can be strengthened

Denmark has one of the smallest immigrant populations among OECD countries, with the foreign born accounting for 8% of the total population in 2013 (13% on average in the OECD). At the same time, immigration has intensified over the past decades, with permanent migration flows reaching close to 1% of the total population in 2013 and increasing by 73% when compared to 2007. The share of immigrants for humanitarian or for family reunification motives, groups that tend to have more difficulties integrating to the labour market across OECD countries, has not been particularly high over the past decade (11% on average between 2005-13 compared with 18% in the OECD), but is currently increasing due to the large inflow of refugees. In particular, the total number of new asylum applications roughly tripled between 2013 and 2015 from around 7 500 to 21 000 persons per year and there is little evidence that the refugee inflow is weakening, adding to the challenge of migrants’ integration. The authorities have enacted a number of changes in response to the increase inflow of asylum seekers (described in more detail in Box 2.3).

Box 2.3. Recent legislative changes for immigration and asylum seekers

In 2015 some 21 000 persons applied for asylum in Denmark. This implies that Denmark received more asylum seekers as a share of the population than the EU average (Figure 2.6), but below the number received in other Nordic countries, and five times less than neighbouring Sweden.

Since mid-2015 a number of measures have been taken in order to curb immigration, including a reduction (up to 50%) in the benefit level for persons who had not been in Denmark seven out of the last eight years. This reduction in the benefit level was also announced through advertisements in a number of Lebanese newspapers in September 2015. The most important legislative and regulatory changes that have been passed cover a wide variety of areas, specifically targeting asylum seekers:

  • Temporary controls on the border between Denmark and Germany were introduced on 4 January 2016 in form of random checks. Police were also given new powers to suspend trains, busses, and ferries entering Denmark.

  • Waiting time for family reunification for refugees and immigrants under temporary protection (around 20% of asylum seekers in 2015) increased from 1 to 3 years. Refugees covered by the UN Refugee Convention, e.g. refugees from Syria and other groups, who due to individual circumstances are subject to prosecution in case of return, are exempted from the new rule.

  • A tightening of access to permanent residency for all immigrants. As a basic requirement, though some exemptions remain, immigrants have to have been in the country for a minimum of six years to be eligible for permanent residency (rather than five under the former rules). The number of additional conditions which must be met, such as language skills and employment effort, has been raised.

  • Increased co-payments. A cap (of 90 days) on the number of days that asylum seekers, who have sufficient means to pay for their stay in asylum centres, may be requested to pay for their stay in the centres is abolished. Furthermore, the police are given a right to search refugees (by frisk searching) and their belongings for valuables that may contribute to the financing of the stay in Denmark. Valuables that are assessed to have a value of less than approximately EUR 1 300 or that has sentimental value (such as wedding bands, medals, etc.), are exempt from confiscation. This assessment is to be made on the scene by the individual police officer, and asylum seekers have the right to dispute seizures of certain items in court.

  • A tightening of the possibility for asylum seekers to live outside asylum centres. The requirements for asylum centre housing are relaxed. This implies that refugees may be housed in tents, and other types of housing which would otherwise not used for residency under previous regulations regarding noise and size, and according to zoning laws.

  • Expanded rules for detainment. Under previous rules the police could detain foreigners, if they found it necessary in order to ensure expulsion from Denmark, for 72 hours before the person was seen by a judge. If it was not possible to get the detention upheld by a judge within the specified period, the detainee would be released. Under the new rules the police may also detain asylum seekers when entering the country, in order to determine their identity and the basis for their application. Furthermore, The Ministry of Immigration, Integration and Housing may temporarily suspend the right to have the detainment tried by a judge in less than 72 hours, in case the inflow of asylum seekers is deemed too high for this time frame to be feasible.

Figure 2.6. Denmark receives less asylum seekers than Nordic neighbours
Asylum applicants as a per cent of total population1 (January-November 2015)
picture

1. Population level as of 1 January 2015.

Source: Eurostat database.

 http://dx.doi.org/10.1787/888933352607

Integration of immigrants is an issue in many OECD countries but is more pronounced in Denmark. Gaps in employment and activity rates between immigrants and their native peers are among the highest in the OECD (see Figure 2.4, Panel A) and have widened since the 2009 crisis. Also, income inequality is particularly high between foreign-born and natives (Figure 2.7, Panel B and C). Unemployment among foreign-born is roughly twice as high as among natives. In 2013, 32.8% of unemployed migrants were out of a job for more than one year, around 8 percentage points more than the unemployed natives. Immigrants are overrepresented in low-skilled jobs, accounting for around 27% of low-skilled employment (vs 11% of total employment) and are also more likely to be poor even when working: in-work relative poverty rate is 2.7 times as high as for the natives in Denmark (Figure 2.7).

Figure 2.7. Inequality between foreign-born and native-born is high
Selected OECD countries, 2012
picture

Source: OECD Indicators of Immigrant Integration 2015.

Moreover, the gaps in employment and activity between immigrants and natives persist across generations (Figure 2.8). Integration of native-born children of immigrants – 9.7% of total youth population – is quite weak and has not improved since 2007. In most OECD countries, having foreign-born parents is associated with worse labour market outcomes, but this is the case to an even larger extent in Denmark. Like their parents, immigrants’ offspring have more difficulty finding a job than natives with equivalent qualifications, and the more qualified to a larger degree. Some 20% of native-born children of immigrants are neither in education, employment nor training, i.e. 6 percentage points more than for those with native-born parents.

Figure 2.8. Employment gap persists across generations and has widened
Difference in employment rates of native-born with foreign parents and native-born with native parents, 15-34 year olds not in education
picture

Source: OECD Indicators of Immigrant Integration 2015.

Integrating migrants and their children into the labour market would improve overall integration into the society, well-being, educational output, and reduce fiscal pressure. Due to a higher probability of being unemployed or inactive, immigrants are over-represented among benefit recipients; though only accounting for 7% of the working age population in 2014, non-western immigrants accounted for 15% of those receiving public benefits and 26% for social benefits. The probability of receiving welfare benefits decreases with the attachment to the labour market, language proficiency and the length of stay in the country (Hinte and Zimmermann, 2014). Improving labour market outcomes of immigrants to those of natives and further investing in targeted education programmes (in particular language courses) would therefore have a positive impact on their overall well-being and also benefit public finances (Box 2.4).

Box 2.4. Impact of immigration on public finances in Denmark

The net impact of immigration on the fiscal purse is small

Using a static accounting model accounting for net contributions of migrants (i.e. including tax revenues), an OECD study shows that the fiscal contribution of immigrant in Denmark is around zero and close to the OECD average (OECD, 2013a). However, on average, migrants’ net fiscal contribution is lower than the one of native born, mainly due to lower employment rates but not due to higher take up of social benefits.

Danish studies using different methodologies (e.g. dynamic general equilibrium model with overlapping generations) find a negative but again relatively small impact of immigration on the public purse. For instance, Wadensjö (2000) estimates a total net negative fiscal impact of immigrants and their children of 0.7% of GDP.*

The fiscal impact of immigration is found to vary with the region of origin, reflecting uneven performance on the labour market (Wadensjö, 2000; Hinte and Zimmermann, 2014). The net contribution is positive for immigrants from high-income countries but negative for immigrants from low-income countries. The contribution of migrants from low-income countries increases with time as integration improves but remains negative (Hansen and al., 2015).

A better integration of migrants has a positive impact on fiscal sustainability

Models accounting for the contribution of migrants over their whole lifespan conclude that no financial benefit from immigration can be expected in the short and in the long run in Denmark, notably because of the below average employment rate of migrants (Schou, 2006; Hansen et al., 2015).

At the same time, bringing the employment rate of the resident immigrant population up to the level of the native-born can improve fiscal sustainability, with a net fiscal gain estimated at 1.3% of GDP (Pedersen and Riishøj, 2008). A more recent study finds that raising employment rates of refugees to the average of other non-western immigrants would improve the primary fiscal balance by at least EUR 0.28 billion (Hansen et al., 2015).

* See also Gerdes and Wadensjö, 2006; Wadensjö, 2007; Hansen et al., 2015

Despite good access to education, the outcomes of immigrants and their children lag behind, with a below average education and skills level (OECD, 2014b). Denmark ranks among countries with the highest gaps in PISA scores, as well as a large gap in educational attainment between students with or without an immigrant background. Around 43% of young migrants who arrived as children have a low education level, more than twice as many as in an average OECD country. The coverage of early childhood education and care could be increased by bringing the attendance rate of migrants to the same level as of natives; 87% of 3-6 year-old immigrant children were enrolled in early childhood education and care in 2012, an attendance rate high by international standards, but 7 percentage points lower than among their native-born peers. Also, finding an apprenticeship contract when having an immigrant background is relatively difficult and partly explains why native-born offspring of foreign-born are overrepresented among early-school leavers. Providing more individualised training courses, implementing regular meetings with case workers as done in Switzerland and intensifying action when outcomes are bad by using multi-disciplinary teams have been identified as good practice to promoting the transition of young migrants who are neither in employment nor education or training (NEET) into employment (OECD, 2014b).

The difference in educational outcomes only partly explains the below average performance on the labour market of the migrant population (OECD, 2014b). Migrants with a high education level also have difficulties in finding a job suggesting a mismatch between their competence and the labour market needs or a lack of recognition of skills acquired abroad. Immigrants have a clear entitlement to assessment of their competences in regulated and non-regulated professions (OECD, 2014b). Public employment services have well established general procedures for assessing and validating skills, competencies, and work experience, which are placed on a digital competence card, the “Competence Card”. Social partners are involved in the certification process, thereby facilitating the recognition of qualifications in the private sector.

Integration programmes can induce lock-in effects for refugees, as they are assigned to municipalities according to quotas for the first three years while they undertake the integration programme. Refugees and immigrants who have been granted a residency are allowed to move between municipalities, but if they move before the integration programme is completed they may lose their entitlement to participating in the integration programme, the right to be assigned housing by the municipality, and the right to cash benefits or the integration allowance. Nevertheless, the entitlement to Danish language training remains, independent of whether the refugee decides to move or not. The aim of the spatial dispersal policy is to promote the integration of refugees in the Danish society by avoiding the development of ethnic ghettos and enclaves of unemployed migrants in urban areas. However, this hampers refugees’ mobility and capacity to find a job and has a negative impact on labour market performance (Edin et al., 2004). Some relaxation of the dispersal policy will come into force as a part of an integration agreement between the government and social partners, which is welcome since refugees should be able to settle in the areas where there are jobs.

Other barriers to employment include sectoral minimum wages (determined locally by social partners, at around EUR 15 per hour in 2015) that can price lower-productivity migrant workers out of employment. Migrants are overrepresented among low-wage workers and are more affected by high labour costs related to the high minimum wages. New migrants, due to a lack of country specific human capital and low language proficiency, may have a relatively lower productivity than native-born (Causa and Jean, 2007). Also, disincentives to work generated by the generous welfare system may be stronger for migrants as their average earnings prospects are much lower than for natives. Finally, differences in cultural human capital, lower access to networks as well as ethnic discrimination can hamper integration like in many countries.

To address these complex issues Denmark has implemented a panel of innovative integration policies in the early 2000s. These include language training, work placement and wage subsidies (Liebig, 2007). Past effort on integration seems to have borne fruit and should be continued to provide full potential. Employment rates of migrants improve with time spent in the country and in particular during the first five years, but at a higher pace than in the past (Figure 2.9). Recent migrants from low-income countries who benefited from integration policies have better outcomes than long-standing migrants of the same origin. Because the impact of integration policies takes time to materialise, stability of policy making and funding in this area is desirable.

Figure 2.9. Employment rates of migrants improve with time
Employment rate by duration of residence, migrants from non-western countries, in 1997 and 2012
picture

Source: Danish Ministry of Finance.

 http://dx.doi.org/10.1787/888933352612

Job training in private firms, language courses and wage subsidy programmes have had a positive impact on employment of migrants (Pedersen, 2013).

  • Wage subsidies proved particularly effective in promoting integration as they compensate employers for a lack of country specific human capital and low language proficiency (Clausen et al., 2009; Heinesen et al., 2011). Nevertheless, immigrants tend to be underrepresented among beneficiaries of the subsidised wages from which they would benefit most (OECD, 2014b).

  • Attention should be given to migrant’s access to training, in particular after the completion of the integration programme. Immigrants are participating less in training than their native peers overall. Participation in job-related training and on-the-job training is relatively lower, for women in particular, and the gap in participation in the latter is among the highest in the OECD. Also, employers’ financial contribution to job-related training is much lower for immigrants than for natives.

  • Job creation for immigrants could also be promoted with diversity plans as has been done in Belgium, encouraging diversification of recruitment and promoting mobility and training for this group. The authorities can audit and certify the company’s diversity management policies, i.e. the promotion of equal opportunity, and in some cases make recommendations on how to avoid discriminatory hiring practices.

A reform of the integration policies has been agreed by social partners and the government. Under the reform, an introduction of a new two-year integration training programme focusing on upskilling and entry into the labour market is envisaged. Such a programme will allow companies to hire persons for up to two years at apprentice wages, i.e. between 50 and 120 kroner an hour (corresponding to approximately 7 to 16 euro) that is significantly lower than the prevailing minimum wage. Companies will also be granted a 5 400 euro bonus, if they employ immigrants for more than two years. Furthermore, the reform agreement aims at standardising language training, and focusing the training more on labour market entry requirements. So far, outcomes of integration programmes have been uneven across municipalities, and monitoring of programmes has not taken place and proves difficult given low data availability (Dagpengekommissionen, 2015). An appropriate monitoring should be put in place at the local level. The quality of language courses should be better supervised as highlighted already in 2007 by the National Audit Office.

The policy focus has shifted to increasing incentives to work for immigrants and discouraging unskilled immigration. Welfare benefits for those who had lived abroad for more than one of the last eight years were approximately halved in 2015 while maintaining other support measures, such as for housing and children. Bonuses are granted to those with Danish language proficiency to encourage the acquisition of skills necessary to enter the labour market. Earlier experience with reduced migrant benefits indicates that such financial incentives will probably have only have some impact on the decision to participate in the labour market and acquire skills, particularly in the short term. A similar measure, i.e. a reduction of immigrant benefits (the Start Help) was implemented already between 2002 and 2012 (though with a smaller reduction in benefits) and has shown mixed results in increasing employment. It did have a positive effect on employment rates, which were increasing with the time spent in the country from 6 percentage points (from 8 to 14%) after 16 months to 12 percentage points (from 30 to 42%)after 52 months(Andersen et al., 2012). Other studies found a positive effect on employment rates for persons who have been in the country for a few years, but that the reduced benefit did not increase labour market participation among newcomers and did not impact the groups the most far away from the labour market, such as women and the low educated (Rosholm and Vejlin, 2010; Pedersen, 2013).

Newcomers with low employability may not have the capacity to find a job even if economic incentives are in place and thus face higher poverty risk. Indeed, poverty increased significantly among the “Start Help” recipients (Andersen et al., 2012). Decreased benefits specifically for migrants are likely to increase inequality between foreign-born and native-born in terms of poverty risks, income and over-qualification from current high levels (see Figure 2.7), as well as to reinforce negative stereotypes on the low willingness of migrants to integrate (Liebig, 2007).

Easing expenditure pressures on social institutions

Thanks to prudent fiscal rules, public finances are deemed sustainable and to comply with EU fiscal requirements and the Budget Law (allowing for a maximum structural deficit of 0.5%) in the long run. According to national projections, public debt – at 45.2% of GDP in 2014 – will remain below 60% of GDP. The S2 sustainability indicator, i.e. the primary deficit needed to stabilise the debt to GDP ratio by 2060, has been estimated at 0.3% of GDP by the Danish Economic Council- the Danish “fiscal watchdog” – in spring 2015 (Danish Economic Council, 2015a).

National projections for the long-term sustainability of public finances are based on a number of assumptions including balanced budget by 2020 and an increase in labour participation in response to structural reforms. In particular, pension reforms are expected to compensate the impact of ageing on the labour force (Figure 2.10). Productivity is also expected to grow by 1.2% in 2020, well above the average growth over the past 10 years, and public spending per capita is to remain stable. Uncertainty surrounding these assumptions is high, notably on future developments of labour force participation (Box 2.5) and healthcare costs. Also, as in most OECD countries, national projections do not take into account contingent liabilities. These are government guarantees, i.e. liabilities related to public-private partnerships and liabilities of government-controlled entities, both classified outside general government balance sheet, estimated at 36.7% of GDP in 2013 (European Commission, 2015b).

Figure 2.10. Structural reforms1 to help cope with ageing according to national projections
picture

1. These mainly refer to the pension reforms of 2006 and 2011, implying an indexation of the statutory retirement age to longevity, and a reduction in the use and duration of the early retirement scheme.

Source: Danish Ministry of Finance.

 http://dx.doi.org/10.1787/888933352627

Box 2.5. Assessing the impact of reforms on potential growth and labour force

OECD and European Commission (EC) projections for potential and labour supply growth in Denmark are quite close, but the estimated impact of reforms on labour supply varies widely between the two institutions (cf. Table 2.1). The impact of pension reforms on old age participation is more than six times higher in percentage points in the EC projections compared with the OECD projections. This is due to differences in the methodology used and highlights the uncertainty related to the projections.

Table 2.1. Projections for labour supply of the Danish government, OECD and European Commission
(percentage point increase between 2013 and 2060)

Government

OECD

EC

Total change

 

 

Participation rate (15-74)

7.7

4.0

4.6

Participation rate (15-64)

2.5

4.6

2.2

Participation rate older (55-64)

13.8

15.3

12.7

Impact of pension reform

Participation rate total (15-74)

7.4

0.7

6.1

Participation rate total (15-64)

1.9

0.5

2.9

Participation rate total (55-64)

10.1

2.4

14.3

Effective retirement age 2060

71

66

68

Source: OECD calculations based on the OECD EO95 and European Commission (2015), Danish Ministry of Finance.

Both institutions are more conservative than the government on the impact of the reforms on labour supply and thus GDP growth. The EC takes into account the uncertainty regarding the implementation of the indexation of the pension age to life expectancy. The OECD measures the impact of pension reforms based on econometric estimates, and an assumption of an indexation of the statutory retirement age, which keeps the proportion of years in retirement fixed as a share of expected life span.1 In this respect the OECD estimates differ from national estimates, as the Ministry of Finance assume a gradual decline in the years in retirement towards 14.5 years.

1. For more details, see Johansson et al. (2013).

In the medium run, complying with fiscal rules will be challenging. According to national projections structural deficits are set to temporarily exceed 0.5% of GDP after 2030 and to reach 1.5% of GDP by 2045 according to government projections (the so-called “hammock challenge”). This period of public deficits results from the fact that the indexation of the statutory retirement age to life expectancy gains will be phased-in, and will therefore only gradually catch up with developments of longevity. Such temporary deficits could be addressed in a number of ways that are, however, not without downsides. Firstly, increases in the retirement age could be speeded up. This would necessitate another change in pension legislation, and might prove politically difficult. Secondly, tax revenues from future pension taxation could be brought forward, as done in recent years. Pensions are taxed when retirement is reached, however such measure includes the risk that tax revenues are not used to reduce public debt and that some economic losses could occur, as the government might not have the same rate of return on savings than private pension funds. A third option is to run down public debt through pre-saving future age-related spending, as done for instance in Belgium, but this would weigh on economic activity. Lastly, revising fiscal rules, more specifically how deficits and debt are measured to take into account deferred taxes on pension savings, can be difficult to implement given the large uncertainty regarding the level of such tax revenues. All in all, although the “hammock challenge” is a temporary issue, it highlights the challenges future governments face in case employment and longevity do not develop according to projections, underlying the need for further action in order to ensure the long term sustainability of public finances.

Further improving the performance of the pension system

The pension system is effective in ensuring high replacement rates and limiting old-age poverty. Combining a taxed-financed defined benefit pension and mandatory occupational defined contribution schemes, it allows diversification of pension risks between idiosyncratic risks on investments of pension funds and macroeconomic risks of national pension schemes (Box 2.6). To maintain its sustainability and its generosity in spite of population ageing, major reforms have been legislated, including the indexation of retirement age to life expectancy gains from 2030 onwards. Reforms of early retirement programmes (notably a progressive increase in early retirement age from 60 in 2013 to 63 in 2019) have already shown an impact on employment rates of workers above 60.

Box 2.6. The Danish pension system: main features and reforms

The pension system

The Danish pension system consists of a mix of pay-as-you-go and defined contribution schemes:

  • A tax financed and means-tested public pension, including a basic flat amount (around 17% of the average wage, reduced if earned income exceeds 75% of the average wage), a pension allowance and a supplementary benefit for the most disadvantaged.

  • A mandatory defined contribution scheme (ATP) for employees who work more than 9 hours a week and those receiving temporary social security benefits. A voluntary defined contribution scheme (SUPP) is also offered to those outside the labour market.

  • Quasi-mandatory fully funded occupational schemes covering around 90% of full-time employees. Most of them are defined contribution schemes with contribution rates ranging from 12% to 18%.

  • Individual voluntary defined contribution schemes.

Adequacy

The pension system is highly redistributive and the replacement rate is 77% on average in 2012 – and higher for public benefit recipients and low income earners (Figure 2.11). Old age poverty rate is 5%, less than half the OECD average, constituted mainly by “fraction” pensioners, i.e. pensioners who have not lived 40 years in Denmark and are not entitled to the full public pension.

Figure 2.11. Gross pension levels and replacement rates are high
Expected pensions for a 20 year old entering the labour market in 2014 retiring at 67 years
picture

Source: OECD, Pension at a Glance 2015.

 http://dx.doi.org/10.1787/888933352631

Recent reforms

With the 2006 and 2011 pension reforms, the statutory retirement age will gradually increase from 65 to 67 in 2022 and to 68 years in 2030. From then on it will be adjusted to life expectancy gains every five years (within a limit of one year and conditional on approval by the Parliament). The average pension payment period should then be reduced to 14.5 years. The duration of voluntary early retirement programmes is also progressively shortened from five to three years with the minimum age raised in line with the statutory retirement age. Overall, the Danish authorities expect the pension reforms to result in a gradual increase in the effective retirement age from around 64 years today to 70 years in mid-2040. If implemented as planned, this will imply that a person entering the labour market at 20 in 2014 will reach the legal retirement age at 73.5. This would make the Danish pension age the highest among OECD countries (OECD, 2015e).

Following the increase in the statutory early retirement age, the employment rate of the age group 60-64 has improved significantly. Like in many other OECD countries, even when increments are provided for longer careers, employment rates tend to fall sharply at the legal retirement age. For the cohort of 60 year olds, affected by the reform, the employment rate increased by more than 6 percentage points, well above what is observed for other cohorts (Figure 2.12). Going forward, the employment rate for persons 61 and older is expected to increase progressively as well, as the reform affects more cohorts.

Figure 2.12. The pension reforms have already an impact on employment rates of older workers
picture

Source: Danish Ministry of Finance.

 http://dx.doi.org/10.1787/888933352649

Participation in the public early retirement scheme has significantly declined. The reform of early retirement in 1998 raised the costs of opting-in in the scheme; measures in 2012 made possible to withdraw previous contributions at a tax discount and changed participation in the scheme from an opt-out system to an opt-in system. The share of 30 to 59 year olds who participate in the scheme has declined from more than 80% in 2000 to less than 30% in 2014. For the 30 to 34 year olds the decline has been more pronounced, indicating a progressive phase out of early retirement in the future.

In 2012, public pensions accounted for around 70% of total gross pension payments and around 90% of individuals older than 65 received means-tested pensions. As private pension schemes introduced in the 1990’s mature, public spending on pension is projected to decline from around 10% of GDP in 2013 to 7% towards 2060 (European Commission, 2015a) together with the share of pensioners relying on means-tested benefits. Also, taxation of private pension savings, both ongoing taxation of returns in pension funds and when they are paid at retirement (payments into the scheme are tax exempt), will increasingly account for a non-negligible share of public revenues which is set to absorb around 80% of the increase in age-related public spending by 2050 (Hansen et al., 2015). As taxes on pension savings increase as a share of total revenues, the volatility of revenues also increases. This poses a challenge for estimating future public revenues and the year-on-year compliance with fiscal rules.

In addition to the uncertainty about the revenues development, spending pressures could be higher than expected. The pension system creates some disincentives to save for workers earning relative low wages, particularly in the years just before the retirement age is reached. Due to means-testing of public pension, income from the occupational funded scheme reduces public pension benefits and housing subsidies and thereby the effective rate of return of private pension savings (Figure 2.13). It is uncertain if and how these economic disincentives will impact retirement decisions for low income earners in the future, as the second pillar matures, but it is likely that this group will retire as early as possible.

Figure 2.13. Pension savings of low-income earners are heavily taxed
Marginal effective tax rates on private pension savings, 2015
picture

Note: Stylized calculation for a single pensioner living in a rented accommodation and receiving means-tested rent subsidy.

Source: Danish Ministry of Taxation.

 http://dx.doi.org/10.1787/888933352659

In principle, means-tested benefits should provide less revenue than work-based pension income. There are number of options to reinforce income gains of staying in employment. These include removing mean-testing of the basic public pension, introducing a tax relief on pensions for low wage earners after retirement, tax exemption on contribution to the occupational scheme during the last years before retirement, excluding late contributions from the income base used from means-testing, introducing compulsory savings or reducing the rate of withdrawal of mean-tested pension allowance. The costs and benefits of these options have to be evaluated.

The marginal effective tax on pension returns is higher than on investment in property, also due to the means-testing of other benefits, and can discourage groups not covered by collective agreements from saving in pension funds. A special tax on large pension pay-outs can also weigh on voluntary savings in pension funds. In addition, the sophistication of the system – the interaction between the different schemes but also difficult access to information on the different tax rates applied to different types of savings – could hamper rational decisions about the right amount of pension savings. The complex incentives to save and the pension system’s interaction with the tax system should be also investigated and analysed.

Over the past five years, around 24% of those aged 30-59 were saving less than 5% of their income in pension schemes and 12% had no savings at all (mainly self-employed and benefit recipients). They are likely to face low replacement rates from the second pillar and be eligible to means-tested pensions. Thus, the pension entitlement of individuals not covered by occupational pensions should be closely monitored and initiatives to increase the pension savings among this group should be introduced. Indeed, mandatory contributions are more effective than subsidies at increasing savings of those who tend not to save, not least because 85% of savers are unresponsive to subsidies and those using subsidies are those who already save for retirement (Chetty et al., 2014).

A sound prudential regulatory framework for the second pillar is in place to protect pensioners and avoid excessive risk taking. Nevertheless, the level of the rate of return of private pension remains uncertain and poses a risk for individuals with limited adjustment capacity. Most of the occupational schemes are defined contribution schemes and the return on pension savings depends on market outcomes. Should the level of private pensions be too low, pensioners will be eligible to tax-financed mean-tested supplements pensions, resulting in the ageing cost partly transferred from pension schemes to social safety net budgets. Thus, investment strategies should ensure exposure to risk declines with age as capacity to adjust also declines with age. Though this option (so-called life-cycle investment strategies) exists, it is voluntary. Moreover, in spite of an overall top evaluation, the Melbourne Mercer Global Pension Index points to an insufficient protection of pensioners in case of fraud, mismanagement and provider insolvency (Mercer, 2015). Also, a requirement of providing an annual report on the pension plan to individual plan holders does not include information on major investment and allocation of assets in asset classes that would allow future pensioners to better assess the risk related to their pension savings.

Increasing effective retirement age and prolonging working lives

According to national plans, the sustainability of the public pension system is ensured by future increases in the effective retirement age initiated by the pension reforms. The indexation of retirement age to life expectancy gains, which aims at reducing the average period in retirement to 14.5 years, is likely to increase labour market participation of older generations but might not be sufficient to sustain the labour force as planned. Furthermore, should life expectancy increase faster than currently projected, the existing rule for adjusting the statutory retirement age (i.e. that the retirement age can only be raised by one year every five years) will be insufficient to keep the average years spent in retirement at 14.5 years. In this case, policy makers should open up for a possibility of a faster than expected adjustment in the retirement age.

Indexation of the legal retirement age to life expectancy is subject to a political decision. The increase in retirement age has to be legislated by Parliament every five years. Political pressures not to reduce the retirement period are likely to intensify in the future as the retirement age increases (under the current national projections it will reach 70 in 2040) and if social inequality in life expectancy continues to broaden. Life and health expectancy have increased but the gap between low and highly educated workers has widened, pointing to increasing inequality in life expectancy among pensioners (Brønnum-Hansen and Baadsgaard, 2008). At the same time, the gap in pension payments between the persons with the highest and the lowest incomes has narrowed between 2000 and 2013, indicating that inequality between pensioners has decreased as private pensions schemes have become more widespread (Based on data provided by the Ministry of Finance).

Also, older workers might not fully adapt to longer working lives, in particular in physically-demanding occupations (Wubulihasimu et al., 2015). A large number of workers in occupations that require good health condition are likely to spend years in disability before reaching the statutory retirement age. In the construction sector in Sweden, for instance, 80% of lost years of working life are among men who retired due to disability at ages above 60 (Järvholm et al., 2014). As a consequence, participation in early retirement programmes, old-age unemployment and reliance on public pensions can rise in the future, inducing additional cost for tax payers and a lower-than-expected contribution base. Another question mark is to what extent occupational differences in the risk of disability pension will increase socioeconomic inequalities in the old-age population.

The effective extension of working lives will depend on the employability of older workers and therefore, improving their working opportunities should be a priority. Remaining regulatory barriers to employment of older workers should be removed and initiatives to adapt the working environment to an ageing work force developed. Most collective labour agreements include a framework on specific schemes for older workers and measures to promote age diversity but only few firms effectively implement them. Older workers can use part of the occupational pension to reduce working hours (“senior freedom”) but there is no evidence on the impact of this measure on the retention of older workers and its cost-effectiveness, among other things, as the use of the scheme is not very widespread. The OECD 2015 review Ageing and Employment Policies, Denmark includes a number of recommendations on how to effectively extend working lives (OECD, 2015b), notably offering flexible working hours and adapted work tasks, reducing seniority elements in wage setting, implementing programmes for early identification and prevention of work-related health issues and removing remaining mandatory retirement ages.

Longer working lives have been supported by reducing generosity of voluntary early retirement programmes (VERP) and unemployment benefits for those above 55. Nevertheless, reforms have not fully translated into higher participation, suggesting other pathways to early retirement have been used (Halvorsen and Tägtström, 2013). Older workers are still over-represented among long-term unemployed (40% compared to 30% for those aged 24-54 in 2013) and disability benefits recipients (e.g. in flex-jobs). They have a quite low level of return to employment after a period of receiving unemployment benefits as they are eligible to a special allowance when they exhaust their rights. Senior jobs are also offered in the municipal sector at a normal wage to all unemployed entitled to a VERP whose unemployment benefit has expired (OECD, 2015b). The duration of these jobs and the wage level should be reduced to avoid lock-in effects and disincentives to remain out of the ordinary labour market. Also, this scheme should be targeted to those with low employment prospects. Disability benefits are another potential pathway to early retirement (see above).

Better controlling healthcare and long-term care costs

Spending on health and long-term care (LTC) – high by international standard – is projected to continue to increase with population ageing (Figure 2.14). Projections are quite uncertain and past experience proved that such spending is difficult to control. Public spending on pharmaceuticals in hospitals increased significantly over the past years (from 5.6% of total health spending by regions in 2007 to 8.3% in 2014), partly because costs are not taken into account in the choice of treatments and no prioritisation mechanism is in place to limit spending pressures in this area.

Figure 2.14. Spending on health and long-term care is high
In 2013 or latest year available
picture

Source: OECD Health database.

 http://dx.doi.org/10.1787/888933352664

The Danish government, the European Commission and the OECD have quite different estimates of future age-related spending (Table 2.2). The spectrum of increases in healthcare spending in OECD projections for Denmark is large, varying from 2% to 5.9% of GDP by 2060 depending on the scenarios. Great uncertainty also surrounds projections for LTC spending, which differs by factor 5 across institutions and scenarios. Future age-related costs will depend widely on the health status of the elderly population and on changes in relative prices of medical goods. Half of the spending variation between 1990s and late 2000s is explained by changes in prices due to innovation which generates more costly medical treatments and products (De la Maisonneuve and Oliveira-Martins, 2013). Due to the large uncertainty that surrounds both health care and long term care projections, measures to improve efficiency are welcome. Disparities across local governments in terms of outcomes and efficiency suggest the scope for improvement is quite large, in particular in healthcare and long-term care (Medeiros and Schwierz, 2015; Heijink et al., 2015).

Table 2.2. The range of healthcare and long-term care spending projections is large
Increase in percentage points of GDP by 2060

OECD

Government

EC

Health care

Average

-

0.0

+0.9

Best scenario

+2.0

+0.1

Worst scenario

+5.9

+3.1

Long-term care

Average

-

+1.4

+2.0

Best scenario

+0.6

+1.9

Worst scenario

+1.0

+3.0

Source: De la Maisonneuve and Oliveira-Martins (2013), European Commission (2015a), Danish Ministry of Finance.

Improving the provision of welfare services

Achieving efficiency gains in the provision of public services would limit increases in age-related costs and strengthen the robustness of the welfare system. Disparities across local governments in terms of outcomes and efficiency analyses suggest the scope for improvement is quite large in Denmark, in particular in the healthcare and the long-term care sectors (Medeiros and Schwierz, 2015; Heijink et al., 2015; Wittrup et al., 2013). Life expectancy is relatively low given the level of spending per capita on health (Figure 2.15, Panel A). Also, international comparison, notably with peer countries, indicates that the large coverage of LTC could be reached at a lower cost (Figure 2.15, Panel B). Denmark has been successful in preventing costly long-term institutional care and older people tend to receive care in their homes. Nevertheless, the unit cost of institutional care is higher in Denmark than in Finland and Sweden (European Commission, 2014).

Figure 2.15. Large efficiency gains are possible in the healthcare and the long-term care sectors
picture

Source: OECD Health database; European Commission (2014).

 http://dx.doi.org/10.1787/888933352679

Reaping more benefits from territorial reform

The 2007 territorial reform has significantly restructured the administrative landscape and led to major changes in the relations between central and local governments (Vammalle et al., 2014). This went in the direction of larger local governments to allow for economies of scale but also of less autonomy for the local administrations (OECD, 2015c). Administrative costs have been significantly reduced but the reform has not triggered efficiency gains in all sectors, in particular not in day care and elderly care (Blom-Hansen et al., 2015). Further gains of decentralisation could be reaped by granting greater autonomy coupled with strong accountability to local government (Sow and Razafimahefa, 2015).

Indeed, limiting autonomy of local government in assigning qualitative and quantitative objectives would risk undermining operational freedom and capacity to adjust. Besides, Danish public employees and organisations have called for a reduction in rules and regulation – a “Trust Reform” of the public sector (Sørensen, 2015). The Productivity Commission also recommended giving more room of manoeuvre to local government by removing unnecessary rules.

During 2012 to 2015 nine local municipalities were granted some exemptions from government rules and documentation requirements in order to test new ways of solving their tasks, in a policy experiment known as the “Free Municipality” initiative. The main focus has been on simplification, innovation, quality and a more inclusive approach to the individual citizen, with many of the experiments focusing on the employment effort. The Free Municipality experiment is currently being evaluated, in order to form the basis for potential future legislation on de-bureaucratisation for all municipalities. The concept of Free Municipalities continues in an adjusted form until 2019, and is extended to more municipalities.

Well defined targets and sanctions helped to keep spending under control at the local level, but the impact on efficiency is unclear. Since 2011, sanctions are imposed on local government which do not comply with budget requirements. In case of non-compliance, poor performance is usually made public, occasionally intensive monitoring is put in place, and the budget is always reduced. Such rules have induced a reduction in spending exceeding budget requirements and the build-up of precautionary budget in municipalities. The question is whether precautionary savings are desirable and funds could not have to be used in a more productive manner. Furthermore, uncertainty on the level of subsidy paid by the central government to the local one hampers the effective management of the budget and cash flows (National Audit Office, 2014). The use of grants to sub-national governments to too specific purposes should continue to be limited, in order to allow more operational freedom and capacity to adjust spending priorities while making sure that appropriate output performance monitoring and minimum service delivery regulation remains in place.

The provision and the quality of welfare services differ widely across municipalities, e.g. in terms of the number of benefit recipients and the implementation of programmes. Improved quality monitoring and performance management can help to identify the sources of these local discrepancies (including differences in work environment and conditions) and best practices. In secondary care, quality monitoring is well developed but has not systematically induced improvement efforts and is not fully used for peer reviews (OECD, 2013b). Data collected should be used to evaluate the impact of financial incentives in hospitals (i.e. to achieve 2% productivity gains per year) on efficiency, notably on over-prescription. In primary and long-term care, there are only few mechanisms to assess and reward the quality of services, and already available data on performance could be better utilised (OECD, 2013b). In the LTC sector, stricter budget constraints to municipalities led a process of retrenchment and restructuring and the impact on the quality and the adequacy of services remains unknown (Kvist, 2014).

Developing an integrated welfare system, prevention and rehabilitation

Integrating the delivery of welfare services – healthcare, LTC and social services – would allow for an early identification of multidimensional needs, improve information sharing between providers, avoid overlaps and duplication, and improve access to services (OECD, 2014d). In particular, a better coordination of care between municipalities, general practitioners, and hospitals would ensure the more cost-effective option is envisaged and prevent unnecessary and expensive hospital admissions. However, transition costs for an integrated system could be elevated as integration requires financial inputs to accompany the restructuration, creating incentives to collaborate and monitoring.

In the healthcare sector, municipalities have been pressed to ensure a better correction between primary and secondary care, to encourage general practitioners (GPs) and hospitals to offer a coordinated package of treatment. Nevertheless, mechanisms to reward the continuity of care are not well developed and incentives for GPs to participate in integrated care pilot projects had a quite low up-take. Better assessing the quality of primary care, publishing performance indicators and giving GPs more responsibility for outcomes across the whole patient pathways would support the transition towards an integrated system (OECD, 2013b). Financial incentives in the form of lump sum for those participating in Disease Management Programmes (DMPs) or a joint payment to the care providers for chronic diseases can also be provided as done in Germany and in the Netherlands.

Many key disease areas and the need for long-term care can be substantially reduced through appropriate prevention and rehabilitation programmes (OECD, 2015c). In particular, prevention targeted on main causes of morbidity or disability at a relatively early stage could support healthy ageing, self-care and home-based services and reduce disability in old age with a significant impact on public finances. Were the longevity gains not to translate into more years in good health, public healthcare spending would increase by around 2 percentage points of GDP from 12.2% to 13.4% (De la Maisonneuve and Oliveira-Martins, 2013).

Since 2007, Danish municipalities have been asked to place more emphasis on prevention and rehabilitation. However, expenses on prevention have not increased as a percentage of GDP over the past decade (Figure 2.16). Also, facilities for providing necessary rehabilitation services were not in place five years after this competence was assigned to municipalities (Olejaz et al., 2012). Financial incentives for municipalities that are successful at rehabilitation and reducing institutionalisation should be strengthened, as done in Germany for the sickness funds. In addition, national targets on alcohol, tobacco and obesity should be set and actions targeted to the most vulnerable. Low educated and low-income earners are more exposed to unhealthy lifestyles and because of geographical access constraints they use less preventive services (Ellins et al., 2014).

Figure 2.16. Spending on health prevention is largely unchanged
picture

Source: OECD Health database.

 http://dx.doi.org/10.1787/888933352685

Developing tax instruments to limit health-damaging behaviours should also be considered as it proves quite effective in improving health outcomes (Sassi et al., 2013). Denmark already imposes excise duties on unhealthy items such as alcohol, tobacco and sugar, but could consider whether more is possible. In 2011 a tax on saturated fat was introduced, but was abolished in 2012, due to arguments that it posed an excessive administrative burden, induced cross-border trading, and potentially introduced barriers to EU single market. Likewise, a plan to expand the chocolate and sugar tax to other sugary foods (e.g. yoghurt, ketchup, jam) was scrapped in 2012.

Increasing competition and private funding for welfare services

Only a small share of welfare services are contracted out to private providers and the state of competition for public contracts varies widely among municipalities (Sørensen, 2015; OECD, 2015d). In 2008, it was reported that leading municipalities opened 35% of their procurements to competition, while in other municipalities less than 20% of procurements were made competitively (OECD, 2015d). Opening welfare services further to competition requires imposing standards and closely monitoring outcomes to ensure high quality services. Related efficiency gains can thereby be small because of induced administrative costs. Empirical analyses find a positive impact of marketization in technical areas (waste collection, road maintenance, cleaning) but evidence is mixed for welfare services (Peterson and Heljmar, 2014). Some studies point to cost savings and higher quality in the healthcare sector as a result of competition exposure, in particular when associated with free choice of providers and dissemination of information on their performance (OECD, 2015c). At the same time, the literature does not identify a one-size-fits-all set of policies to increase efficiency by competitive pressure and the modalities should be assessed on a case by case basis.

Enforcement of competition law is limited in the Danish public sector. Central and local governments can provide ad hoc exemptions to the operation of the Competition Act for practices deemed being “direct or necessary consequence of public regulation” without having to find alternative means that do not produce anticompetitive effects. There are no implementation guidelines or regulation for applying these exemptions. Clear standards should be defined and the use of competition exemptions should be made or assessed by an independent body. At minimum, the Competition Council should be in a position to seek a judicial review of the administrative decision to override the application of the Competition Act (OECD, 2015d).

Competition could be extended in the LTC and the healthcare sectors.

  • Public spending on LTC is mostly provided in cash (45% vs. 28% on average in the EU) and people eligible for home care can choose their provider. Diversity is limited however for nursing homes: independent institutions make up only 11% of the market.

  • In the healthcare sector, secondary care is concentrated in public hospitals. Patients are allowed to choose private providers only if the public sector is not able to provide care within a guaranteed period. While extending public procurement to private providers can lead to savings for certain services (e.g. for scans; Holstein, 2014), the development of private hospital activity remains limited. Also, contract prices for hospital medicine are confidential and practitioners cannot choose the most cost-effective treatments (Hostenkamp, 2012). Transparency on prices and an obligation to prescribe the cheapest treatment can help to contain costs.

  • In the retail sector, regulation for pharmacies creates barriers to competition. Entry and ownership are restricted, prices and profits are controlled (e.g. a collective financial equalisation system requires pharmacies with above-average turnovers to contribute to pharmacies with below-average turnovers). The number of pharmacists per inhabitant in Denmark is low compared with other OECD countries (Figure 2.17). Relaxing conditions for establishing internet pharmacies, for multiple ownerships, and for opening hours, removing the limit on the number of pharmacies and the equalisation scheme as recommended by the Competition Council would improve access to care, and foster productivity gains in the pharmaceutical sector.

Figure 2.17. The retail market for pharmaceuticals is more concentrated than in most OECD countries
Number of practising pharmacists per 1 000 inhabitants in 2013 or latest year available
picture

Source: OECD Health database.

 http://dx.doi.org/10.1787/888933352699

Increasing private funding from the current low level would limit overreliance on tax-financed public budget for the provision of healthcare and LTC services. In Denmark, most of the healthcare is provided for free, except for in particular pharmaceuticals, dental services and glasses. As a result, out-of-pocket payments (OPP) account for around 13.7% of total health spending (vs 19.6% on average in the OECD, Figure 2.18, Panel A). Extending co-payments would limit demand without adverse effect on access to medical services and thereby on health status if kept at relatively low levels. Excluding high value health care and introducing a ceiling on OPP can limit the negative impact of cost-sharing on health outcomes on the poorest or the sickest (Pisu, 2014).

Figure 2.18. Co-payments are quite low in the healthcare and the long-term care sectors
Out-of-pocket expenditure on health and long-term care in 2013 or latest year available
picture

Source: OECD Health database.

 http://dx.doi.org/10.1787/888933352708

In the LTC sector, municipalities have the legal obligation to offer a range of services to elderly in need. They have the freedom to make the decisions about the amount and type of services, based on concrete assessment of each individual case (European Commission, 2011). Some rationing of care is already in place, and since January 2015 municipalities are obliged to offer an individual rehabilitation and training to make the elderly more self-sufficient, before granting home care. This is both aimed at promoting independence and quality of life, and to limit the strain on municipal budgets from an aging population.

Home care services are provided for free and municipalities have only a limited freedom to charge for home help and non-health related expenses. As a result, OPP payments account for 8.6% of total spending on long-term care, 8 percentage points less than on average in the OECD (Figure 2.18, Panel B). Consideration should be given to the role OPPs could play in increasing well-being for patients with LTC needs. At the same time, national eligibility criteria for different types of home care should be elaborated, to ensure that the elderly get a comparable level of services across municipalities.

Recommendations to sustain effective social institutions

Making work more attractive for benefit recipients

Key recommendations

  • Reduce the effective taxation of returning to work by reforming benefits in order to make work pay. Allocate a permanent disability pension only to those with permanent incapacity to work regardless of age.

  • Improve the quality and the implementation of integration programmes for migrants.

  • Monitor work requirements for social assistance recipients.

Other recommendations

  • Review existing exceptions, and reassess eligibility to disability benefits on a regular basis.

  • Assess the lifetime pattern of benefit recipients and the costs of their inactivity.

  • Promote diversity in firms with information campaigns, audits on management policies, programmes for early identification and prevention of work-related health issues.

Addressing expenditure pressure in social institutions

Key recommendations

  • Eliminate disincentives and barriers to work at older ages originating from public and occupational pension design.

  • Use autonomy of local governments better for drawing lessons from the diversity of approaches in providing social services, which should also strengthen accountability for outcomes. Integrate welfare, prevention and rehabilitation services to improve provision of public services.

  • Make general practitioners more responsible for continuity of care.

  • Encourage the private supply of welfare services to offer more choice while at the same time reaping efficiency gains.

Other recommendations

  • Implement initiatives to increase the pension savings among individuals not covered by occupational pensions.

  • Support competition on welfare services, notably by revising exemptions to the Competition Act in the public sector and relaxing regulation in the pharmaceutical sector.

  • Improve employability of older workers and develop initiatives to adapt working environment to an ageing work force.

  • Consider diversifying funding sources of healthcare and long-term care by raising co-payments, mean-testing public support for home care, and expanding “sin taxes”.

Bibliography

Andersen, L., H. Hansen, M. Schultz-Nielsen et al. (2012), “Starthjælpens betydning for flygtninges levevilkår og beskæftigelse”, Working paper No. 25, The Rockwool Foundation Research Unit.

Blom-Hansen, J., K. Houlberg and S. Serritzlew (2016), “Elusive Scale Effects: Jurisdiction Size and the Costs of Providing Public Services”, forthcoming.

Brønnum-Hansen, H. and M. Baadsgaard (2008), “Increase in Social Inequality in Health Expectancy in Denmark”, Scandinavian Journal of Public Health, No. 36, pp. 44-51.

Causa,O. and S. Jean (2007), “Integration of Immigrants in OECD Countries: Do Policies Matter?”, OECD Economics Department Working Papers, No. 564, OECD Publishing, Paris, http://dx.doi.org/10.1787/162367775052.

Causa O. et al. (2016), “Inequality in Denmark through the Looking Glass”, OECD Economics Department Working Papers, forthcoming.

Chetty, R. et al. (2014), “Active vs. Passive Decisions and Crowd-Out in Retirement Savings Accounts: Evidence from Denmark”, The Quarterly Journal of Economics, Vol. 129 (3), pp. 1141-1219.

Clausen, J. et al. (2009), “The Effect of Integration Policies on the Time Until Regular Employment of Newly Arrived Immigrants: Evidence from Denmark”, Labour Economics, Vol. 16, Issue 4.

Dagpengekommissionen (2015), “Frivillig eller obligatorisk forsikring mod ledighed”, Working Paper, October.

Dansk Arbejdsgiverforening (2015), “Work Life, Safety and Health”, Labour Market Report 2014.

Danish Economic Council (2014), “Effekter af en toårig dagpengeperiode på beskæftigelse og ledighed: En foreløbig evaluering af dagpengereformen”, De Økonomiske Råds Arbejdspapir, N° 2014:03.

Danish Economic Council (2015a), “Dansk Økonomi, efterår 2015”, English Summary, Autumn

Danish Economic Council (2015b), “Dansk Økonomi, forår 2015”, English Summary, Spring.

Danish Ministry of Finance (2015), “National Reform Programme 2015”, March.

Diop-Christensen, A. (2015), “Is ‘Making Work Pay’ Effective for the ‘Unemployable’? The Impact of Benefit Sanctions on Social Assistance Recipients in Denmark”, Journal of European Social Policy, Vol. 25(2), pp. 210-224

De la Maisonneuve, C. and J. Oliveira-Martins (2013), “A Projection Method for Public Health and Long-Term Care Expenditures”, OECD Economics Department Working Papers, No. 1048, OECD Publishing, Paris, http://dx.doi.org/10.1787/5k44v53w5w47-en.

Edin, P.-A., P. Fredriksson and O. Aslund (2004), “Settlement Policies and the Economic Success of Immigrants”, Journal of Population Economics, Vol. 17(1), pp. 133-155.

Ellins, J. et al. (2014), “International Responses to Austerity”, Evidence Scan No. 22, Health Foundation, Birmingham.

European Commission (2015a), “The 2015 Ageing Report: Economic and Budgetary Projections for the 28 EU Member States (2013-60)”, European Economy 3, May 2015, Brussels.

European Commission (2015b), “Contingent Liabilities”, Brussels.

European Commission (2014), “Identifying fiscal sustainability challenges in the areas of pension, health care and long-term care policies”, European Economy, Occasional Papers 201, October.

European Commission (2011), “Closing the gap – between expanding care needs and limited resources”, European Economy, Peer review, October.

Gerdes, C. and E. Wadensjö (2006), “Immigration and the Welfare State: Some Danish Experiences”, AMID Working Paper Series, No. 60/2006.

Halvorsen, B. and J. Tägtström, (2013), “Det dreier seg om helse og arbeidsglede: Om seniorer, arbeid og pensjonering i Norden”, Nordic Council of Ministers, Copenhagen.

Hansen J., S. Hougaard Jensen and P. Stephensen (2015), “Occupational Pensions, Aggregate Saving and Fiscal Sustainability in Denmark”, PeRCent’s First Annual Conference, “Taking the Danish Pension System to the Next Stage”, held at Copenhagen Business School, June 16.

Hansen M., M. Schultz-Nielsen and T. Tranæs (2015), “The Impact of Immigrants on Public Finances: A Forecast Analysis for Denmark”, IZA Discussion Paper, No. 884, February.

Heijink R. et al. (2015), “Comparative Efficiency of Health Systems, corrected for Selected Lifestyle Factors”, Funded by the European Commission and carried out by the Dutch National Institute for Public Health and the Environment (RIVM).

Heinesen, E., L. Husted and M. Rosholm (2011), “The Effects of Active Labour Market Policies for Immigrants Receiving Social Assistance in Denmark”, IZA Discussion Papers, No. 5632.

Hinte, H. and K. Zimmermann (2014), “Does the Calculation Hold? The Fiscal Balance of Migration to Denmark and Germany,” IZA Policy Paper, No. 87.

Holstein, M. (2014), “Regionernes udbud af MR-Scanninger”, CEPOS, Copenhagen.

Hostenkamp, G. (2012), “The Market for Hospital Medicine in Denmark”, Nordic Journal of Health Economics, Vol. 1(1), pp. 57-71.

Huefner F.and C. Klein (2012), “The German Labour Market: Preparing for the Future”, OECD Economics Department Working Papers, No. 983, OECD Publishing, Paris, http://dx.doi.org/10.1787/5k92sn01tzzv-en.

Immervoll, H. (2012), “Reforming the Benefit System to ’Make Work Pay’: Options and Priorities in a Weak Labour Market”, IZA Policy Paper, No. 50.

Immervoll, H., S. Jenkins and S. Königs (2015), “Are Recipients of Social Assistance ‘Benefit Dependent’? Concepts, Measurement and Results for Selected Countries” OECD Social, Employment and Migration Working Paper, No. 162, OECD Publishing, Paris, http://dx.doi.org/10.1787/5jxrcmgpc6mn-en.

Jamet, S. and V. Koen (2014), “Making the Most of Skills in Denmark”, OECD Economics Department Working Papers, No. 1119, OECD Publishing, Paris, http://dx.doi.org/10.1787/5jz2qt4crff8-en.

Järvholm B. et al. (2014), “Heavy Work and Disability Pension – A LongTerm FollowUp of Swedish Construction Workers”, Scand J Work Environ Health, Vol. 40(4), pp. 331-333.

Johansson A. et al. (2013), “Long-Term Growth Scenarios”, OECD Economics Department Working Papers, No. 1000, OECD Publishing, Paris, http://dx.doi.org/10.1787/5k4ddxpr2fmr-en.

Kvist, J. (2014), “Pensions, Health and Long-Term Care, update 2014, Denmark”, ASISP Country Document, European Commission, March.

Langenbucher, K. (2015), “How Demanding are Eligibility Criteria for Unemployment Benefits, Quantitative Indicators for OECD and EU Countries”, OECD Social, Employment and Migration Working Papers, No. 166, OECD Publishing, Paris, http://dx.doi.org/10.1787/5jrxtk1zw8f2-en.

Liebig, T. (2007), “The Labour Market Integration of Immigrants in Denmark”, OECD Social, Employment and Migration Working Papers, No. 50, OECD Publishing, Paris, http://dx.doi.org/10.1787/233783261534.

Medeiros, J. and C. Schwierz (2015), “Efficiency Estimates of Health Care Systems”, European Commission, Economic Papers, No. 549, June.

Mercer (2015), Melbourne Mercer Global Pension Index, Australian Centre for Financial Studies, Melbourne.

National Audit Office (2014), “Report to the Public Accounts Committee on the Financing of Unemployment Benefits for Insured Unemployed People”, English summary, Rigsrevisionen.

National Audit Office (2015), “Report on the Integration Effort”, English Summary, Rigsrevisionen.

OECD (2015a), Government at a Glance, OECD Publishing, Paris, http://dx.doi.org/10.1787/gov_glance-2015-en.

OECD (2015b), Ageing and Employment Policies, Denmark, OECD Publishing, Paris, http://dx.doi.org/10.1787/9789264235335-en.

OECD (2015c), Fiscal sustainability of health systems, OECD Publishing, Paris, http://dx.doi.org/10.1787/9789264233386-en.

OECD (2015d), In-depth review of the competition law and policy in Denmark, OECD Publishing.

OECD (2015e), Pensions at a Glance 2015: OECD and G20 indicators, OECD Publishing, Paris, http://dx.doi.org/10.1787/pension_glance-2015-en.

OECD (2014a), OECD Economic Surveys: Germany 2014, OECD Publishing, Paris, http://dx.doi.org/10.1787/eco_surveys-deu-2014-en.

OECD (2014b), “Labour Market Integration of Immigrants and Their Children: Developing, Activating and Using Skills”, in International Migration Outlook 2014, OECD Publishing, Paris, http://dx.doi.org/10.1787/migr_outlook-2014-5-en.

OECD (2014c), OECD Economic Surveys: Denmark 2013, OECD Publishing, Paris, http://dx.doi.org/10.1787/eco_surveys-deu-2014-en.

OECD (2014d), “Integrating the Delivery of Social Services for Vulnerable Groups, Bridging Sectors for Better Service Delivery”, OECD Publishing, Paris, http://dx.doi.org/10.1787/9789264233775-en.

OECD (2013a), “The Fiscal Impact of Immigration in OECD Countries”, in International Migration Outlook 2013, OECD publishing, Paris, http://dx.doi.org/10.1787/migr_outlook-2013-6-en.

OECD (2013b), “OECD Reviews on Health Care Quality: Denmark”, OECD Publishing.

OECD (2013c), “Public Spending on Health and Long-Term Care: A New Set of Projections”, OECD Publishing.

OECD (2013d), “Mental Health and Work: Denmark”, OECD Publishing, Paris, http://dx.doi.org/10.1787/9789264191136-en.

Olejaz M. et al. (2012), “Denmark: Health system review”, Health Systems in Transition, Vol. 14(2), pp. 1-192.

Pareliussen, J. K. (2014), “Overcoming Vulnerability of Unemployment Insurance Schemes”, OECD Economics Department Working Papers, No. 1131, OECD Publishing, Paris, http://dx.doi.org/10.1787/5jz1592jj48r-en.

Parsons, D., T. Tranæs and H. B. Lilleør (2015), “Voluntary Public Unemployment Insurance”, IZA Discussion Paper, No. 8783.

Pedersen, E. and J. Riishøj (2008), “Growth, Public Finances and Immigration”, Danmarks Nationalbank Monetary Review, 3rd quarter 2008, pp. 119-129.

Pedersen, P. (2013), “Immigration and Welfare State Cash Benefits: The Danish Case”, International Journal of Manpower, Vol. 34, No. 2, pp. 113-125.

Petersen, O. and U. Hjelmar (2014), “Marketization of Welfare Services in Scandinavia: A Review of Swedish and Danish Experiences”, Scandinavian Journal of Public Administration, Vol. 17(4), pp. 3-20.

Pisu, M. (2014), “Overcoming Vulnerability of Health Care Systems”, OECD Economics Department Working Papers, No. 1132, OECD Publishing, Paris, http://dx.doi.org/10.1787/5jz159228n6j-en.

Rosholm, M. and R. Vejlin (2010), “Reducing Income Transfers to Refugee Immigrants: Does Start-Help Help You Start?”, Labour Economics, No. 17, pp. 258-275.

Sassi, F., A. Belloni and C. Capobianco (2013), “The Role of Fiscal Policies in Health Promotion”, OECD Health Working Papers, No. 66, OECD Publishing, Paris, http://dx.doi.org/10.1787/5k3twr94kvzx-en.

Schou, P. (2006), “Immigration,Iintegration and Fiscal Sustainability”, Journal of Population Economics, Vol. 19, pp. 671-689.

Sørensen P. (2015), “Reforming Public Service Provision: What Have We Learned?”, EPRU Working Paper Series, No. 2015-01.

Sow, M. and I. Razafimahefa (2015), “Fiscal Decentralization and the Efficiency of Public Service Delivery”, IMF Working Paper, No. 15/59, International Monetary Fund.

Taylor Fry (2013), “MSD Actuarial Valuation of the Benefit System for Working-Aage Adults”, Ministry of Social Development and the Treasury, New Zealand.

Vammalle, C., R. Ahrend and C. Hulbert (2014), “A Subnational Perspective on Financing Investment for Growth II – Creating Fiscal Space for Public Investment: The role of Institutions”, OECD Regional Development Working Papers, 2014/06, OECD Publishing, Paris, http://dx.doi.org/10.1787/5jz3zvxc53bt-en.

Wadensjö, E. (2000), “Immigration, the Labour Market, and Public Finances in Denmark”, Swedish Economic Policy Review, Vol. 7, pp. 59-83.

Wadensjö, E. (2007), “Immigration and Net Transfers Within the Public Sector in Denmark”, European Journal of Political Economy, Vol. 23, pp. 472-485.

Wittrup, J. et al. (2013), “Kommunale serviceniveauer og produktivitet,” KORA.

Wubulihasimu P., W. Brouwer and P.van Baal (2015), “Does Living Longer in Good Health Facilitate Longer Working Lives? The Relationship Between Disability and Wworking Lives”, European Journal of Public Health, 1-5.