“Shelter from the storm?” - Danish flexicurity and the crisis

Denmark received considerable international attention during the decade before the economic crisis. The flashing of the highest employment rate in the EU, the low level of unemployment and an overall positive macroeconomic performance made Denmark stand out as a best practice for Europe (Madsen, 2006; Bredgaard et al, 20062007).

Furthermore Denmark showed some interesting traits, when it came to the country’s combination of the basic building blocks of a Nordic welfare state with some characteristics of more liberal market economies. The Danish development of the welfare state and labour market was thus seen as a successful hybrid between the flexible labour markets in the liberal welfare states characterised by high numerical flexibility (liberal hiring-and-firing rules) and the Scandinavian welfare regimes of generous social security. The hybrid model seemingly managed to reconcile the dynamic forces of the free market economy with the social security of the Scandinavian welfare states (Madsen, 2006).

Under the heading of “flexicurity”, the Danish labour market model was therefore cast in terms of a well-functioning relationship between low job protection, a flexible labour market, high levels of unemployment insurance and active labour market policies (Bredgaard et al, 2006). In the wider European discourse of flexicurity, Denmark was rapidly endorsed as inspiration for a European Social Model that should guide the Members States in the development of their employment strategies (European Commission, 2006).

Since 2008, Denmark has like most other European countries fallen into a deep recession. This was the combined result of the steep decline in the international economy and the bursting of a national housing bubble that had developed in since the beginning of the decade and accelerated after 2005. From the peak of the business cycle in the spring of 2008 to the bottom of the crisis a year later, Danish GDP fell by 8 percent on a quarterly basis. Reacting more slowly to the downturn, total employment reached its lowest level in the first quarter of 2012, having declined by 7.7 percent. The unemployment rates increased dramatically from 3.1 percent in the second quarter of 2008 to 8.2 percent in the beginning of 2011. From having had one of the lowest levels of unemployment in the EU, Denmark moved close to the middle of the rank.

This has stimulated debates about the ability of the Danish model to sustain external chocks and a strand of literature has developed, which discusses the fate of the Danish model of flexicurity in times of crisis sometimes in a comparative perspective (Andersen, 2012; Andersen and Svarer, 2012; Eichhorst et al, 2010; Leschke & Watt, 2010; OECD, 2010).

The present article takes up the comparative perspective by applying more recent data and a detailed identification of the national business cycles. Based here on it is discussed, whether the concrete manner in which the crisis has spelled itself out on the Danish labour market is related to the specific balance between flexibility and security that characterises the Danish model.

Following the analysis of the Danish labour market performance during the crisis in a comparative perspective, the article furthermore gives an broad overview of the actual Danish policy responses to the crisis both at the macro-level and with respect to labour market policy. It then discusses the degree to which the policies do – or do not – reflect a distinct approach linked to the Danish flexicurity-regime. To which degree do the policy responses support or divert from the main characteristics of the model? This analysis will be based on a mapping of Danish employment policy since 2008.

Finally the article assesses the outcomes of the functioning of the Danish flexicurity model during the crisis. Has the Danish model in effect been able to provide security for wage earners in times of severe economic crisis?

One of the aspects of Danish labour market performance, which has been noted in recent years, is the rapid rise in unemployment. In June 2008, Denmark flagged the lowest rate of unemployment in the European Union. In December 2012 Denmark had dropped to being no. 12 in the rank and only 2.7 percentage points below the average unemployment level of EU-27 (Eurostat, 2013).

In Figure 1 the relative increase in unemployment rates in the EU Members States is shown, while Figure 2 depicts the increase in unemployment rates in percentage points. In both figures the underlying data cover the period from the beginning of 2007 to the third quarter of 2012. However, since the Member States were hit by the crisis at different points in time, the increases in unemployment rates in the figures are measured from individual starting and ending points for each country and thus reflect the specificity of the timing of the respective national business cycles. While this procedure takes into account the different timing of the national downturn, it implies a bias in the sense that a few of the observations are “right-censored” by the last observation (2012Q3), which will not for all countries be the bottom of the national business cycle measured by the unemployment rate. However, until information is available for all countries for the full period of the present crisis, this is the only practical approach.

Open image in new window

Open image in new window

Based on the information in Figures 1 and 2 there is little doubt than Denmark is found among the countries, where both the relative and the absolute increase in unemployment has been most dramatic. In relative terms Denmark is only surpassed by the Baltic States, Ireland, Greece and Spain. Since Denmark entered the crisis with a very low level of unemployment, the relative increase is somewhat biased. However, also measured by the absolute increase shown in Figure 2, Denmark is clearly above the average for the EU-27.

The rapid increase in Danish unemployment could of course reflect an equally dramatic fall in employment. In Figure 3 the fall in employment is compared. Again the change in employment is measured taking into account the national development of the business cycle. The starting quarter is thus the quarter with the highest employment level since the beginning of 2007 and the ending quarter is the lowest level found since the peak, which only for Greece, Spain and Portugal is the last quarter in the sample (2012Q3).

Open image in new window

Based on the evidence in Figure 3, the Danish crisis on the labour market also appears dramatic with a decline in employment, which is significantly above the average for the EU-27. The same goes for the decline in GDP, cf. Figure 4.

Open image in new window

Finally Figure 5 presents the employment elasticity with respect to GDP defined as the relation between the relative change in employment and in GDP (both calculated from peak to through). Poland is omitted from Figures 4 and 5 due to the exceptional development in its GDP, which hitherto has not been declining during the economic crisis.

Open image in new window

The immediate impression from Figure 5 is that the relationship between the change in GDP and in employment in the Danish case clearly above the European average. Given the wide-spread notion of Denmark as having a labour market characterised by a high level of numerical flexibility, this comes as no surprise. Especially one also notes that Denmark has a relatively high elasticity of employment to GDP, if one solely looks at the old Member States. Here Denmark is only surpassed by Spain, Portugal and Ireland. Similar calculations presented in European Commission (2010:28) indicate a similar ranking of countries with respect to their of employment elasticities of GDP, although there are some differences in the level of the elasticities probably due to differences in the time periods for which data have been available at the time of the calculation and also because the Commissions’ calculations are based on the same time period for all countries.

Space does not allow for a more detailed analysis of the relationship between the change in GDP and employment (measured in persons) and other macro-economic variables like productivity and average working hours, which act as “buffers” between the change in GDP and the resulting change in employment. However, the employment elasticities calculated above are somewhat higher than those published elsewhere, which have indicated that Denmark is among the countries, where the response of employment to the decline in production has been around average (Andersen, 2012; Eichhorst et al, 2010; Leschke & Watt, 2010; OECD, 2010). Possible explanations for this new result can be that the present calculations are based on newer data, and that they also take the specific timing of the national business cycle into account.

Summing up the impression of the impact of the economic crisis on the Danish labour market one can thus make the following observations:

Both these observations are thus consistent with the hypothesis that the high level of external numerical flexibility of the Danish labour market – corresponding to a relatively low level of EPL – will lead to more volatile changes in unemployment and employment over the business cycle. This of course will go both ways and also imply that employment will catch up rather rapidly, when an upswing occurs. Furthermore one must stress that a high level of internal numerical flexibility (working time flexibility) also plays an important role on the Danish labour market. Without the flexibility of working hours one would have expected the decline in employment (measured in persons) to have been even higher.

Finally one however should point to the fact that there are – in a cross-country comparison - no simple correlations between the elasticity of employment (in persons) with respect to GDP and institutional variables like the level of employment protection legislation. As indicated by the case of for instance the UK, where employment protection is generally assessed as rather low, other intervening factors like productivity and wage flexibility can play an important role and lead to deviations from a simple relationship between EPL and the employment elasticity. Another example is Spain, where the EPL for ordinary workers is high, but one still observes a high employment elasticity due to the large share of temporary employees on the Spanish labour market. At the other end of the spectrum Germany stands out with a very low employment elasticity as a reflection of the importance that internal numerical flexibility (adjustment in working time) has played on the German labour market during the crisis.

One of the most striking observations from the crisis has thus been the diversity in the way in which the crisis has played itself out on the national labour markets of the EU. Also when it came to the consequences of the crisis on the labour market, one size did not fit all.

Following this brief comparative exposition of the changes on the Danish labour market, we now turn to the political reactions to the crisis. Which policies have been enacted in order to provide “shelter against the wind”? Do the Danish policy responses to the crisis reflect a particular “flexicurity profile” and – following that – can the responses be conceived at sustainable in the sense that there are no indications of serious barriers to the implementation of such policies? Or has the stress of the crisis lead to a dismantling of the basic pillars of the model?

This section therefore presents the major policy initiatives, which have been taken as a response to the crisis since 2008.

The policy initiatives are grouped according to the traditional division between macro-economic policy and labour market policy. After an overview of the policy shifts in the respective areas, the text goes on to discuss the issues of conformity with the Danish flexicurity model and its sustainability.

3.1 Fiscal policy initiatives

The rapid rise in unemployment during the fall of 2008 and the following winter put the issue of “growth packages” high on the political agenda. Since the spring of 2009 a number of concrete expansionary measures have been undertaken. The measures introduced during 2009 and subsequent years include access for the municipalities to increase their investments beyond the existing spending limit, a state subsidy to renovation of private and several public “investment packages” in infrastructure, energy saving etc. Furthermore, taking effect from 2010, a tax reform was introduced, which was not fully financed during the initial years.

To these discretionary fiscal policy measures must be added the fact the Danish economy due to its rather high tax level and the large share of the workforce covered by unemployment benefits, has some of the largest built-in automatic stabilisers in the EU, cf. Figure 4 (Dolls et al, 2009). Thus the effects of the economic downturn on income and unemployment are significantly dampened. The backside of this is of course that the public budgets have deteriorated rather dramatically during the crisis.

Later developments in Danish macro-economic policy have however represented a stepping back from the expansionary fiscal policy of 2009 and 2010. Triggered by rising deficits on the public budgets and a recommendation from the European Commission to restrict the deficits, the Conservative-Liberal Government in May 2010 began political negations with the Danish People’s Party, which rapidly resulted in a political agreement for fiscal recovery to be implemented in 2011 and subsequent years. Important elements in the agreement were combined savings and reallocation of public expenditure and a temporary hold on the indexation of the tax scale, which implies a de facto tax increase. The agreement also implied a cap over the yearly amount of membership fee for trade unions, which the members can deduct, when calculating taxable income and rising fees for labour market training. These elements were of course fiercely opposed by the trade unions.

However, the most controversial elements in the agreement was a shortening of the duration of unemployment benefits from four to two years and a tightening of the criteria for regaining access to unemployment benefits, once the right to benefits had been exhausted. These elements are further discussed below.

In the subsequent years the issue of “growth packages” and “kick starts” has remained high on the political agenda, also stimulated by the change of Government in October 2011. The new Government was formed by a coalition of socialists, social democrats and social liberals under the social democratic prime minister Helle Thorning Schmidt. The Government has however chosen a cautious strategy with respect to fiscal expansion putting great emphasis on the need to obey the recommendation from the European Commission.

Taking the period as a whole, according to the latest available estimates from the Ministry of Finance shown in Figure 6, fiscal policy in 2009 and 2010 contributed to the growth of GDP by 1.2 and 0.5 percentage points respectively, measured by the one-year fiscal effect. For 2011 the fiscal effect was strongly negative mainly due to a decline in public consumption of -1.5 percent in real term. Also rising taxes contributed to the tightening of fiscal policy. In 2012 fiscal policy was almost neutral, while the fiscal effect in 2013 is again estimated to be negative by -0.3 percent. Thus there is a clear shift in fiscal policy from the first two years of the crisis to more recently, where the austerity pressure has set in. Not surprisingly this development has caused a heated policy debate about, whether the Government is conducting a fiscal policy, which is overly tight and implies the risk of causing severe structural damages to the economy and the labour market. Going further into this issue is however outside the scope of present article.

Open image in new window

3.3 Changes to the content of ALMP

In February 2009, a broad political agreement was reached inspired by a proposal from the social partners. It implied a number of changes in the rules for active labour market programmes. The changes were all aimed at targeting the programmes more towards upgrading the skills of the unemployed in the light of the composition of labour demand. Furthermore one of the initiatives made by the Minister of Employment in March 2009 implied that the funds allocated to assist employees that are affected by collective dismissals are increased. Furthermore, the Minister of Employment in January 2010 announced new initiatives to assist employers that are restructuring and employees that are about to be dismissed. These initiatives were in part inspired by a joint set of proposals from the social partners, who in January 2010 published 23 concrete proposals to improve employment policy and combat long-term unemployment (LO and DA, 2010). All in all, these initiatives implied a slightly higher priority to education and training and also more intensive contact with the unemployed in the form of frequent meetings with counsellors at the job-centre.

As mentioned above, in October 2011 a new Government headed by a Social Democratic Prime Minister took office. The Government announced a number of reforms of labour market policy and retirement schemes, the latter including the disability pension and the flexi-job-scheme. The most important elements can be summarised in the following points:

•  The access to education for unemployed was widened and recipients of cash benefits, which have other problems than just unemployment, were now to be activated every 12 months (instead of 6) to allow for a more flexible activation of this vulnerable target group.

•  More effective efforts to reduce youth unemployment including more job-rotation for young persons, more young to take part in the adult-apprenticeship-scheme, targeted programmes for young persons with lack of basic education and more traineeships for graduates.

•  Strengthened efforts to reduce the drop-out-rate especially from the vocational education system.

Thus, while not representing a drastic break with previous policies, the many minor reforms can be interpreted as pointing toward a more flexible approach to active labour market policy and also implies the allocation of more resources to interventions.

The issue of reforms of the organization, steering and financing of active labour market policy has been postponed several time (following failed tripartite negotiations in June 2012). A new attempt to deal with these complicated matters was announced by the Government in February 2013. An expert committee and a committee of social partners shall work in parallel to develop solutions during 2013 and 2014.

In quantitative terms Figure 7 sums up the development in the priorities of active labour market policy during the crisis. The main message from Figure 7 is that the element of counselling, training and education in active programmes has been reduced significantly, while traineeships and subsidised employment has increased. This policy shift during times of weakening labour demand has been criticised also from leading politicians. Actually the mandate of the expert committee just mentioned call for an investigation of the potential for increasing the role of education in active labour market policy.

Open image in new window

In assessing the policy responses to the crisis since 2008, one must highlight that Danish labour market policy to a large degree is “rule-based” in the sense that the law on active employment policy in detail specifies a number of rights and obligations for the unemployed and the jobcentres. Among the most important ones are individual dead-lines for contacts with the job-centre and for taking part in mandatory active programmes. This implies for instance that an unemployed member of an unemployment insurance fund must be activated after 9 months of unemployment or 3 months, if the person is aged less than 30 years. Unemployed aged 60 years and above are activated after 6 months of unemployment

This rather strict regime of monitoring and activation is not related to the business cycle. As also noted by the ILO, Danish labour market policy therefore has the character of an automatic stabiliser, where there is little need for discretionary actions to be taken under a downturn (ILO, 2009:23).

This role is reinforced by the universal character of Danish labour market policy, which encompasses the vast majority of wage-earners (and self-employed) due to their membership of an unemployment insurance fund or access to means tested social security. Thus, according to a study from the Economic Council of the Labour Movement, 85 percent of the employed would qualify for unemployment benefits or social security, if becoming unemployed (Arbejderbevægelsens Erhvervsråd, 2010a).

The major challenge in this context is of course the necessity to adjust the resources of the jobcentres to accommodate to the larger inflow of unemployed that follows from fall in employment. There are indications that the jobcentres have faced severe difficulties in meeting the deadlines. Thus in January 2010 only 58 percent of the insured unemployed started activation in time. For young unemployed aged less than 30 years this was the case for only 42 percent (Arbejderbevægelsens Erhvervsråd, 2010b).

Both when it comes to macro-economic policies and labour market policies a characteristic feature of the Danish policy response is thus its non-discretionary nature. This goes both for the important role of strong automatic stabilisers in the macro-economy and for the rule-based and universalistic nature of labour market policy. In spite of the, one could of course have expected that the crisis had lead to large scale policy reforms that more profoundly changed the Danish labour market model . As described above the response to the crisis has hitherto been within well-known policy boundaries. Applying the classic taxonomy of Hall (1993) one could say that the Danish employment policy has been mainly characterised by first first-order changes, where existing policy instruments are adjusted, but not radically changed.

However one can also identify in some examples of second-order changes, where new instruments are introduced. Thus although “growth packages” have also been applied during previous economic downturns, the strong focus on the need to stimulate job creation in 2009-10 can be interpreted as a policy shift.

Within labour market policy, the shortening of the duration of unemployment benefits from 4 to 2 years that was decided in May 2010 can be interpreted as another important second-order change, given the fact that the duration of benefits had not been applied as a policy instrument since the late 1990s and then in a much less controversial setting given the steady decline in unemployment at the time.

Table 1 sums up the picture.

Open image in new window

Among Danish academics it is sometimes argued that the first- and second-order changes described above have already lead to a fundamental change in the Danish flexicurity-model, which makes it relevant to consider it as now being fundamentally different from the model that we have previously known. Whether or not to subscribe to this view is a matter of judgment. When one is in the middle of a historical process it is always hard to assess to degree to which it will later be seen as leading to more fundamental transformations.

However, the present author tend to subscribe to the view that the Danish model of flexicurity is still basically intact and retains the same characteristics that is has had for decades. Although some reforms have been introduced in recent years, policy-makers still support the main characteristics of the model. An explanation for this situation could be that the Danish policy arrangements have in fact managed to handle many of the challenges of the economic crisis. So if it works, don’t fix it. One observation in support of this argument is that the speed in the rise of registered unemployment has leveled out during the spring of 2009 and that the forecasts for economic growth are improving somewhat from the disastrous decline in GDP of between 4 and 5 percent in 2009. Thus forecasters publish expected growth rates for 2012 and 2013 around 1 percent. This will not be sufficient to rapidly restore employment and unemployment to their pre-crisis levels, but will at least prevent them from a continued dramatic deterioration.

Pointing to the stability of the basic elements of Danish flexicurity does not imply that nothing has changed or that nothing will change in the coming years. To the contrary it is evident that the policy changes described in section 3 have put the traditional pillars of the Danish model of flexicurity under stress, but without fundamentally corrupting the functioning of the model.

As described in detail above, the crisis has implied a dramatic rise in unemployment and substantial fall in employment. Also the labour force has been reduced due to the lack of demand. Long-term unemployment has more than doubled. Especially young workers, unskilled workers and immigrants and non-insured unemployed are severely affected by unemployment.

A further issue is the adequacy of income support. As already described, about 85 percent of the employed will receive some form of income support, if becoming unemployed. The backside hereof is of course, that 15 percent will be without any safety net and also not eligible for receiving the offers of active labour market policy. The rising number of persons reporting as unemployed in the Labour Force Surveys, but not being registered as unemployed at the job-centres, could be another indication of the limits of Danish income security. Also, the declining membership of the unemployment insurance funds may to some degree be caused by the erosion of the replacement rates of unemployment benefits, which has taken place since the 1980s.

But on the other hand, there are also indicators that the Danish version of flexicurity is still providing the Danish labour market with some shelter against the crisis.

Second, the basic security arrangements in the form of income security and active labour market policy are still functioning in spite of the increased pressure from the crisis on the labour market. One indication hereof is the observation that Danes rank very low in studies of the economic hardship caused by the crisis (Eurobarometer, 2012). When asked about whether the household has run out of money to pay ordinary bills, buy food or other daily consumer items during the last 12 months, 8 percent of the Danes answer positively. This is together with Sweden the lowest share in the EU-27, where the average is 18 percent. When asked about, whether they are able to keep up with household bills and credit commitments, 79 percent of the Danish household state that they can keep up without any difficulties. This is the highest share in the EU, where the average is 44 percent.

When it comes to the “political sustainability” of the Danish version of flexicurity, the support to the present state of affairs with respect to Danish flexicurity seems to be there - even in times of crisis. Or to quote the former Minister of Employment:

Still, I have faith in flexicurity. It will serve us well in both the best and the worst of times. Flexicurity allows us to adjust to the changes in the market, and it secures the livelihood of the unemployed (Frederiksen, 2009 , p. 2)

One should mention that he also in the same speech stresses the need for reforms, but only to preserve the model, not to dismantle it. A similar positive assessment of flexicurity is also expressed by the OECD (2012:13-14):

Flexicurity should help Denmark both during hard times and to achieve strong eco- nomic growth over the longer term…This model fosters low unemployment and high employment but it may be tested by prolonged periods of low labour demand…

The Danish welfare system acts as a buffer in periods of crisis. While GDP per capita has lost some ground relative to the upper half of OECD countries, well-being in terms of both material conditions and quality of life is very high . This is the result of a well-functioning labour market and a well-developed and generous welfare system that includes social policies directed at helping those with the lowest incomes, broad access to education and free access to most health services. This system is costly, however, with Denmark spending more than 20% of GDP on social policies. Even so, net public social spending is higher in a number of OECD countries where poverty is more prevalent.

Of course there can be a thin line between conducting a more wide-ranging renovation and genuine demolition. Thus the shortening of the duration of unemployment benefits and the tightening of the access to regaining the right of benefits was conceived – with support from the OECD – by the same Minister Frederiksen, who is quoted above.

This critique was also expressed by the leader of the Social Democratic Party, now Prime Minister, who in several interviews during the summer of 2010 argued that the long duration of unemployment benefits was essential for keeping the balance in the Danish model of flexicurity. However, given the present balance of power in the Danish parliament, it is unlikely that one will see a return to the former four year duration.

Summing, up the available evidence, Danish flexicurity seems to be able to provide some shelter against the storm of the crisis. It also seems to be reasonably robust against political interventions of a more drastic nature that would go far in dismantling the traditional Danish version of flexicurity. On the other hand it is evident that the combination of policy changes and the economic crisis has put the model under stress compared to its latest heydays during the long upswing of from 1993 to 2008.

The balance between flexibility and security has without doubt drifted away from the equilibrium that until recently gave the Danish model the widespread support from most political parties and the social partners. Whether a new balance can be found – not far from the old one – will depend on the political developments and the reactions of the trade unions over the coming years. If institutions providing income and employment security are further eroded, and if the response from the trade unions is claims for more traditional job security, it could mean a gradual farewell to “flexicurity in Danish”.