The global economic crisis that began in September 2008 – with roots that go back much further – has had a devastating effect on incomes, government finances and, not the least, labour markets. Over time, impacts in these three areas will feed into ‘social’ outcomes. With high unemployment and fiscal austerity, increases in inequality, poverty and social exclusion seem likely unless effective counter measures are taken. This report provides an overview of the different ways in which the crisis has impacted on European labour markets in different countries and for different labour market groups. It also looks at the way in which policy has attempted – and to some extent succeeded – in mitigating the worst effects of the crisis on the labour market.[1]

Labour markets: Achievements during the Lisbon period and impact of the crisis

The employment targets set in the Lisbon Strategy in 2000 for 2010 were ambitious: an overall EU employment rate of 70% and employment rates of 60% for women and 50% for older workers. While there were some positive developments with strong employment growth prior to the crisis, particularly among women and older workers (European Commission 2006, p 38), even at its peak in 2008 at 65.9% the overall EU employment rate remained well below the 2010 target. With the global economic crisis employment rates have fallen by more than 2 percentage points (Figure 1). In the first quarter of 2010, EU employment stood at 63.7% and unemployment was 10.2%. Employment has slipped back to its 2005 level and unemployment is higher than at any time during the Lisbon period. This situation is likely to deteriorate further.

Prior to the global economic crisis, a considerable share of employment growth in Europe was due to the increasing proportion of part-time and temporary employment (European Commission 2006, p 24). Part-time employment as a percentage of total employment increased by about 2 percentage points between the beginning of the Lisbon Strategy and 2008. Since the onset of the crisis, part-time employment has increased further and stood at 18.6% in the first quarter of 2010. Temporary employment (all contract forms with limited duration such as fixed-term employment and temporary agency work) has also increased since the introduction of the Lisbon Strategy – its share of total employment reached 14% in 2008. However, temporary employment fell with the crisis, with workers on temporary contracts, particularly temporary agency workers but also those on fixed-term contracts, being the first to lose their jobs. In the first quarter of 2010, temporary employment accounted for 13.2% of all employment. By mid-2010, temporary employment had increased in a number of countries, because in uncertain economic times newly employed are often hired on the basis of temporary contracts.

The Lisbon employment rate target can be used to benchmark European countries. By the first quarter of 2010, only five countries exceeded the Lisbon target: Netherlands, Denmark, Sweden, Germany and Austria (Figure 2). High employment rates among women (considerably in excess of the 60% target for women) contributed to this positive outcome. The worst performers with employment rates of around 55% (Malta, Hungary) have very low female and/or elderly employment rates. The economic crisis has had an adverse effect on employment rates in all countries but four: Malta, Germany, Luxembourg and Poland. Countries especially hard hit in terms of employment are the Baltic countries, Ireland and Spain.

From the first quarter of 2008 to the first quarter of 2010, only Germany saw no increase in unemployment rates. Unemployment more than doubled in Ireland, Spain and Denmark, and more than tripled in the Baltic countries. By the first quarter of 2010, only the Netherlands, Austria and Luxembourg had unemployment rates of below 5%. In Spain, Estonia and Latvia, every fifth person was unemployed (Figure 3).

In all countries, unemployment rates are considerably higher among youth than among other age groups (Figure 4). At 21.4%, average youth unemployment in the EU is more than double the total unemployment rate. In almost all countries, older workers, who are more likely to enter early retirement or functional equivalents of prolonged unemployment, have the lowest unemployment rate. Italy, Sweden and Luxembourg have youth unemployment rates that are more than three times the total unemployment rate, and a large number of countries have youth unemployment rates that are more than double the total rate. Germany, with its dual vocational training system, which cushions the transition from school to work, is the best performer on this indicator. Persons with low qualification levels have considerably higher unemployment rates than those with medium and particularly high qualification levels. On average, unemployment is higher among migrant workers (and particularly non-EU migrants) than among nationals (Figure 5).

The economic crisis has affected different groups of workers in different ways. Between the first quarter of 2008 and the first quarter of 2010, overall unemployment increased by more than 40%. Due to the fact that certain male-dominated sectors (manufacturing, construction) were particularly hard hit, especially in the first phase of the crisis, growth in unemployment was greater among men than women. Men in the EU now have a slightly higher unemployment rate than women, a reversal of the situation before the crisis (Figure 5). Youth were, in general, much more affected by increases in unemployment in the wake of the crisis than prime-age or older workers. Between the first quarter of 2008 and the first quarter of 2010, young workers experienced an increase in unemployment of more than 6 percentage points[2] and prime-age and older workers of less than 3 percentage points. In terms of qualification levels, measured in percentage point changes, those with low qualification levels were most affected by rising unemployment. Also, in general, foreigners (both from the EU and outside the EU) were more affected by rising unemployment than nationals, but again with large country variations around the average.[3]

Forms of non-standard employment have been actively promoted at the EU and national levels as a remedy for unemployment and a way of helping to boost employment rates. Particularly part-time work can be in the interest of workers who want to, or have to, combine work and other activities such as caring. However, both part-time and temporary employment have been shown to lead to spells of unemployment or inactivity, and to have an adverse effect on wages, social security benefits and career advancement (Eurofound 2003; Leschke 2007).

The incidence of part-time employment is much more pronounced in the old Member States; in five countries (the Netherlands, Denmark, Sweden, Germany and the United Kingdom) more than 25% of the working population are employed on a part-time basis (Figure 6). In the Netherlands, this share amounts to 48%. At the other end of the scale, we find Bulgaria and Slovakia with the proportion of part-time employment at below 5%. Reflecting the traditional gender division of wage and care work, there are significant gender differences in part-time employment rates. Part-time employment increased during the global economic crisis from 17.8% in the first quarter of 2008 to 18.6% in the first quarter of 2010. This trend is visible for most countries. The largest growth in part-time work took place in Latvia, Slovakia, Hungary and Estonia, all of which have well below average part-time employment rates.

Temporary employment as a share of total employment varies greatly. At well above 20%, it is highest by far in Poland, Spain and Portugal (Figure 7). Spain has recorded huge declines in temporary employment since the onset of the crisis. Temporary employment is below 4% in Romania, Lithuania, Estonia and Bulgaria.

The share of temporary employment decreased in the initial phase of the crisis and picked up slightly during the last phase: the EU27 average decreased from 13.9% in the first quarter of 2008 to 13% in the first quarter of 2009, and increased again slightly to 13.2% in the first quarter of 2010 (Figure 7). Country trends in this regard were somewhat more diverse. Some countries with comparatively low initial levels of temporary employment – such as Latvia and Malta – saw relatively large increases, possibly because during the crisis employers preferred contracts of short duration, which can easily be terminated. Other countries saw relatively large declines (e.g., Spain, Sweden, Bulgaria and Luxembourg), as temporary agency workers and workers on fixed-term contracts were first to lose their jobs during the crisis. About half of the countries replicate the EU average trend with decreasing temporary employment in the initial phase of the crisis (2008Q1–2009Q1), picking up in the second phase of the crisis (2009Q1–2010Q1).

Although not the main focus here, it should be noted that an additional important labour market effect is on wages (for details see O’Farrell 2010; Glassner and Watt 2010). In many countries, real wages – for those who kept their jobs and did not suffer cuts in working hours – initially held up during the crisis, helping to stabilise demand; they were boosted by lower than expected inflation, the lagged effect of existing collective agreements and, in some countries, statutory minimum wages. Increasingly, however, wage cuts and freezes were implemented, both in the private sector – by firms threatening job losses unless wage concessions were made – and in the public sector. In the latter, the drive to reduce wages was fuelled by a perceived need to reduce government deficits quickly and, in the euro area, as a way of driving down wage and price levels in order to regain lost competitiveness. It is too early to see this in the data, but it is clear that such (relative) wage losses will affect those with the weakest labour market position most severely. In the absence of counter measures, this can be expected to increase income inequalities at the bottom of the labour market.

Role of labour market policies in the crisis

Because economic output (goods and services) is produced by human labour, when output contracts sharply, as happened in the crisis, the demand for labour, and thus employment, is expected to fall and unemployment to rise. If productivity growth remains constant, the fall in employment will be proportional to the fall in output.

Figure 8 compares the changes in output, employment and unemployment for EU countries between the first quarter of 2008 and the first quarter of 2010. Overall, there is a reasonably high correlation between short-term changes in output and those in employment and unemployment. However, there is considerable variation in the sensitivity of employment to output changes for different countries. The Baltic countries and Ireland show the expected pattern of large output losses combined with large declines in employment and a substantial hike in unemployment. Spain, however, although experiencing below average output losses, displayed large labour market reactions. Germany represents the opposite case: despite larger than average output losses, employment and unemployment levels barely changed.

There are a number of reasons for these differences including so-called ‘buffers’ between output, employment and unemployment (discussed in detail in Leschke and Watt 2010). To some extent they reflect productivity differences (for instance, job losses concentrated in Germany’s high-tech export sector compared with Spain’s low-productivity construction sector). But other buffers between changes in output and employment are of greater relevance here. Firms can simply hoard labour, keeping workers on despite the lower output; economically this is reflected in falling labour productivity. Or employment can be maintained through various work-sharing schemes (annual accounts, short-time[4] working schemes, temporary lay-offs); economically this is reflected in falling average working hours. Both labour hoarding and work-sharing schemes are positively affected by workers’ (firm-specific) skill levels because firms are keen to retain qualified workers that they may have difficulty rehiring once the crisis is over. But also, and crucially, they are conditioned by labour market institutions (such as unemployment benefits that can be used flexibly to finance parts of the lost wages at short-time working). These include both ‘preventive’ measures (employment protection legislation) and ‘supportive’ measures (notably government-supported work-sharing schemes). Lastly, job losses can be prevented from showing up in open unemployment by various labour market and other policies (and changes in labour supply). All of these measures have been used to different extents by EU countries during the crisis. Thus, the situation experienced by outliers[5], such as Germany and Spain, can be explained, at least in part, by the application of (or lack of) labour market policies or other institutions that help to cushion the effects of the crisis on labour market outcomes.

As an illustration of the importance of short-time working measures, Figure 9 (taken from Arpaia et al. 2010, p 34) indicates the quantitative effect of measures taken in eight countries expressed as a proportion of the labour force. Without such measures – which also have the positive effects of maintaining workers’ skills and attachment to the labour market and aiding firms in recovering quickly once demand picks up – the drop in employment and concomitant rise in unemployment would have been considerably greater in these countries, even allowing for the fact that the figures are not expressed as full-time equivalents.

In addition to such employment-maintaining measures, active labour market policies (ALMPs) have been deployed, in most cases with a view to preventing a rise in unemployment given the existing scale of job losses. In the relatively short term considered here, the effectiveness of ALMPs depends primarily on the scale of existing measures, on timing (whether or not ‘activation’ policies kick in at an early stage of unemployment) and on the mix of measures implemented (e.g., long-term training versus short-term employment subsidies). Financing systems are also important: where both passive (i.e., unemployment benefits) and active measures are financed from the same ‘pot’, active measures risk being crowded out when unemployment rises sharply, which is when they are needed most.

Looking at the expenditure figures (expressed here as a share of GDP), we see that in the year the crisis broke expenditure on unemployment benefits (passive labour market policies) and active labour market policies including labour market services varied greatly among EU Member States (Figure 10). There is also no apparent link between the amount of expenditure and the level of unemployment. Taking the three countries on the right of the graph, Denmark has massive ALMP spending despite (at that time) very low unemployment; the figures for Denmark are very similar to Belgium where the unemployment rate was several times higher. Active spending in Spain, with the highest unemployment rate in Europe, was comparatively modest.

Unfortunately the most recent data available are for 2008, so we cannot analyse the extent to which ALMP expenditure has reacted to the crisis across the whole of Europe. This is also reflected in the almost identical 2007 and 2008 figures for the EU27 (Figure 11). However, it is interesting to compare a number of countries where the crisis struck early and where the rise in unemployment was already well underway in 2008. Unsurprisingly, there was a substantial increase between 2007 and 2008 in the share of GDP spent on the payment of unemployment benefits of various types (passive measures). This reflects both the hike in unemployment and measures implemented to extend coverage. There has been no parallel increase in spending on active measures, however. In fact, Estonia and Ireland appear to be the only countries that increased spending on active measures, although the increases are small. In most other cases, active spending has remained unchanged or decreased slightly as a share of GDP, and in Lithuania and Latvia it has actually fallen considerably. This may well be an expression of the ‘crowding-out’ phenomenon mentioned above. There is also no evidence of a short-term increase (i.e., already in 2008) in the staffing of public employment services to cope with the increased caseload. However, it is too soon to conclude that active labour market policy has not been effective in the crisis. We need to wait for more recent data: for example, an expansion of public employment services was announced as part of some countries’ stimulus packages in 2009 (Watt 2009).

Conclusion

The economic crisis – the result of macroeconomic mismanagement and an out-of-control financial sector – has had a devastating effect on labour markets, and thus on the livelihoods and prospects of ordinary working people. A heavy price has been paid in terms of lost jobs, reduced hours and associated income losses. In some countries, the effective use of policy instruments by both governments and social partners (Glassner and Keune 2010) has helped to ease the burden on working families and stabilise the economy. Active labour market policies and effective labour market institutions have proved their worth in the crisis, and should be maintained and strengthened.

Increasing inequality was a feature of the prevailing economic growth model prior to the crisis. Given high unemployment and fiscal constraints, there is a real risk of poverty and social exclusion increasing as we come out of the crisis. The question of ‘who pays’ for the crisis will, in various forms, dominate political debate for the foreseeable future, as is already evident in the discussions about the austerity measures announced in a number of countries. So that the burden is not foisted upon those who can least afford it, thereby increasing social exclusion, appropriate measures, such as financial transaction taxes and taxes on inherited wealth or high incomes, will need to be taken. •