Miroljub IGNJATOVIC* LABOUR MARKET FLEXIBILITY AND SECURITY IN THE EU Abstract. The article deals with the profound changes (reducing various forms of security) that modern European labour markets are experiencing due to changes in dominant paradigms in the current period and to processes such as globalisation and flexibilisa-tion. After analysing Eurostat's labour market and social security indicators, we could confirm that there are different flexibilisation 'models' in the EU since EU countries are using different forms of flexible employment as the primary source of flexibilisation. We also found that the 'selection' of a particular form of flexible employment is strongly correlated, among other factors, with the level of wealth accumulated in a particular society and that the different forms of flexible employment produce different levels of risk of poverty. 901 Keywords: European labour market(s), flexibilisation, security, risk of poverty Introduction Modern labour market(s) can be defined as the "central institution in modern society" (Jensen, 1989: 406) in which through its own mechanisms processes are underway concerning the formation, interactions and regulation of two quite abstract categories - the supply and demand of labour. It is only through individuals acting as carriers of these two functions (supply and demand of labour) and the holders of different cultural and social patterns, patterns of behaviour (action and reaction, tactics and strategies) that the labour market acquires its final form and importance in society. The labour market's importance is determined by both individuals and society as a whole. First, by individuals, because in one way or another they appear in the market and are more or less dependent on activity -work, which should be the most valued in the labour market and in society. Second, the labour market's importance is determined by society as a whole. With the help of economic and social criteria, society determines * Dr. Miroljub Ignjatovic, Assistant professor, Faculty of Social Sciences, University of Ljubljana. the price of work, sanctions and limits it in the form of socially recognised work. As the central institution of modern society, the labour market has a large impact (which increases in periods when jobs or any kind of work possibilities are scarce) on the formation of certain social structures, the prosperity of modern societies and the position of the individual in society, their security or chance of survival and the choice of a particular lifestyle or habitus, by providing them with their primary source of income. With increasing flows and dynamics within the modern labour market, chiefly due to the processes of globalisation and flexibilisation, the individual's position in the labour market and in society is ever more diversified and, in the case of those involved in some flexible forms of employment even more dependent on the redistributive measures put in place by the welfare system of society1. As such, the labour market is a significant factor in social reproduction and the (re)production of social stratification. The story of the labour market is thus also a story of an attempt to reconcile two fundamental principles of modern developed societies: (economic) efficiency and (social) security. The problem of modern societies is how to ensure the simultaneous fulfilment of both. Recurring problems 902 of high unemployment and a desire for strong economic growth have put a great deal of pressure on European societies and labour markets in the past 30 years, making them shifting the balance between the two principles in favour of economic efficiency. One generally accepted strategy for pursuing efficiency in the past 30 years has been flexibilisation of the labour market. However, there are considerable differences in the extent, scope and direction of the actual flexibilisation of different models of welfare states2. In this article, we argue that this shift holds a much deeper meaning for modern societies and the labour market than simply an adaptation in response to the current financial and economic crisis. It represents the gradual yet fundamentally important redistribution of risks related to economic activity in modern societies. As employers are encouraged to take even greater risks in performing economic activity, the price of this risk would be redistributed (mostly unevenly) between social groups in society. Since the burden of this risk largely takes the form of rising social insecurity (reducing different forms of security, see footnote 3) and consequently increasing risk of falling into poverty. Since both consequences (increasing social insecurity and at-risk-of-poverty) are related to the growing shares of various flexible forms of employment, European model(s) of the welfare state are 1 It could be said that the labour market is the framework in which the relationship between social and economic policies are most clearly defined (Rus, 1990: 352). 2 For more on welfare state models, see Esping-Andersen (1990) and Sapir (2005). taking on the task of compensating the risks by redistributing the accumulated wealth, thereby increasing social security (mostly by providing income security) for those parts of the workforce dealing with flexible employment and unemployment. It seems that the current financial and economic crisis has reopened questions about the structure and effectiveness of Europe's labour markets, the need for their further flexibilisation, the consequences flexibilisation brings to different forms of security3 especially for those involved in flexible forms of employment and the unemployed and about levels of compensation and the redistribution of wealth. Individual countries or regions have reacted differently to these pressures. The response has largely been in line with the concept of an individual welfare state, which is dictated by the dominant cultural model, economic situation and is in accordance with the objectives and long-term strategy developed within these concepts and models. Put differently, "there are in reality different European social models, with different features and different performance in terms of efficiency and equity" (Sapir, 2005: 1). Following the arguments presented above, in this article we attempt to analyse the following hypotheses: 1. There are different 'models' of flexibilisation in the EU since EU countries are using various forms of flexible employment as the primary 903 source of flexibilisation. 2. The 'selection' of a particular form of flexible employment is strongly correlated, among other factors, with the level of wealth accumulated in a particular society (measured by GDP per capita). 3. Different forms of flexible employment produce various levels of risk related to entering into poverty (measured by the in-work at-risk-of-pov-erty rate). In order to test these hypotheses, we will analyse Eurostat's data for EU countries on the labour market structure, the shares of flexible forms of employment (part-time, temporary and self-employment), and on different indicators of the risk of poverty and the redistribution of wealth. Changes in flexibility - a security balance in European labour markets After a period (1945 - mid-1970s)4 of seeking a social consensus and the extension of labour rights and entitlements (Standing, 1997), in the 1980s 3 Standing (1997; 2011) distinguishes between seven different forms of security appearing in the labour market: labour market security, employment security, job security, work security, skill reproduction security, income security and representation security. 4 When it seemed that the principle of efficiency may also contribute to ensuring the principles of social security and that there should not always be a trade-off between them. the rising unemployment, inflationary pressures and problems of sustaining the growing welfare state shifted the pendulum of the flexibility-security balance towards the flexibility side. As Sapir (2005) suggests, all the labour market and social policies that were established in a relatively stable and predictable environment (and based on a male breadwinner as the unit of production and source of security for the whole household) have increasingly become dysfunctional in the more variable economic environment in the late 20th and early 21st century. Further, according to Standing (2011) informal mechanisms of community help which supported individuals and families in the past are also losing their strength. Several authors (Standing, 1997; Gray, 2005; Heyes, 2011; Cook, 2008; Burroni and Keune, 2011) have pointed to this shift towards the neo-liberal policies of the 1990s on "committing the EU to a supply-side, workfarist5 approach to unemployment and to flexibilisation (read 'de-regulation') of the labour market" (Gray, 2005: 3) despite the simultaneous efforts, at least in many documents (the Social Europe project, the Amsterdam and Maastricht treaties, the Lisbon Strategy), to pursue targets such as job quality, social inclusion and the 'make work pay' principle. The continuous flexibilisation of the EU labour market which is also vis-904 ible in the first decade of the 21st century means increasing the shares of those working in flexible forms of employment in the total labour force. As Chart 1 shows, the share of part-time employment (performed predominantly by women - in the second quarter of 2012, 32.1 % of women and 8.5 % of men worked part-time, respectively) rose the most in the 2000-2012 period, while the share of temporary employment reached the share of self-employment. Flexibilisation of the labour market was seen as a strategy which would bring more jobs and increase both employment rates and economic growth. On the other side, the principle of full employment was abandoned in many EU countries and "securities that had been regarded as the primary objectives of economic and social policy in the previous era became regarded as obstacles and rigidities to be overcome, in the name of economic growth" (Standing, 1997: 14). Further, as Cook (2008: 43) suggests, "the diagnosis of the cause of unemployment shifted from recognition of inadequate demand to the proposition that unemployment was rooted in supply deficiencies. In particular, the unemployed were viewed as lacking the necessary skills and abilities to obtain employment, or were labelled welfare dependent and unwilling to work". 5 Workfarism' or the 'work first'principle is a set of "benefit rules and employment service practices which are designed to lower jobseekers' aspirations and wage expectations" (Gray, 2005: 2) in order to get jobseekers as soon as possible back into the active part of the labour market, and to avoid their 'state dependency'. Chart 1: TRENDS IN THE THREE MOST WIDESPREAD FLEXIBLE6 FORMS OF EMPLOYMENT IN THE EU, 2000-2012 18,7 18,9 .......19,3..... 17,6 17,7 17,7 16,7 15,8 15,7 16,1 J17............. 14 7 ... ,............14,5....... *— — _ K.5....... .....14,2........ 13,9 ""H,'5..... M,6..... 14,2 14 14,2 13,9 , 13,2 13,5 12,2 12,4 UA...........12'6........................ 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 — Self-employment Part-time employment ■ Temporary employment Data source: Eurostat, 2012 Such a diagnosis also demanded a new treatment - the proportion of longer vocational courses declined in favour of shorter and cheaper courses 905 focussing on job search and employability skills, thus promoting the 'work first' (instead of 'train first') approach as a prescribed cure. Yet this cure also had a negative side-effect - an increasing proportion of jobseekers in low quality jobs with a high probability of entering into a vicious circle of quickly returning to the employment office and/or finding another low quality job, thereby also producing a multi-tiered or segmented labour market. Moreover, while painting a similar picture as Cook above, Standing (2011) highlights another change with an important influence on the social security of unemployed persons. While built on the principle of social insurance (based on a person's previous contributions during their working period) unemployment benefits are nowadays available to a decreasing share of the unemployed. Due to the shorter spells of employment (mostly in low paid flexible jobs), along with the related more frequent spells of unemployment ('unemployment trap') and tighter entitlement conditions for receiving unemployment benefits7, ever more unemployed are finding themselves in poverty. 6 Shares of part-time employment and self-employment are calculated as percentages of total employment, while the share of temporary employment is calculated as a percentage of the total number of employees. 7 Calmfors (2007:3) maintains that there is overwhelming empirical evidence that higher unemployment benefits (passive labour market policies) tend to raise unemployment by reducing the search intensity of the unemployed and the taking up of job offers. On the other hand, activation policies (active labour All of the abovementioned consequences of the increased flexibilisation and changes in the redistribution of risks have enhanced the need to introduce a new concept which would preserve the 'European social model' and offer greater security to those at risk. Flexicurity as a European solution for the flexibility-security balance in European labour markets Europe of the first decade of the 21st century has seen some new attempts to implement a more balanced application of flexibilisation and security through the flexicurity8 concept, which had proved successful in the Dutch and Danish cases at the end of the 1990s. In both cases, especially the Danish, the high risks of job insecurity were compensated with generous income security and active labour market policies which greatly contributed to the feeling of higher employment security. The Danish and Dutch examples both proved it is possible to avoid a trade-off between flexibility and security by using labour market policies that are coherent with cultural, institutional and labour market frameworks that promote high levels of social protection in society. 906 In contrast, the current financial and economic crisis has turned the tables on the promotion and development of strong flexicurity concepts in EU countries. Standing determined 15 years ago that "the state apparatus is facing budgetary cutbacks in the name of fiscal discipline, which makes it hard to provide adequate social transfers to placate the poor and even hard to police the poor efficiently" (Standing, 1997: 28), and the same conclusion could apply to the majority of EU countries today. Moreover, Heyes (2011: 653) suggests that during the current crisis "there is evidence of convergence in respect of labour market and social protection policies, but the dominant tendency is towards 'less security' rather than 'flexicurity'. To varying extents, EU member states have weakened employment protections, reduced benefit entitlements and linked those entitlements ever more closely to workers' preparedness to participate in ALMPs or accept a job, regardless of its quality". But while old social security systems, largely based on job security (its stability) with unemployment insurance depending upon that, are beginning to show weaknesses in providing decent social security to those faced with flexible forms of employment and market policies) mainly seem to have had ex ante threat effects, shortening the unemployment duration by changing the behaviour of the unemployed prior to programme participation. 8 Flexicurity is a »policy strategy that attempts, synchronically and in a deliberate way, to enhance the flexibility of labour markets, the work organisation and labour relations on the one hand, and to enhance security - employment security and social security - notably for weaker groups in and outside the labour market on the other hand« (Wilthagen and Tros, 2004:169). to unemployed persons, the need is increasing to introduce a new social security system that would ensure security for individuals based on their actual labour market activity and their employability9. In addition, this same subsystem should also act as a safety net or 'trampoline' for those finding themselves in flexible employment and especially for those who are also excluded from this group (such as the unemployed or inactive). As mentioned, the flexicurity concept was introduced by the Dutch and Danish welfare states in the late 1990s to compensate for the decreasing job security (fewer permanent jobs and easier dismissals) by improving employment opportunities and social security benefits in the labour market. Later, mostly in the second half of the 2000s, it was adopted by the European Commission as part of the European Employment Strategy and from 2007 promoted as eight common principles of flexicurity (Council of the European Union, 2007) which suggest that member states introduce a deliberate combination of flexible and reliable contractual arrangements, comprehensive lifelong learning strategies, effective active labour market policies, and modern, adequate and sustainable social protection systems tailored to the specific circumstances of each member state. Introduction of the flexicurity concept on the EU level seemed to be a specific European response to the challenges posed by the tougher global 907 competition and the demand to increase the labour market's flexibility. In contrast to the classical neo-liberal approach, the flexicurity approach recognises the fact that in many cases increased flexibilisation reduces the employment and income security of workers engaged in flexible forms of employment. Consequently, it offered policies that should address the issue of reduced (especially) employment and income security in accordance with the specific circumstances of a particular country. Thus, the flexicurity concept was accepted by many as the winning strategy which should solve the trade-off problem. Critiques of the flexicurity concept The flexicurity concept has also attracted many, more or less justified critiques. Some authors have questioned its transferability10 to different welfare models outside the two countries (Denmark and the Netherlands) that 9 "Security for an individual, it is suggested (Bridges, 1995; Kanter, 1989; Mirvis and Hall, 1994), will be anchored not in a particular organisation but in their own portable skills and employability. People are encouraged to weaken their ties with organisations" (Mallon, 1998:363). 10 Schmid (2009a: 39), for example, claims that: "As successful employment systems demonstrate, flexicurity has to be embedded in sound macro-economic and macro social policy. Without sustainable job creation dynamics, all employability and stepping-stone strategies are in danger of ending in a cul-de-sac or in displacing other categories of workers successfully implemented it and criticised the vagueness11 of the flexicurity definition, which "has become a 'catch all' concept that can easily accommodate a variety of (sometimes contradictory) policies. It includes the whole menu of both active and passive labour market policies, training and education and social protection, as well as labour law and collective bargaining. With the principle of 'no one size fits all', meaning that each of the 27 member states can have its own version, no systematic approach towards flexi-curity is possible" (Auer, 2010: 374). Further, when questioning the transferability of the flexicurity concept Schmid (2009: 1) mentions the "illusion of flexicurity as a guiding principle for all countries", while Calmfors (2007) thinks that the concept too easily promotes a 'win-win' position, and underestimates important trade-offs like between generous social protection and the unemployment trap. Similarly, in their critical assessment Burroni and Keune (2011) focus on four aspects of the flexicurity concept: its conceptual ambiguity from which stem different views on how to translate this abstract concept into policy; its failure to problematise the assumption that there can be a positive-sum game between regulations increasing flexibility and those reinforcing security; its lack of attention to conflicts of interest (especially between capital 908 and labour) and to the heterogeneity of the European labour market; and its reductionist view of the sources of flexibility and security in the labour market on the EU and national levels assuming that there is high degree of homogeneity. Regarding the latest point, Burroni and Keune list numerous internal12 and external13 factors on the national level that have an important influence on the ability to implement the flexicurity concept on the national level. In his analysis, Tangian (2010a: 6) suggests that "high labour flexibility shows no macroeconomic advantages under crisis conditions. When a crisis occurs, both economic losses in firms and labour adjustments occur on a massive scale, aggravating both the economic and the social situation (increase in the output gap and in unemployment). The burden on public finance (size of bailout packages and aid to the unemployed) further aggravates the situation. One possible explanation is that in 'good times' the availability of an external flexibility option encourages employers to take higher risks, since potential losses can be recovered through unproblematic labour adjustments in the event of a crisis. 11 For example, Calmfors (2007: 2) suggests that "there seems to be a tendency for everyone to have their own definition of flexicurity and then to subsume everything they like under that label". 12 Important types of regulation like collective agreements, company policies or informal rules, roles played by families and other actors, macroeconomicpolicies and more. 13 International competitiveness, global and regional economic situation, functioning of financial markets, ... Low flexibility, on the other hand, restricts labour adjustments and thereby constrains risky economic behaviour. As a result, firms (a) operate in a more secure and stable way, (b) carry out fewer labour adjustments, which is positive for employment and, accordingly, (c) burden the state with less additional social expenditure for supporting the unemployed. In other words, flexibility is disadvantageous in times of crisis". While Tangian rightly determines that the flexicurity concept means a burden on public finances, it should be reiterated that this burden is a consequence of the actual redistribution of risks and the related compensation in the form of the redistribution of wealth through the social security system. But such a scenario and the whole flexicurity concept are endangered by the current austerity measures and attempts to reduce public debt in many EU countries. Thus, it is not surprising that Heyes (in O'Reilly, 2011) suggests there has been no widespread move towards flexicurity among EU countries. As Burroni and Keune (2011) emphasise the importance of approaching national examples as heterogeneous cases which contain significant within-country differences, we will continue with an empirical analysis of differences regarding flexibilisation and its influence on security and risk of poverty among EU countries. 909 Empirical analysis As presented in the introduction, in order to test the presented hypotheses we will analyse Eurostat's data for EU countries (for 2010) on the labour market structure (the selected variables are: employment rate; unemployment rate; and long-term unemployment rate), shares of flexible forms of employment (the selected variables are: share of part-time employment in total employment, share of temporary employment in the total number of employees; and share of self-employment in total employment), and different risk-of-poverty indicators (the selected variables are: at-risk-of-poverty rate before transfers; at-risk-of poverty rate after transfers; and in-work at-risk-of poverty rate) and the redistribution of wealth (the selected variables are: share of GDP for labour market policies (LMP); share of GDP for active labour market policies (ALMP); share of GDP for passive labour market policies (PLMP); and GDP per capita). At the beginning, we performed a simple cluster analysis by including in the analysis only three variables related to flexible employment (shares of part-time employment, temporary and self-employment) in total employment in a particular EU country. After using Ward's minimum variance method as a criterion in a hierarchical cluster analysis, we obtained four clusters of countries (see Chart 2) with distinctive differences concerning the use of the abovementioned flexible forms of employment. Chart 2: DENDROGRAM OF CLUSTERED EU COUNTRIES BY FLEXIBLE FORMS OF EMPLOYMENT, LFS 2010 910 Thus, the first cluster (Germany, Sweden, Denmark, Luxembourg, Bel- gium, Austria, United Kingdom, Ireland, Netherlands), which generally represents more developed countries from Northern and Continental Europe, shows a particularly high share of persons in employment employed parttime (27.1 %), whereas temporary employment and self-employment are underrepresented compared to the EU-27 average (see Chart 3). The second cluster, mostly including Central and Eastern European countries (Estonia, Lithuania, Latvia, Bulgaria, Hungary, Czech Republic, Slovakia, Malta), reveals an overall underutilisation of flexible forms of employment, with only self-employment (12.2 %) coming close to the EU-27 average (14.4 %). In contrast, the third cluster (Italy, Greece and Romania) represents countries with a high share of self-employed (25.4 %) among persons in employment and low shares of part-time and temporary employment. Finally, the fourth cluster, which is the most geographically dispersed and represents countries from the Mediterranean region as well as Central and Northern Europe (Spain, Poland, Portugal, Slovenia, Finland, France and Cyprus), chiefly uses temporary employment (19.9 %), with shares of self-employed (15.5 %) also above the EU-27 average. It could be argued that different social, cultural, political and institutional settings also influence a country's and cluster's labour market structure and the utilisation of particular forms of flexible employment. Nevertheless, since we will be unable to analyse all those factors in this short analysis, we will focus on one factor that is available and seems closely related to the particular cluster patterns of labour market flexibilisation14. This factor is the wealth of a country/cluster expressed through the variable GDP per capita. Chart 3: CLUSTERS OF EU COUNTRIES BY FLEXIBLE FORMS OF EMPLOYMENT, LFS2010 911 * in 1000 Data source: Eurostat, 2012 As Chart 3 shows, it seems that the 'choice' of a particular flexible employment form is closely related to the level of GDP per capita in a given cluster. Thus, it seems that the predominance of part-time employment in the first cluster is 'enabled' by the high level of development and high accumulation of wealth15 in those countries. On the other end of the scale, the low overall flexibility in the second cluster is accompanied by low GDP per capita. Between those two extremes lie two other clusters with predominant self-employment (with GDP per capita at 17,20016) and temporary employment (with GDP per capita at 21,371). We were initially tempted to conclude that there is indeed a significant influence of a country's/cluster's level of development on the predominance of a particular flexible form of employment as presented in Chart 4. 14 As Schmid (2009) suggests, the similar pattern of flexibility clusters strongly correlates with economic prosperity in terms of GDP per capita. 15 In this cluster, Luxembourg represents an outlier with GDP per capita of 79,500. The next closest country is Denmark with 42,500. Regardless of that, we chose to leave Luxembourg in the cluster since, even without it, the average GDP per capita in this cluster was around 35,000 - thus there was no particular difference. 16 With Romania as a negative outlier in the cluster (GDPper capita of 5,800). Chart 4: RELATIONSHIP BETWEEN THE PREDOMINANCE OF A PARTICULAR FLEXIBLE FORM OF EMPLOYMENT AND GDP PER CAPITA IN THE EU COUNTRY CLUSTERS low flexibility - (€10,412) temporary employment (€21,371) part-time employment (€39,333) 10000 15000 20000 25000 30000 35000 40000 45000 _ self-employment (€17,200) GDP per capita, in € Data source: Eurostat, 2012 However, after performing a bivariate correlation analysis (see the Appendix) between the different forms of flexible employment and GDP per capita, we could only partially confirm that presumption. This is partly because there was only a statistically significant strong positive correlation (.589 at p<0.001) between part-time employment and GDP per capita. The 912 correlation between self-employment and GDP per capita also pointed to the same presumption - a moderate negative correlation (-.302) (meaning lower shares of self-employment in countries with higher GDP per capita and vice versa), but it was not statistically significant. The correlation between temporary employment and GDP per capita was almost non-existent (.076). Chart 5: RELATIONSHIP BETWEEN GDP PER CAPITA AND THE SHARE OF GDP ALLOCATED TO LABOUR MARKET POLICIES, EU, 2010 Data source: Eurostat, 2012 In the case of the labour market, the impact of wealth is visible through the share of GDP allocated to (passive and active) labour market policies. As Chart 5 shows, there is a positive moderate statistically significant correlation (.455 at p<0.017) between GDP per capita and the share of GDP allocated to labour market policies. It should be pointed out that there is a significant difference in the direction of such allocation - thus among countries on the right side of Chart 5, in 2010 Denmark, Belgium, the Netherlands and Finland allocated more to active labour market policies, while Spain and Ireland allocated more to passive labour market policies (mostly unemployment benefits). Here we should again mention Calmfors (see footnote 7). Our analysis confirms the positive correlation between higher unemployment and a bigger share of passive labour market policies, but we cannot confirm the outcome of this relationship. Similarly, we can confirm that higher investments in active labour market policies are correlated to lower unemployment rates. At the same time, it is noted that different forms of flexible employment generate different levels of risk for in-work poverty17. Thus, according to the Eurostat data from 2010, the higher shares of part-time employment (see Chart 6) characterising the first cluster of more developed European coun- 913 tries reduces (correlation at -.291, although not statistically significant) the in-work at-risk-of-poverty rate. Chart 6: RELATIONSHIP BETWEEN PART-TIME EMPLOYMENT AND IN-WORK AT-RISK-OF-POVERTY RATE, EU, 2010 bU NI 30 - eAt J. £ )e F It Lu ♦ | rp ♦ _L t G 0 » Hu^ Sk ♦ Bg ♦ Lv ♦ 4 ♦ 1 0 1 2 1 4 1 6 1 8 2 In-work at-risk-of-poverty rate Data source: Eurostat, 2012 17 The share of employed persons aged 18 years or over with an equivalised disposable income below the risk-ofpoverty threshold, which is set at 60 % of the national median equivalised disposable income (after social transfers). In this constellation, temporary employment (see Chart 7) has a neutral effect on in-work poverty (correlation -.001, not statistically significant). Chart 7: RELATIONSHIP BETWEEN TEMPORARY EMPLOYMENT AND IN-WORK AT-RISK-OF-POVERTY RATE, EU, 2010 ::::: Pt ♦ « E 4 5::: :::::: NI ::::: V Se i::: ::::: r A1Hu & it O G. ♦ s4 nk ^r Et -Lv ♦ Lu ♦ i 4 Ro ♦ 0 2 4 6 8 10 12 14 16 18 20 In-work at-risk-of-poverty rate Data source: Eurostat, 2012 914 Finally, self-employment (see Chart 8) as a form of flexible employ- ment increases the in-work risk of poverty as it is statistically significantly moderately correlated (0.463 at p<0.015) with the In-work at-risk-of-pov-erty rate. Chart 8: RELATIONSHIP BETWEEN SELF-EMPLOYMENT AND IN-WORK AT-RISK-OF-POVERTY RATE, EU, 2010 cz f V H r- It «1* ♦ PI l-c G ♦ :.: ;:; f n :±: ft t ^ïï IZS* — -V h ♦ ♦ Lt ♦ Zt'.'.t'". i ; : ::;.::;:: : 0 2 4 6 8 10 12 14 16 18 20 In-work at-risk-of-poverty rate Data source: Eurostat, 2012 It thus seems that self-employment, as one of the flexible forms of employment, brings particularly high risks of entering into poverty. One can relate this finding with the high mortality rate of small businesses (high job and employment insecurity for the self-employed) and the fact that self-employed persons are usually expected to finance their own security (to contribute to pension and health funds). High levels of accumulated wealth also enable more developed countries (the first cluster) to compensate (through various forms of the flexi-curity concept) the risks of high flexibility and high risks of poverty with a high redistribution of wealth through various forms of transfers (various social security benefits and assistance) - even though the relationship is less clear for the other clusters. More precisely, countries from the first cluster have the highest at-risk-of-poverty rate before transfers (Chart 9), but manage to reduce it to one-half after transfers (and to the lowest level among the clusters). On the other hand, while the second and fourth clusters have similar rates before and after transfers, the second one manages to reduce the risk to the same levels with half of GDP per capita as in the fourth cluster, thus showing greater efficiency. The small difference between at-risk-of-poverty rates before and after transfers points to the low level of efficiency of transfers in Italy, Greece and Romania. Chart 9: RELATIONSHIP BETWEEN AT-RISK-OF-POVERTY RATES BEFORE AND AFTER SOCIAL TRANSFERS AND GDP PER CAPITA FOR PARTICULAR EU COUNTRY CLUSTERS, 2010 915 ■ At-risk-of-poverty rate before transfers Data source: Eurostat, 2012 Conclusions The primary objective of this article was to highlight the difficult regulation of the flexibility-security balance in modern European labour markets. As labour markets become ever more dynamic and fluent due to processes such as globalisation and flexibilisation, different forms of security are becoming more and more compromised for an increasing number of workers and thus for European societies too. In the last decade, the European Union has been trying with varying levels of success to address this problem by introducing the flexicurity concept, trying to preserve at least income security and to thus compensate the risks of those working in flexible forms of employment and the unemployed through the redistribution of wealth. On the other hand, the flexicurity concept has various shortcomings - an increasing burden on public finances being one of the most important. For many experts, implementation of the concept was stopped or even reversed by the current economic and financial crisis and related austerity measures, even though the current situation (increasing unemployment and poverty rates) demands the opposite. The current global economic crisis has thus emphasised the problems related to the labour market and especially 916 highlighted the need for more efficient social protection time when interna- tional competition demands even more flexible labour markets. While addressing the EU mostly on the general level, we must acknowledge that there are important differences among EU members in their starting positions, the labour market structure and the ability to address problems emerging in the current situation. In our analysis we confirmed all three hypotheses posited at the beginning. We thus confirmed the presence of different 'models' of flexibilisation in the EU since EU countries are using different forms of flexible employment as the primary source of flexibilisation. Following a similar analysis performed 10 years ago (Ignjatovic, 2002), we may conclude that developed EU countries predominantly use part-time employment as a form of flexible employment. It offers higher job security and lower income than temporary employment and self-employment. On the other hand, less developed new members from Central and Eastern Europe do not have high shares of flexible forms of employment at all. We also found that the 'selection' of a particular form of flexible employment is strongly correlated, especially in the case of part-time employment, with the level of wealth accumulated in a particular society (measured by GDP per capita). We also determined that more developed countries (with high levels of accumulated wealth) can afford to implement the flexicurity concept since they can more efficiently compensate the risks of higher flexibility and in-work poverty with higher levels of wealth redistribution than other clusters of countries. Further, we confirmed our hypothesis that different forms of flexible employment produce different levels of risk related to entering into poverty (measured by the in-work at-risk-of-poverty rate). Our analysis thus proved that there is a relatively high correlation between shares of self-employment and the in-work at-risk-of-poverty rate. The recent developments in European labour market(s) and societies also form part of more profound long-term changes concerning the emphasis a particular form of security has for members of European societies. As modern labour markets are becoming more dynamic, the probability of retaining one's job for a longer time is becoming smaller for a growing proportion of the workforce. The flexicurity concept has tried to encourage policy makers and the workforce to shift the emphasis of security from job security towards employment security. As the European Commission (2007: 9) suggests: "Individuals increasingly need employment security rather than job security, as fewer have the same job for life". But here again the differences arise in the concept's understanding. The Commission emphasised external flexibility by defining job security as a job with a single employer and employment security as the potential to hold jobs with multiple employers, i.e. employ-ability security rather than employment security, according to Auer (2010). On the other hand, Auer puts the emphasis on internal flexibility by defining 917 job security as related to the probability of workers retaining employment in their current job, and employment security to retaining a job with their current employer. According to Auer, this shift already occurred in the 1980s when the first wave of flexibilisation hit European labour markets. What is even more important from the long-term perspective is, as Auer suggests, that yet another shift is needed: "The decisive and critical shift is thus not from job security to employment security, but towards what can be called labour market security. Labour market security implies that security for workers in today's labour markets cannot stem from job - and employment security alone. It has to be complemented by additional layers of security" (Auer, 2010: 381). Relying on the transitional labour market theory Auer suggests that, from the perspective of further development of the labour market and the lifelong professional trajectory of an individual, labour market security should offer security regardless of the individual's position (employment, unemployment, inactivity, training, partial work) or transition18 at a particular point of time, rightly due even more to the dynamic nature of the labour market. 18 Transitional labour market theory concentrates on five critical events: transitions from education to employment, transitions from one job to another, transitions between employment and unemployment, transitions between private household activities and gainful work, and eventually transitions from employment to retirement. Chart 10: HISTORICAL EVOLUTION OF WORKERS' SECURITY IN LABOUR MARKETS Job security Employment security Labour market security Source: Auer, 2010: 381 Chart 10 illustrates this shift schematically along a time axis (in the last 30 years and the future). Such a transformation of labour-market-related risks toward more life-related risks demands, as Schmid (2009a: 12) puts it, "a recalibration of the conventional social security systems. Three general strategies of social risk management need thereby to be distinguished: prevention of risks, mitigation of risks and coping with risks". Schmid proposes (Schmid, 2009: 22) a "way in which to extend the social insurance principle to a broader set of life course risks than unemployment would be to establish a system of work-life insurance. Such a system would build on three pillars: first, a universal minimum income guarantee that ensures a life without persistent poverty; second, the extension of unem-918 ployment insurance to employment insurance; third, private or collectively negotiated insurance accounts targeted especially to life course risks such as lifelong learning accounts, time-saving accounts or care-leave systems. Governments could join such ventures at various levels through tax subsidies, standard setting and co-financing partners". It is difficult to predict how soon such a transformation of the current social security systems will occur, if at all. REFERENCES Andersen, M. Torben, Bosch, Nicole, Deelen, Anja, Euwals, Rob (2011): The Danish flexicurity model in the Great Recession. VOX. Available at: http://www.voxeu. org/index.php?q=node/6332. Auer, Peter (2006): Protected Mobility for Employment and Decent Work: Labour Market security in a Globalized World. Journal of Industrial Relations, 48: 21 Auer, Peter (2010): What's in a Name? The Rise (and Fall?) of Flexicurity. Journal of Industrial Relations 52: 371. Auer, Peter (2010a): Does flexicurity work in economic crises? Draft version, May 2010 ILO. Begg, Iain (2012): Is 'Flexicurity' Post-Crisis Europe's New Social Model? Bloomberg, http://www.bloomberg.com/news/2012-04-15/is-flexicurity-post-crisis-europe-s-new-social-model-.html. Bredgaard, Thomas, Larsen, Flemming, Kongsh0j Madsen, Per (2005): The flexible Danish labour market - a review. CARMA Research Papers 2005: 01. Burroni, Luigi, Keune, Maarten (2011): Flexicurity: A Conceptual Critique. European Journal of Industrial Relations, 17: 75. Calmfors, Lars (2007): Flexicurity - An Answer or a Question? European Policy Analysis, November Issue 6-2007, Swedish Institute for European Policy Studies Cook, B. (2008): Active Labour Market Policies in the Neo-Liberal Era. Centre of Full Employment and Equity, Working Paper No. 08-03. Cornelifien, Thomas (2007): Flexibility and Security in the Labour Market Institutions and Perceived Job Security in Europe. ILO, Subregional Office for Central and Eastern Europe, Budapest. Council of the European Union (2007): Towards Common Principles of Flexicurity - Draft Council Conclusions. Esping-Andersen, G0sta (1990): The Three Worlds of Welfare Capitalism. Princeton University Press, New Jersey. European Commission (2007): Towards Common Principles of Flexicurity: More and better jobs through flexibility and security. European Foundation for the Improvement of Living and Working Conditions (2010): Extending flexicurity - The potential of short-time working schemes. ERM Report 2010. European Foundation for the Improvement of Living and Working Conditions (2010a): Flexicurity: perspectives and practice. Foundation Findings. European Foundation for the Improvement of Living and Working Conditions (2012): The second phase of flexicurity: An analysis of practices and policies in 919 the Member States. European Commission (2007): Towards Common Principles of Flexicurity: More and better jobs through flexibility and security. EUROSTAT (2012): Statistics database, Available at: http://epp.eurostat.ec.europa. eu/portal/page/portal/statistics/search_database. Gazier, Bernard (2007): »Making Transitions Pay«: The »Transitional Labour Markets« Approach to »Flexicurity«. Gray, Anne (2005): 'Unsocial Europe'; The Future of the European Model in the Face of Globalisation. Workshop Stockholm, Arena Centre for European Studies, Available at: http://www.sv.uio.no/arena/english/research/projects/ cidel/old/WorkshopStockholm/Gray.pdf . Heyes, Jason (2011): Flexicurity, Employment Protection and the Job Crisis. Work, Employment & Society, 25: 642. Ignjatovic, Miroljub (2002): Družbene posledice povečanja prožnosti trga delovne sile. FDV, Znanstvena knjižnica 47. Jensen, H. Per (1989): Employment and Unemployment Policies and the Functioning of the Labour Market in a Comparative Perspective. In Acta Sociologica Vol. 32, No. 4, Norwegian University Press, Oslo. J0rgensen, Henning (2011): Danish »Flexicurity« in Crisis - Or Just Stress-tested by the Crisis? International Policy Analysis, Friedrich Ebert Stiftung. Klindt, Mads Peter (2008): 'Old wine in a new bottle' or 'something new under the sun'? A Comparative Assessment of the European Commission's Flexicurity-concept. Paper prepared for CARMA's (Centre for Labour Market Research) 25th Anniversary Conference, October 9-10 2008, Aalborg University Denmark Mallon, Mary (1998): The Portfolio Career: Pushed or Pulled to it? Personnel Review, Vol. 27 No. 5, MCB University Press. Mitchell, William and Muysken, Joan (2008): Full employment abandoned: Shifting sands and policy failures. Centre of Full Employment and Equity, Working Paper No. 08-01. Muffels, Ruud and Ruud Luijkx (2008): Labour Market Mobility and Employment Security of Male Employees in Europe: 'Trade-Off' or 'Flexicurity'? Work, Employment & Society, 22: 221. O'Reilly, Jacqueline, Lain, David, Sheehan, Maura, Smale, Bob, Stuart, Mark (2011): Managing Uncertainty: The Crisis, its Consequences and the Global Workforce. Work, Employment & Society 25: 581. Rodgers, Gerry (2007): Labour Market Flexibility and Decent Work. DESA Working Paper No. 47. Ross, Andrew (2008): The New Geography of Work: Power to the Precarious? Theory, Culture & Society, 25: 31. Rus, Veljko (1990): Socialna država in družba blaginje. Domus, Ljubljana. Sapir, André (2005): Globalisation and the Reform of European Social Models. Bruegel, Brussels. Sapir, André (ed.) (2009): Europe's economic priorities 2010-2015, memos to the new Commission, Brussels: Bruegel. Schmid, Günther (2008): Full Employment in Europe: Managing Labour Market 920 Transitions and Risks. Edward Elgar. Schmid, Günther (2009): Transitional Labour Markets, from Theory to Policy Application. Transitional Labour Markets and Flexicurity: Managing Social Risks over the Life Course. CES Working Papers 2009.75. Schmid, Günther (2009a): Theory of Transitional Labour Markets and »Flexicurity«. Lessons for Transition and Developing Countries. Paper presented to the European Training Foundation, Torino, 14 May 2009. Schmid, Günther (2011): Non-standard Employment in Europe: Its Development and Consequences for the European Employment Strategy. German Policy Studies, Vol. 7. No. 1. Standing, Guy (1997): Globalization, Labour Flexibility and Insecurity: The Era of Market Regulation. European Journal of Industrial Relations, Vol. 3, No. 1, SAGE, London. Standing, Guy (2011): The Precariat. The New Dangerous Class. Bloomsbury Academic, London. Str0by Jensen, Carsten (2011): The Flexibility of Flexicurity: The Danish Model Reconsidered. Economic and Industrial Democracy, 32: 721 Tangian, Andranik (2010): Not for bad weather: Macroanalysis of flexicurity with regard to the crisis. Working paper 2010.06, European Trade Union Institute. Tangian, Andranik (2010a): Not for Bad Weather: Flexicurity Challenged by the Crisis. ETUI Policy Brief Issue 3/2010. Viebrock, Elke and Clasen, Jochen (2009): Flexicurity and Welfare Reform: A Review. Socio-Economic Review, 7. Wilthagen, Ton, Tros, Frank (2004): The Concept of »Flexicurity«: A new approach to regulating employment and labour markets.Transfer: European Review of Labour and Research 2004 10: 166. SAGE. Wilthagen, Ton (2008): Mapping out flexicurity pathways in the European Union. Available at: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1118725. Appendix: Table 1: STATISTICALLY SIGNIFICANT CORRELATIONS BETWEEN SELECTED VARIABLES At-risk-of-poverty rate after transfer In-work at-risk-of-poverty rate Self-employment Active labour market policies Passive labour market policies Employment rate Unemployment rate Long-term unemployment rate GDP per capita At-risk-of-poverty rate before transfer .396* At-risk-of-poverty rate after transfer .802** -.512** .534** .425* In-work at-risk-of-poverty rate .463* -.424* Part-time .524** .419* .704** -.443* -.502** .589** Active labour market policies .638** .420* .475* Employment rate -.571** -.667** .526** Unemployment rate .905** -.468* Long-term unemployment rate -.510** 921 *correlation is significant at the 0.05 level (2-tailed); ** correlation is significant at the 0.01 level (2-tailed)