Abstract
This article investigates employees’ attitudes towards job protection legislation and attitudinal differences between employees with different levels of job security. National surveys from three Nordic countries, using different measures of insider–outsider positions in the labour market, do not support the assumption that outsiders (those with insecure jobs) prefer laxer job protection legislation. On the contrary, workers in secure jobs seem more likely to prefer laxer regulation.
Keywords
Introduction
Employees differ in the resources they possess and their exposure to labour market risks. Consequently, they may have different interests and orientations towards social and political issues. This article studies employees’ preferences regarding job security regulation in Finland, Norway and Sweden and asks whether the preferences of insiders (those with high job security) differ from those of outsiders (those with insecure employment).
In recent decades, many European countries have liberalized their employment protection legislation (EPL) (OECD, 2013): their laws and rules governing employers’ ability to dismiss individual workers with permanent contracts; regulations affecting the scope and cost of collective dismissals; and legislation governing the use of temporary work contracts or agency workers (Venn, 2009). However, there have been few studies of employees’ attitudes towards changes in job security regulations. This is surprising because these regulations affect the distribution of goods and burdens and have power implications (Elster, 1992; Emmenegger, 2014; Mahoney and Thelen, 2010: 9). EPL can empower trade unions; moreover, like any institution that regulates and patterns action, it can also distribute resources to some actors in preference to others. The distributional consequences of job security regulations may maintain or create cleavages between labour market groups (defined in terms of socioeconomic class, specific jobs or types of employment contract).
Studies have shown that class can significantly influence labour market policy preferences since class structures can express and shape the distributional conflicts of power in society (Bengtsson et al., 2013; Edlund, 2007; Korpi, 1980). Class is also likely to be a structuring principle with respect to policy preferences, especially in the Nordic countries with their well-established class cleavages (Brooks and Svallfors, 2010). Class and attitudes are known to be linked; Rueda and colleagues have repeatedly argued, using insider–outsider theory, that labour market insiders and outsiders have conflicting interests concerning job security regulations (Lindvall and Rueda, 2014; Rueda, 2005, 2006). Insiders may try, both individually and through unions, to secure their jobs and wage advantages using various power resources, for instance by invoking seniority rules and other regulations governing dismissal (Lindbeck and Snower, 2001: 168). From this perspective, the use of power by insiders keeps wages higher than the market-clearing level and increases unemployment levels since outsiders willing to work for lower wages remain unemployed. In other words, high wages and high job security create barriers for outsiders seeking to obtain more secure employment.
It is thus assumed that insiders care more about their own job security than the employment of outsiders, while outsiders have an interest in lax job protection legislation because it reduces mobility barriers and thus offers more opportunities for labour market integration (Rueda, 2005: 64). If insiders and outsiders have divergent interests regarding job security, reforming job security regulations may influence the configuration of interests and alter public support for further labour market and social policy reforms (Blekesaune and Quadagno, 2003).
From an international perspective, the small Nordic labour markets seem to display high levels of labour mobility, with wage floors set through centralized negotiations. This limits low-wage competition and forces companies with low productivity out of the market. The reallocation of labour takes place within labour markets with high employment levels supported by generous social policies. This combination of labour market flexibility with robust social security has been described as ‘flexicurity’ (Burchell, 2009: 368; Burroni and Keune, 2011; Madsen, 2004; Muffels and Luijkx, 2008: 225; Wilthagen and Tros, 2004). In contrast, insider–outsider theory is based on a rather segmented image of the labour market, which is assumed to feature insiders with secure jobs and outsiders who have precarious jobs or are unemployed, with only limited mobility between these segments. Countries with dualized labour markets such as Germany and France have very different levels of labour market equality to the Nordic countries, which score highly on measures of egalitarianism (Thelen, 2014: Chapter 1). Segmentation of the labour market into insider and outsider interests may thus be less valid for the Nordic countries; but Rueda does not distinguish between dualistic and egalitarian labour markets, claiming that his theories retain their explanatory power even in the Nordic context (Lindvall and Rueda, 2014: 463–464). We aim to test this claim by investigating differences in attitudes to job security between insiders and outsiders in the Nordic countries.
Another objective is to investigate differences among the Nordic countries with respect to insiders’ and outsiders’ attitudes to job security regulation. Although all Nordic countries have egalitarian labour markets and with forms of flexicurity, they also have different regulations governing job protection and differ in their levels of mobility from temporary to permanent employment; Sweden, in particular, exhibits more pronounced dualistic tendencies than Norway or Finland (Thelen, 2014: 175). These national differences make it possible to investigate the relationship between labour market dualization and the divergence of insiders’ and outsiders’ attitudes to job protection regulations. We therefore consider whether differences between insiders’ and outsiders’ attitudes are more pronounced in Sweden than in Norway and Finland. Our original study design also included Denmark, but lack of funding for a Danish survey prevented this.
The following section discusses the attitudes of insiders and outsiders in the Nordic context and how these two groups are best operationalized in labour markets such as those of the Nordic countries. We then present the institutional frameworks relevant to the questions posed, followed by the data used in the analysis, the methodology and the results obtained.
Insider–outsider attitudes towards employment protection policies
Insiders and outsiders in labour markets
According to insider–outsider theory, the labour market consists of two segments with conflicting interests and hence also different attitudes towards job security: insiders, defined as those ‘employed full-time with a permanent job or as those with part-time or fixed-term jobs who do not want a full-time or permanent job’, and outsiders, ‘who are unemployed, employed full-time in fixed-term and temporary jobs (unless they do not want a permanent job), employed part time (unless they do not want a full-time job) and studying’ (Rueda, 2005: 63).
Insider and outsider positions are thus defined by individuals’ objective position in the labour market. Several studies have shown that workers holding temporary employment contracts face greater risk of job loss than those with permanent contracts because temporary employment provides poorer job protection (Burgoon and Dekker, 2010; Ellonen and Nätti, 2013; Svalund, 2013b). Part-time work is often considered a ‘non-standard’ and less secure employment relationship (Kalleberg, 2000) but has traditionally been rather normalized for large sectors of the labour market in Norway and Sweden. This is less true in Finland although the percentage of female part-time workers in Finland has grown in recent years (Kangas and Saloniemi, 2013: 55).
Insider–outsider theory assumes limited labour market mobility; but as noted above, this is not true of Nordic labour markets. For theorists of flexicurity, generous unemployment benefits (income security) and good opportunities to find a new job (employment security) are assumed to reduce the stress of losing a particular job and also encourage voluntary employee mobility, making it possible to relax job security regulations (European Commission, 2007; Wilthagen and Tros, 2004). Critics of the flexicurity approach argue that flexibility for employers and security for employees are incompatible goals and that liberalizing job security regulations create a detrimental trade-off for employees or groups of employees, who are left with less job security (Burroni and Keune, 2011; Tangian, 2008).
In labour markets such as the Nordic ones, with intermediate levels of job security and high mobility, differences in job security between insiders and outsiders may not be captured well by the objective positions used in Rueda’s definition. In such countries, it may be more appropriate to link labour market integration and job security to individuals’ subjective job security and employment (in)security: their self-perceived risk of losing their job and their opinion on the likelihood of finding new, similar work (Vulkan et al., 2015). We therefore combine Rueda’s understanding of insiders and outsiders based on objective labour market positions with an alternative categorization based on individuals’ subjective assessments of their job (in)security and likelihood of obtaining new employment (De Witte, 2005).
Differences in attitudes towards job regulation between insiders and outsiders
There is limited empirical research on insider–outsider preferences regarding job security. By analysing data from the Eurobarometer surveys (including Finland, Norway and Sweden), Rueda (2005: 70) found that insiders support EPL policies more than outsiders. However, the difference was rather small; the mean agreement scores for the item ‘job security is very important’ were 6.4 and 5.4 respectively, on a scale ranging from 0 to 10 (Rueda, 2005: 65). Svalund (2013a) found that in Norway, employees with temporary jobs, part-time workers and those experiencing subjective job insecurity are less likely to desire relaxation of regulations on dismissing employees or hiring temporary workers, and that insiders’ views on seniority rules are identical to those of outsiders (about 23% of both groups disagree with the rules). Guillaud and Marx (2014: 6), in France, asked a group of respondents whether they would support the establishment of a single employment contract, which would relax job security regulations, and found no difference in attitudes between those with permanent and with temporary work contracts; Rueda’s arguments were only confirmed among the unemployed. Emmenegger (2009) analysed ISSP data from 1997 where respondents were asked how important they considered job security. The results provide limited support for Rueda’s theory: unemployed individuals and insiders supported job security regulations more than those outside the labour market (the non-employed), while part-time and temporary employees were more critical of them than the non-employed (Emmenegger, 2009: 138).
Institutional frameworks
Employment protection regulation
Employees’ attitudes to job security regulation may be related to its strength. Furthermore, liberalization of the regulation of temporary work has led to dualization in some countries, with most employees on secure permanent contracts while others remain in precarious temporary employment (Emmenegger, 2012; OECD, 2013). Although all the Nordic countries have moderately strict regulations governing the dismissal of employees with permanent contracts, there are clear differences regarding temporary employment according to the crude but widely used OECD index. According to this index, Sweden has the most lax regulation of temporary contracts of the three considered here, while Norway has the strictest. However, Sweden also has the strictest regulations regarding individual and collective dismissals (Figure 1). Sweden’s combination of strict regulation of individual and collective dismissal with lax regulation of temporary contracts resembles the situations in Germany and Great Britain, countries with large insider–outsider divides. Thus, significant insider–outsider divides are more likely in Sweden than in Finland and Norway.
EPL strictness regarding dismissals (collective and individual) and fixed-term contracts, 2011.
The regulations governing the dismissal of individuals in these three countries are similarly strong according to the OECD (FI = 2.38, NO = 2.23 and SE = 2.52) (OECD, 2012): none has detailed regulations governing individual dismissals, but collective redundancies are tightly regulated. Employers facing lower demand because of economic or production-related issues can dismiss employees with no costs other than paying them during their notice period: severance pay is not legally required. Legislation or collective agreements in these countries stipulate a period of notice before permanent employment contracts can be terminated, varying between 14 days and 6 months depending on the prior duration of employment (Berglund et al., 2010). The regulations also have other dimensions, with more variation between countries.
The rules on selecting employees to be dismissed in collective redundancies are important and differ between the Nordic countries. Discriminatory decisions are forbidden in all three countries. Finnish legislation states that dismissals must be based on financial and production-related grounds related to the employee or the company, but no criteria for prioritization are defined. Norwegian laws are similarly unspecific. The Basic Agreement between the Norwegian Confederation of Trade Unions (LO) and the Confederation of Norwegian enterprises (NHO) states that seniority shall apply, but deviation is permitted with due cause. In contrast, Swedish law requires employers to uphold the ‘last in first out’ principle (Lagen om änstellningsskydd), with the prerequisite that the preferred employees must be qualified to perform the work. However, deviations are possible based on collective agreements. The use of the seniority principle might influence employees’ attitudes since it benefits those with seniority and may thus protect older, more vulnerable workers (Elster, 1992: 75). These legal differences, as well as case studies from manufacturing and construction, suggest that employers have much more control over dismissal selection in Finland than in Norway or Sweden (Svalund et al., 2013; Svalund and Kervinen, 2013).
Past experience of EPL reforms can also affect employees’ attitudes towards liberalization of regulations. In recent years, individual and collective dismissal regulation has only changed slightly in Finland and Sweden, and there have been no significant changes in Norway. However, Sweden’s law on temporary contracts was liberalized in 1997, 2003 and 2007, making it among the most liberal in the EU. These changes have become prominent in the Swedish public debate, with business representatives and some members of centre-right parties claiming that the current labour laws are still too strict (Christensen and Engelbrecht, 2012). Others have criticised the practice of successive fixed-term contracts which keep many employees in long-term temporary employment. Demands for EPL relaxation are also made in Finland, where the main legislative changes concerning fixed-term work have been to protect the position of fixed-term employees rather than to make temporary contracts more attractive. For example, in 2008, a new law stating that fixed-term contracts could only be offered under strictly defined conditions was introduced. However, the status of temporary employees has recently become a topic of public dispute because the new government programme proposes to allow temporary employment relationships of less than a year without special justification (Valtioneuvoston Kanslia, 2015: 14). In Norway, there have only been minor adjustments in the rules regarding the use of temporary contracts, but the issue has been prominent in the public debate. In 2004, the centre-right government proposed to change the Work Environment Act (Arbeidsmiljøloven), liberalizing the use of temporary contracts. The law was passed after fierce debate, but the 2005 election brought in a new centre-left government which reversed the changes and made the existing regulations slightly stricter. The issue remained fiercely debated, and the right-wing government elected in 2013 again liberalized the rules in the spring of 2015 despite strong opposition.
Income security
The flexicurity approach defines a triangulation between employees’ possibilities of getting a new job (employment security), protection of the current job and income security. Employees’ preferences regarding job security may thus be influenced by the income security and unemployment benefit systems as well as by employment protection rules. Unemployment insurance is mandatory in Norway but voluntary in the other Nordic countries. Recent changes in the Swedish ‘Ghent’ system, before and during the economic crisis of 2008–2009, have increased the number of uninsured workers and reduced the compensation level (Dølvik et al., 2015: 261–262). The net replacement rates for an average worker were in 2010 lowest in Sweden, intermediate in Finland and most generous in Norway (Ferrarini et al., 2012: 32). The replacement rate in Norway has been stable over the last decade but has decreased since 2002 in Sweden (Bengtsson and Berglund, 2012; Ferrarini et al., 2012). The Finnish replacement rates decreased from 2002 to 2008, then increased again (Vulkan et al., 2015: 38).
Dualization within the egalitarian Nordic labour market?
As insider–outsider theory is premised on the idea of segmented labour markets, the proportion of employees holding temporary contracts and involuntarily working part-time is important. Reflecting national differences in regulation, 8 percent of all employees in Norway were in temporary employment in 2011, compared to 16.5 percent in Sweden and 15.5 percent in Finland (Eurostat, 2014). Both Finland and Sweden saw large increases in the number of temporary contracts in the early 1990s, whereas the number of temporary contracts in Norway has been falling (Berglund and Esser, 2014: 64–65).
Part-time workers are considered outsiders in Rueda’s studies (2005, 2006). In 2011, 27.3 percent of the labour force in Norway had a part-time job, 25.3 percent in Sweden and 13.9 percent in Finland. Some of these workers would have preferred longer working hours: 4.5 percent in Sweden, 3 percent in Norway and Finland (De La Fuente, 2011: 6, Table 2).
Studying mobility patterns inside and outside the labour market during the period 2000–2006, Berglund and Furåker (2011: 125) found that Norway had the highest mobility from unemployment to employment, while that from employment to unemployment was highest in Finland. Swedish rates were somewhat lower. In an analysis of the same data, Svalund (2013b) found that the possibility of moving from temporary to permanent employment within 1 year was higher in Norway and Finland than in Sweden, while the probability of moving from a temporary job to unemployment was higher in Sweden than in the other two countries.
These data indicate that Sweden’s labour market is the most dualized of the three countries, followed by Finland, where the duration of temporary employment is longer than in Norway. The combination of lower mobility and more temporary employment may mean that insecure labour market positions are less likely to be considered transitional. Thus, on the basis of insider–outsider theory, one might expect more pronounced differences in attitudes towards job protection in Sweden, where outsiders’ objective positions seem more permanent, than in Finland and, particularly, Norway. And finally, as mentioned, unemployment benefits are also least generous in Sweden, thus income insecurity in case of job loss is highest there.
Data and methodology
The data used in this article are derived from three separate postal surveys sent out to employees participating in the regular Labour Force Survey (LFS). A number of LFS variables could therefore be added to the dataset to supplement the responses to the postal surveys. The analysis focuses on respondents aged between 19 and 64 years.
The surveys were conducted during 2010–2011 (Sweden autumn 2010, Finland winter 2010–2011 and Norway spring/summer 2011). Response rates were 40 percent in Norway (1634 responses), 53 percent in Finland (2252 responses) and 54 percent in Sweden (2023 responses). Regarding the representativeness of the datasets, there is no evidence of strong biases. People under the age of 24 years, the temporarily employed, and men in the Finnish data, are somewhat overrepresented among the non-respondents. However, the samples depict the eligible population in each country relatively adequately.
We analyse the data using logistic regression, a method appropriate for the treatment of categorical data that makes it possible to control simultaneously for many variables.
Dependent variable
The dependent variable measures attitudes towards job security regulation. Respondents in Norway and Sweden were informed that the relevant national legislation limits employers’ ability to dismiss employees and asked: ‘what do you think of this?’ Three responses were possible (in addition to ‘do not know’ in Norway): ‘it should be easier for employers to dismiss employees’, ‘the current rules are good’ and ‘it should be harder for employers to dismiss employees’. In the regression analysis in Table 2, these answers were dichotomized by comparing ‘it should be easier for employers to dismiss employees’ (1) with the other two categories (and ‘do not know’ in Norway). Respondents in Finland were asked a different question: ‘all in all, do you think that statutory employment laws should provide more or less security for the employed?’ The possible answers were ‘they should provide much more security’ (1), ‘they should provide somewhat more security’ (2), ‘they are good as is’ (3), ‘they should provide somewhat less security’ (4), ‘they should provide much less security’ (5) and ‘they should be completely removed’ (6); ‘do not know’ was also an option. The Finnish question was also given to the Swedish respondents, and we therefore compare Finland to Sweden. In the regression analysis in Table 2, the values for responses 4–6 are combined under the heading ‘less security’, while 1–3 and ‘don’t know’ comprise the rest. The Pearson R value for the two questions in the Swedish survey indicates a moderate level of correlation (R = 0.58). The relationship between the two questions is noteworthy with regard to flexicurity theory because while both questions concern the employment relationship, the question asked in Finland and Sweden stresses employee job security, whereas that in Norway and Sweden emphasises employer flexibility and its potential reduction by job security.
Independent variables
The focal independent variables in this study are those operationalizing insiders and outsiders. Following Rueda, who classifies employed outsiders as those who involuntarily hold a temporary contract or part-time job, we define insiders and outsiders by the nature of their employment contracts (permanent for insiders and temporary for outsiders) and their working hours (full-time for insiders and part-time for outsiders). However, as in previous research (Guillaud and Marx, 2014; Rueda, 2005), we could not distinguish between those who were in such positions voluntarily and involuntarily.
Subjective job insecurity was based on the question: ‘how do you assess the risk that you will lose your job in the coming 12 months?’ The possible responses were ‘very high’, ‘quite high’, ‘neither high nor low’, ‘quite low’ and ‘very low’. The first two alternatives were combined as ‘there is a risk’ (1, outsiders) and the latter as ‘there is little risk’ (0, insiders). It was also possible to answer ‘don’t know’; these respondents were disregarded.
Proponents of flexicurity suggest that the likelihood of obtaining another job influences employees’ attitudes towards job security regulation, thus also separating insiders and outsiders. Subjective employment (in)security was investigated using the question: ‘in general, what do you think of your current opportunities for finding another job that is comparable to or better than your current one?’ The possible answers were ‘good, ‘quite good’, ‘neither good nor poor’, ‘quite poor’ and ‘very poor’. The responses were separated into two groups: the first three were combined in the ‘employment security’ category (0, insiders), while the last two were combined as ‘employment insecurity’ (1, outsiders). It was also possible to answer ‘don’t know’; these respondents were disregarded.
We included a number of control variables such as gender and age. Age was included because it can increase concern over economic risks and because previous research has shown that older employees experience higher levels of job insecurity (Hartley, 1991) and are more inclined to support job security regulation (Elster, 1992: 75; Guillaud and Marx, 2014: 1181). Employers must consider seniority during dismissal selection in Norway and Sweden, which largely helps secure the jobs of older employees. We also controlled for occupation, income and union membership. Occupation is used as an indicator of socioeconomic class position. Income is also expected to reflect socioeconomic status and ideology and thus to correlate with class (Bengtsson et al., 2013). Union membership tends to correlate positively with more egalitarian and redistributive attitudes concerning public policy (Pontusson, 2013). Education was also included as a control variable; it has class-related aspects such as cultural and human capital but is also related to social status and cognition. Earlier research has shown that the class components of education can be accounted for by other factors such as income (Brooks and Svallfors, 2010). We therefore expected education to have only a modest effect on attitudes towards policy issues such as EPL. Finally, company size might matter (Svalund, 2013a) and was thus controlled for.
The analysis first describes the distribution of the focal independent variables (Table 1). Logistic regressions are then used to estimate the effects of the independent variables (Table 2) using three separate models.
Preferences regarding job security legislation (%).
Should it be easier, as it is, or more difficult for employers to dismiss employees?
Should the law provide less, as it is, or more security for those employed?
Binominal logistic regression, preferences regarding job security legislation: Odds ratios.
It should not be easier to dismiss employees (0); it should be easier to dismiss employees (1).
The law should not provide less security for those employed (0); it should provide less security (1).
Low = under €2000 or approximate equivalent in NOK or SEK; 2 = €2000–2499; 3 = €2500–2999; 4 = €3000–3999; high = over €4000.
Coefficients that differ significantly (p < .05) between country pairs are shown in bold. Significance levels: *<5 percent, **<1 percent, ***<0.1 percent.
Descriptive analysis
Table 1 shows the preferences of employees in Sweden, Finland and Norway regarding job security legislation. There are large differences across the three countries: only 24 percent of respondents in Norway state that it should be easier to dismiss employees, compared to 37 percent in Sweden. In addition, the percentage of respondents in Norway saying that EPL regulations should remain unchanged is higher than in Sweden. Table 1 also shows that while 10 percent in Finland think that the law should provide less security for the employed, 17 percent think so in Sweden. The proportion of respondents saying that the law should provide more security in Finland is higher than in Sweden. This may be because Finland’s laws are more liberal in this respect and give employers more freedom in dismissal selection. The results obtained for Sweden show that the proportion of respondents desiring liberalization falls when attitudes towards job security are measured with a question that emphasizes perceived job security rather than employers’ prerogatives and flexibility.
Based on objective measures of insider–outsider positions, Table 1 indicates that in Norway, the temporarily employed are more favourable towards laxer dismissal rules, but the insider–outsider difference is only moderate. No such difference exists in Sweden. Furthermore, there is no difference between full- and part-time employees in either Norway or Sweden. Analysis using the subjective measures of insider–outsider position reveals only minor differences between those with insecure and secure jobs in Norway, but employees with insecure jobs in Sweden are more likely to prefer stricter regulation. Those with insecure employment prospects in both Sweden and Norway are less likely to desire relaxation of the regulations governing dismissal.
There is almost no difference in Finland between permanent and temporary employees’ opinions when asked whether the employed should have more or less security. In Sweden, a much larger percentage of temporary employees support stricter regulation. This is probably a reflection of the country’s very liberal legislation concerning temporary employees. In both Finland and Sweden, full-time employees are more likely than part-time workers to support a reduction in the security of those in employment although the difference is small. Using the subjective insider–outsider measures for Finland and Sweden, a lower percentage of those with insecure jobs (especially in Sweden) or insecure employment prospects want it to be easier to dismiss employees compared to those with secure jobs and employment. This demonstrates clearer subjective insider–outsider differences in Sweden than were apparent in the responses to the question about employers’ ability to dismiss employees. Among the respondents who experience subjective insecurity, those with perceived employment insecurity outnumber those with job insecurity in all three countries. In principle, flexicurity allows employee security to be achieved by combining moderate or high levels of job insecurity with high levels of employment security: it may be easy to dismiss workers from their current job, but they will readily find new employment and are thus secure overall. However, current conditions do not appear to provide this combination in any of the three countries.
Table 2 shows the results of logistic regression analyses using three different models. Model 1 relies solely on the independent variables that define insider and outsider position by objective criteria. Under this model, there are no significant attitudinal differences in any of the three countries between employees with temporary and permanent contracts or between part-time and full-time workers. Model 2 was introduced to determine whether more subjective definitions of insider–outsider position can better capture differences in attitudes; but the results obtained are almost identical to those for Model 1: we find no significant difference in any of the three countries in attitudes to liberalization between employees with secure and insecure jobs. However, workers who think it unlikely that they could obtain a new job (thus experiencing employment insecurity) appear somewhat less likely to prefer relaxation of laws governing dismissal in Norway and Sweden. We find that perceived employment insecurity makes respondents in Sweden less likely to believe that the law should provide less security, but no such difference exists in Finland. In Model 3, which includes the control variables, employment insecurity ceases to be a significant factor for any of the questions in Sweden. It also suggests that class has profound effects on attitudes, at least in Sweden (both questions) and Finland. Union members as well as manual workers, in both Norway and Sweden, are less likely than non-union members and managers and professionals to think that it should be easier for employers to dismiss employees. Moreover, attitudes to this question in Sweden correlate more closely with class-related variables than in Norway: non-manual workers in Sweden were more likely than manual workers to answer that dismissal should be easier. High-income employees in Sweden were also more likely than those with lower incomes to support easier dismissal; no such difference was observed for Norway. The attitudinal gap between union and non-union members in Finland was identical to that in Sweden. In addition, in both countries, manual workers were less likely to think that the law should provide less security for employees. Model 3 also shows that in Sweden, employees with lower incomes are less likely to agree that the law should provide less security. In Finland, compared to those with the lowest income, the low-income groups are more likely than the middle-income group to favour a reduction in security for those employed. Thus, while the Swedish data indicate a rather linear relation between income and job security preferences, the same is not true for the Norwegian and Finish results.
As noted above, the selection of employees for dismissal is determined by seniority in Norway and Sweden but not in Finland. This may explain why young employees (19–29 years) are more likely to prefer deregulation than those aged 30 to 39 years in Sweden and those aged 50 to 59 years in both Norway and Sweden. This age difference was not observed for Finland but was apparent for both questions in Sweden. The difference in attitudes towards job security regulations between younger and older workers could indicate that attitudes towards job security change over the course of a working life. Alternatively, it could indicate societal changes, whereby younger workers have more liberal beliefs than earlier generations. However, if so one would expect the results for Finland and Sweden to be similar.
Discussion and conclusion
We have studied employees’ attitudes to job security and examined whether the attitudes of insiders and outsiders differ in Finland, Norway and Sweden – three small Nordic countries whose labour markets and societies are structured according to broadly similar principles. As a theoretical starting point, we used Rueda’s thesis that employment protection regulations create and maintain insider–outsider divides. While Lindvall and Rueda (2014) find that such interest differences have in fact influenced Swedish politics and Swedish elections in recent decades, we find no support for this theory in the preferences of employees in any of the countries. The differences in attitudes are small, or even the opposite of those predicted by Rueda, making it hard to believe that social-democratic regimes cater to core insider constituencies that are engaged in an intra-class conflict against the interests of outsiders.
To capture attitudinal differences between holders of secure and insecure jobs in high-mobility, flexicurity-type labour markets such as the Nordic countries, we also applied the concepts of job and employment insecurity. The results still did not support Rueda’s claims. Where statistically significant relationships were observed, they contradicted our expectations: those experiencing subjective job or employment insecurity were less supportive of deregulation than those with greater security.
Transition rates into insider positions are lower in Sweden and Finland than in Norway (Berglund and Furåker, 2011; Svalund, 2013b), which may mean that insecure positions in the labour market are less likely to be considered transitional; one might therefore expect insider–outsider differences to reflect this. However, in comparative terms, all three countries have low rates of long-term unemployment and rather high rates of mobility into the labour market, so segmentation is low and objective outsider status is more transitional than permanent. It is therefore quite possible that the lack of national variation in insider–outsider attitudinal differences is at least partially explained by the similarity of the three labour markets.
It is possible that different results would have been obtained if we had asked other questions about job security. However, the surveys all asked about respondents’ beliefs concerning the effect on unemployment of liberalizing EPL, and we do not find insider–outsider attitudinal divides based on this alternative measure; the results are not shown here but can be provided on request. Our study considered two dependent variables in Sweden, enabling pairwise comparisons with Finland and Norway. The results for the two different questions in Sweden were quite similar, giving a degree of mutual validation. It is perhaps less surprising that no insider–outsider differences were observed in the responses to the question that emphasised employees’ job security. The responses to the other question, which emphasised employer flexibility, might be expected to be more supportive of Rueda’s argument that outsiders should favour employer flexibility; but no difference consistent with this claim was observed, no matter which aspect of the employment relationship was emphasised.
It is possible that the dichotomy between insiders and outsiders is less static than insider–outsider theory assumes. Those who hold insecure jobs or are unemployed today may hold secure jobs in the future, and vice versa, especially in flexicurity-type labour markets (Emmenegger, 2009: 132). This may lead to an overestimation of attitudinal differences between insiders and outsiders. In addition, the consequences of specific job security regulations and reforms may be less straightforward than is assumed by insider–outsider theory. Although some have argued that increasing employers’ ability to hire temporary workers will enable greater integration of outsiders into the labour market, this proposition is not universally accepted in the Nordic countries. It is difficult to predict actors’ preferences based on their interests alone, as Rueda does, because actors may be uncertain about what is rational for them.
While there is little evidence of insider–outsider differences in this study, the control variables introduced to express class seemed much more effective in capturing differences in policy attitudes. Union membership has a clear effect in all three countries, with members being more supportive of employee security. Since membership can be taken to reflect a recognition of class position and class organization, this increased preference for job security is unsurprising; the existing Nordic labour laws and collective agreements were heavily shaped by the exercise of union and employee power. The existence of encompassing and centralized trade unions with broad membership as expressions of class-based alliances may also facilitate the consideration of general goals over particularistic interests, reducing insider–outsider differences (Häusermann and Schwander, 2012; Lindvall and Rueda, 2014: 464). With respect to occupational status (which can be regarded as a reflection of class membership), manual workers were more likely to support employee security than managers and professionals. This was true in all three countries but particularly in Sweden, showing that differences in attitudes to job security are structured along class divides. The same pattern was observed in Sweden when using income as an indicator of socioeconomic status. A similar pattern was seen in Finland although here the lowest income group deviated from the trend. Based on the differences between these expressions of class and the (apparently non-existent) gap between the attitudes of insiders and outsiders, it seems that horizontal rather than vertical power relations shape differences in policy attitudes in these Nordic countries. Notably, Sweden has the most dualized labour market with respect to EPL and the lowest labour market mobility of the three countries and also shows the clearest class divides concerning policy attitudes.
Hence the observation that class indicators correlate with policy attitudes but insider–outsider status does not may result from the strength of egalitarianism in these countries. Egalitarian values may encourage moral convictions and obligations that form a basis for solidarity between (future) recipients of job protection and those less dependent on it, making insiders more concerned about the employment opportunities of those with precarious working situations than insider–outsider theory would predict. Furthermore, norms and preferences are not only influenced by one’s labour market position but also by socialization, upbringing and social interaction in general. Potential interest divides between insiders and outsiders (and the influence of such divides on policy attitudes) may thus be reduced by egalitarian labour markets, the specifics of national politics and ongoing political debates, and the level of trust between people at all levels of society.
Finally, seniority rules are necessarily age discriminatory because seniority tends to increase with age (Elster, 1992). Age-related differences in policy attitudes were observed for Norway and Sweden, where seniority is an important selection criterion during collective dismissals. Thus, interest divides do seem to influence policy attitudes in these countries to some extent.
Denmark is unfortunately not covered by our study. The regulation of permanent contracts is, according to Figure 1, as lax in Denmark as in Norway, while the regulation of temporary contracts is somewhat stricter than in Finland and much stricter than in Sweden. But the regulation of both permanent and temporary contracts is mostly anchored in collective agreements, with large variations in job security, such as periods of notice, between skilled and unskilled blue-collar workers on the one side (often very short notice periods) and salaried employees with middle- or high-level education and qualifications on the other (Jensen, 2011: 727–728). The job security of civil servants (about 10% of all employees) is regulated by legislation; many other salaried employees, in total, 53 per cent of all employees, have contracts with similar terms, granting them a notice period similar to that in the other Nordic countries (Larsen and Madsen, 2015: 105). Compared to Norway and especially Sweden, Danish employers have more prerogatives concerning selection of dismissals in case of mass redundancies (Svalund et al., 2013: 199). The collective agreement of 2008–2010 for Danish manufacturing increased severance pay provision for workers with more than 3 years tenure, supplementing unemployment benefits up to 85 percent of previous pay levels (Ibsen et al., 2011: 333). Thus, some insiders have managed to increase their job security; while as in the other Nordic countries, there have been no public demands by outsiders for laxer job protection.
Mobility rates between employment and unemployment and from temporary to permanent employment have been high in Denmark (Berglund and Furåker, 2011; Svalund, 2013b). The lax regulation of permanent contracts and high mobility rates may encourage fewer insider–outsider differences in attitudes towards job security, compared to the other Nordic countries, while the limited legislation and the large variation between different collective agreements may lead to attitudinal variation between employee groups. There are no studies of Danish employee preferences regarding job security rules, but in a 2013 survey, employees were asked, in a time of crisis, whether they would prefer job or employment security. In all, 59 percent preferred job security, while 39 percent prioritized good opportunities to be able to find other work (employment security) (Bredgaard et al., 2015: 167). Thus, job security is important also for Danish employees. The study showed no difference between permanent and temporary employees on this matter, and there were no differences between those who had been unemployed within the last 5 years, and those who had not, also when the authors controlled for notice periods (Bredgaard et al., 2015: 168–169). Thus, while the question is related to the balance between job and employment security rather than the strictness of job security regulations, the results are in line with our findings. Furthermore, employees with limited education prioritized job security over employment security to a higher degree than those with longer education (Bredgaard et al., 2015: 168–169).
There are limitations that must be borne in mind when evaluating our conclusions. First, no unemployed persons, who are usually regarded as a sub-group of outsiders, were included. Second, using cross-sectional data made it impossible to definitively elucidate the causal relationships between the variables; panel data research would be needed to achieve this. Additionally, when comparing countries, it was not possible to pool the survey responses, so it was impossible to control for differences in labour market composition. Further research using a wider selection of countries including some with clearly dualized labour markets could clarify the extent to which the Nordic countries, with their more class-based and egalitarian structures, behave differently with regard to policy attitudes.
Footnotes
Acknowledgements
The authors are grateful to Tomas Berglund, Jon Erik Dølvik, Bengt Furåker, the Journal’s Editor and the referees for helpful, challenging and constructive comments.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was funded by the Research Council of Norway’s programme on welfare, working life and migration (Velferd, arbeidsliv og migrasjon, VAM), the Finnish Work Environment Fund (Työsuojelurahasto rahoittaa) and the Swedish government agency VINNOVA.