Understanding Perceived Worker Insecurity in Europe,2002–2016: Economic Freedom and Neoliberalism as Alternative Theories?

Abstract

This study engages theories of economic freedom and neoliberalism as alternative ways of understanding workers’ perceived job and labor market insecurity. The results of hierarchical ordered logit and linear models of multiple rounds/waves of European Social Survey and European Working Conditions Survey data between 2002 and 2016 reveal some support for one hypothesis derived from each set of theories in basic models: As theories of economic freedom anticipate, the levels of worker insecurity are generally lower in countries that are currently more “economically free”; as theories of neoliberalism anticipate, the levels of insecurity are higher in countries that have increased levels of economic freedom, or “neoliberalization,” in about the past five years, but this relationship is limited to pre-2010 survey data. Altogether, the findings, including of more complex models, suggest these theories account for worker insecurity in Europe, but neoliberalization’s effects are selective, unemployment also matters, and synthesis is needed.

Keywords

labor and labor movements global and transnational sociology comparative and historical sociology job insecurity neoliberalism liberalization European union

Introduction

Worker insecurity is a central focus of research in light of claims about an “age of insecurity” (Burchell 2009; Fevre 2007), the 2007/2008 economic crisis (Chung and van Oorschot 2011), and what some see as the global turn toward neoliberalism (Standing 2016). Such insecurity may be objective, as in the case of job loss, or perceived (see Lee, Huang, and Ashford 2018). Workers’ perceptions of insecurity are not only important in the context of theoretical debates, but they are also related to workers’ job performance (see Cheng and Chan 2008), health (László et al. 2010), and actual job loss (Dickerson and Green 2012). Workers’ perceptions of insecurity vary across countries (see Keim et al. 2014). Some countries, particularly in Europe, attempt to address worker insecurity more generally through policy reforms (Wilthagen and Tros 2004).

Cross-national research is mired in debate over the (relative) roles that institutions—defined as “a set of rules, formal or informal, that actors generally follow” (P. A. Hall and Soskice 2001:9; see North 1990:3)—and economic factors play in accounting for workers’ perceived insecurity¹ (e.g., Anderson and Pontusson 2007; Green 2009). Yet, this debate is theoretically narrow and, despite the contributions of prior research, methodologically difficult to settle. As Kathleen Thelen (2012:138–39) argues, broader “liberalization” processes challenge institutional theories’ assumed distinctions between types of economies because developed economies are increasingly similar. Empirically, the effects of countries’ institutions, such as their levels of employment protection legislation (EPL), on levels of worker insecurity are mixed, and at times, the opposite of predictions (Clark and Postel-Vinay 2009). Research on Europe, specifically, is at an impasse because there are too many theoretical factors and often not enough countries with which to properly test them (see Bryan and Jenkins 2016; Stegmueller 2013). Some seemingly conflicting conclusions are also based on different forms of worker insecurity, the measures of which may vary by survey and year (Chung and Mau 2014). In short, the literature is in need of alternative theories, more parsimonious models, and analyses of different forms of insecurity.

The present study contributes to this literature and partly addresses its gaps by engaging theories of economic freedom and neoliberalism that encompass and extend explanations of worker insecurity. The concepts of economic freedom and neoliberalism are similar (Foucault 2008; Harvey 2005), referring to an institutional context promoting privatization, free markets, and at least minimal government intervention. Theories of economic freedom and neoliberalism, however, differ in their predictions: Based on the former set of theories (Gwartney and Lawson 2003; Gwartney, Lawson, and Holcombe 1999), economic freedom and its increase over time should mitigate worker insecurity. In contrast, the latter set of theories propose that neoliberalism and neoliberalization—the process of adopting neoliberal policies over time (Harvey 2005, 2006)—foment workers’ insecurity (Bourdieu 1998; Standing 2016) but are rarely directly examined in quantitative research on worker insecurity (Kalleberg 2018; Kalleberg and Vallas 2017). Using an economic freedom index (Gwartney et al. 2018) and its over-time change to proxy the main concepts of both sets of theories (Centeno and Cohen 2012; Cohen 2011; De Haan, Lundström, and Sturm 2006), I test the hypotheses implied above via hierarchical ordered logit and linear models on two forms of worker insecurity—job and labor market insecurity—which, respectively, refer to workers’ perceptions that they will lose their jobs and be unable to find comparable ones (Anderson and Pontusson 2007). The analysis makes use of multiple rounds/waves of European Social Survey (ESS) data between 2002 and 2016 and European Working Conditions Survey (EWCS) data in 2005, 2010, and 2015 linked to macro data, but I analyze each round/wave, dataset, and dependent measure separately: This is not a time series analysis, as there are few time points; the results also indicate somewhat different findings across years of the survey data. This article begins by reviewing cross-national research on worker insecurity, focusing on Europe, before turning to theories of economic freedom and neoliberalism. After discussing the methods and data I use, I present the results: The basic models lend some credence to both sets of theories, but overall, the relationships anticipated by theories of neoliberalism are more selective than their counterparts and theoretical synthesis is needed, as noted in the conclusion.

Background

The starting point of cross-national research on worker insecurity is that such insecurity takes different forms (Anderson and Pontusson 2007; Chung and van Oorschot 2011; Green 2009). This study examines what Christopher J. Anderson and Jonas Pontusson (2007:214–15) call cognitive job insecurity (hereafter, “job insecurity”)—defined as “an individual’s estimate of the probability that he or she will lose their job in the near future”—and labor market insecurity, which are perceptions of “a low probability of finding another job with more or less equivalent characteristics.”

Anderson and Pontusson’s (2007) model holds that job and labor market insecurity are rooted in different institutions, such as countries’ EPL and active labor market policies (ALMPs), which, respectively, protect workers from dismissals and help the unemployed find jobs. Whereas EPL levels mitigate the levels of job insecurity, ALMP levels mitigate the levels of labor market insecurity, based on Anderson and Pontusson’s (2007) analysis of 15 Organization for Economic Co-operation and Development (OECD) countries using 1997 International Social Survey Programme (ISSP) data. Arguing that economic conditions also provide “cues” for workers, Anderson and Pontusson (2007:223) find one-year changes in unemployment to be associated with greater levels of job and labor market insecurity as well.

Subsequent research lends inconsistent support to the above institutional hypotheses, but more support for hypotheses concerning the role of economic conditions. Research using ESS data, which is one data source used in the current study, shows that countries’ EPL levels are not generally related to the levels of job insecurity (Erlinghagen 2008; Esser and Olsen 2012; Lübke and Erlinghagen 2014). Still, EPL, or some of its components, accounts for the gap in the levels of job and other forms of insecurity between temporary and permanent workers (Balz 2017; Chung 2016). ALMP spending and labor market insecurity levels are not related in the 2004 and 2010 ESS (Lübke and Erlinghagen 2014), but unemployment assistance accounts for variation in the levels of labor market insecurity across 23 OECD countries in the 2005 ISSP (Hipp 2016). Job and labor market insecurity levels are often higher when unemployment is greater, regardless of the data, years, or countries examined (Dixon, Fullerton, and Robertson 2013; Erlinghagen 2008; Esser and Olsen 2012; Green 2009; Hipp 2016; Lübke and Erlinghagen 2014).

Heejung Chung and Wim van Oorschot (2011:289) offer a different perspective, proposing that “employment insecurity”—that is, “having (the potential for) secure and continuous employment, which might entail changing employers and/or jobs”—is better explained by countries’ economic conditions than institutions. They suggest that models should account for recessionary conditions especially in the wake of the 2007/2008 economic crisis. Coupled with other findings on economic conditions above, research lends credence to this view: Examining 2008 ESS data, Chung and van Oorschot (2011: 299) find that one-year changes in countries’ unemployment rates are positively—and one-year gross domestic product (GDP) growth rates are negatively—related to the levels of employment insecurity, acknowledging their results may be specific to this year (see also Chung 2016; Chung and Mau 2014). Using 2004 and 2010 ESS data, Christiane Lübke and Marce Erlinghagen (2014) show that a one-year change in GDP per capita is negatively related to the levels of job insecurity, but not labor market insecurity. Yet, Marcel Erlinghagen (2008) does not find a relationship between average GDP growth and the levels of job insecurity in the 2004 ESS.

To account for the role of countries’ economies, more broadly, and specific institutions, research also draws on the Varieties of Capitalism (VoC) framework (P. A. Hall and Soskice 2001) and welfare state models (Esping-Andersen 1990). These theories, respectively, suggest that worker insecurity should be greatest in “liberal” market economies, characterized by competitive and hierarchical relationships (P. A. Hall and Soskice 2001:8), and “liberal” welfare states, typified by high commodification and low levels of public support (Esping-Andersen 1990). In contrast to these expectations, studies find that job insecurity levels are higher in post-Socialist countries (Dixon et al. 2013; Gash and Inanc 2013). Along with other theories (e.g., Korpi 2006), the VoC and welfare state models also highlight institutional mechanisms that bear on worker insecurity, such as union strength (Esser and Olsen 2012): Whereas job insecurity (Esser and Olsen 2012) and employment insecurity (Chung 2016) levels are lower in European countries with higher union density, labor market insecurity levels are not (Dixon et al. 2013). Although these and similar theories (e.g., Gallie 2013) account for many work outcomes (Esser and Olsen 2012; Gash and Inanc 2013; Mai 2017), they have difficulty in explaining worker insecurity, per se.

Attempts to move past the VoC framework and welfare state models have had mixed success. Arguing that corruption undermines the rule of law and the possibility of a “class compromise” (Wright 2000) assumed by some theories, Jeffrey C. Dixon et al. (2013) find that perceived corruption accounts for the cross-national variation in the levels of some forms of insecurity (e.g., job insecurity), but not others (e.g., labor market insecurity). Labor market flexibility, as manifested by the proportion of nonstandard workers, is also inconsistently related to worker insecurity: Whereas countries with a greater proportion of part-time workers have lower levels of job insecurity (Dixon et al. 2013), the proportion of temporary workers does not appear to be related to the levels of job or employment insecurity (Chung 2016; Hipp 2016).

In sum, there are inconclusive results in this literature (Hipp 2016): With the exception of the effects of unemployment, findings often vary by the form of insecurity. Empirical analyses also tend to include several country-level variables and/or specify cross-level interactions, both of which can result in potentially unwieldy models if relatively few countries are examined (Bryans and Jenkins 2016; Stegmueller 2013), as in Europe. This problem is exacerbated by the use of (inconsistently significant) OECD institutional measures, which are usually only available for OECD countries and thus limit the country samples accordingly (see Mai 2017 on the latter point). As such, the literature is in need of alternative ways of understanding worker insecurity.

Toward Alternative Ways of Understanding Worker Insecurity

Implicit in some theory that guides research is the notion that “liberalism” bears on workers’ insecurity. Peter A. Hall and David Soskice (2001:50–51) suggest that the VoC framework and welfare state models above assume a common definition of liberalism: “Virtually all liberal market economies are accompanied by ‘liberal’ welfare states.” Yet, conceptual and other challenges have until now limited the application of liberalism and similar concepts to research on worker insecurity.

One such concept is “economic freedom,” which emerges from some literature in economics. James D. Gwartney and Robert A. Lawson (2003:408) partly conceive of economic freedom as an “institutional and policy environment.” James D. Gwartney et al. (2018:1) argue that the “cornerstones of economic freedom are personal choice, voluntary exchange, open markets, and clearly defined and enforced property rights.” Terry Miller, Anthony B. Kim, and James M. Roberts (2018) assert that economic freedom involves freedom of (economic) choice and minimal government intervention into the economy.

“Neoliberalism” is another such concept, which appears in studies in sociology and allied disciplines. John L. Campbell and Ove K. Pedersen (2001:5) conceptualize neoliberalism partly as institutions, such as “minimalist welfare-state, taxation, and business regulation programs; flexible labor markets and decentralized capital-labor relations.” For David Harvey (2005:2), neoliberalism is “a theory of political economic practices that proposes that human well-being can best be advanced by liberating individual entrepreneurial freedoms and skills within an institutional framework characterized by strong private property rights, free markets, and free trade.” In research focused on work, the global neoliberal revolution—or neoliberal globalization—emphasizes “the centrality of markets and market-driven solutions, privatization of government resources, and removal of government protections” (Kalleberg 2009:3; see also Bandelj, Shorette, and Sowers 2011). Although there continues to be debate over the meaning of neoliberalism (see Birch 2015), extant theory suggests some common characteristics of it are privatization, free markets, and minimal or selective (Jessop 2002) government intervention in the economy.

So defined, the concepts of economic freedom and neoliberalism are similar (Foucault 2008). Importantly, both concepts not only encompass mechanisms in studies on worker insecurity, such as deregulation, government spending, and flexibility, but also extend beyond the specific labor market and other institutions examined in extant research. Only the former concept has a standard measure—economic freedom indexes (Gwartney et al. 2018; Miller et al. 2018)—but scholars suggest these indexes can be used as proxies for neoliberalism (e.g., Centeno and Cohen 2012; Cohen 2011; Peck 2010). For instance, Arne L. Kalleberg (2018:38, Footnote 4) notes, “Neoliberalism can be measured by indicators such as the degree of ‘economic freedom’ or market openness . . .” As such, the present study makes use of an economic freedom index as a proxy for the main concepts of both sets of theories, the details of which are elaborated later.

Despite conceptual similarities, theories of economic freedom and neoliberalism offer different accounts of socioeconomic phenomena. Next, I discuss these sets of theories, deriving hypotheses from each about the cross-national variation in job and labor market insecurity.

Theories of Economic Freedom

Theories of economic freedom argue that economic freedom promotes countries’ growth. This occurs through the expansion of personal choice, according to Gwartney and Lawson (2003:406):

When economic freedom is present, the choices of individuals will decide what and how goods and services are produced . . . Personal ownership of self is an underlying postulate of economic freedom. Because of this self ownership, individuals have a right to choose—to decide how they will use their time and talents.

By offering people choice and promoting growth, these theories argue, economic freedom should reduce poverty and increase national welfare (Gwartney et al. 2018; Miller et al. 2018).

Particularly relevant to the present study is that a growing strand of this literature shows that economic freedom is associated with people’s perceptions. Research finds that economic freedom—as an institutional/policy context and measured with an economic freedom index—is positively related to people’s life control perceptions (e.g., Pitlik and Rode 2016), which are assumed to be intermediating mechanisms in these theories’ causal models. Countries’ levels of economic freedom are also positively related to the levels of perceived happiness and/or life satisfaction (Veenhoven 1999), which hold after controls for growth (Ovaska and Takashima 2006) and unemployment, but some relationships vary by levels of economic development (Gehring 2013). In a review of research using the Economic Freedom of the World (EFW) index employed in the current study (Gwartney et al. 2018), Joshua C. Hall and Robert A. Lawson (2014) claim that less than five percent of studies find a relationship between economic freedom and a “bad” outcome.

The notion that economic freedom promotes national welfare suggests it will also mitigate job and labor market insecurity, which is partly buttressed by the above research on people’s perceptions. From this perspective, economically freer institutional environments provide individuals with choice and opportunity—for example, to develop human capital, change jobs, or open a business (Gwartney and Lawson 2003). Where such possibilities are in shorter supply, workers are more likely to perceive their jobs as insecure and have a bleaker outlook of other comparable jobs. Hence, this set of theories suggests the following hypothesis:

Hypothesis 1 (H1): Countries with higher levels of economic freedom are expected to have lower levels of job and labor market insecurity, all else equal.

This literature also recognizes the possibility that the effects of economic freedom are dynamic. According to James D. Gwartney et al. (1999), the extent to which countries are economically free is less telling than changes in countries’ economic freedom. They continue,

Because credibility must be earned, there will often be a time lag between a change in economic freedom and when the change exerts an impact on economic activity. For example, when a nation moves toward a more stable monetary policy or more liberal trade regime, it will take time to convince economic decision-makers that the change is permanent, rather than temporary. (Gwartney et al. 1999:647)

This theoretical emphasis on time is supported by research indicating that the levels of economic freedom have less robust effects on growth than changes in economic freedom over time (De Haan et al. 2006; Doucouilagos and Ulubasoglu 2006), the latter of which Gwartney et al. (1999) measure in (multiple) five-year intervals. Following from H1 and the research above, worker insecurity may thus be (further) mitigated as institutions/policies become “freer” over time. Formally, this suggests the following hypothesis in which over-time change is operationalized in an approximate five-year interval also due to issues of data comparability/availability noted later:

Hypothesis 2 (H2): Countries with greater increases in levels of economic freedom in about the last five years are expected to have lower levels of job and labor market insecurity, all else equal.

Theories of Neoliberalism

In contrast to theories of economic freedom, some “theories of neoliberalism” (Barnett 2010)—particularly class-based theories (Birch 2015)—argue that neoliberalism fails to deliver on its promise of economic growth (Harvey 2005, 2006). Instead, it is oft-theorized to yield inequality, unemployment, and other negative consequences (e.g., Calhoun 2011; Duménil and Lévy 2011).

It has long been assumed that neoliberalism foments risk, uncertainty, and insecurity (e.g., Beck 2000; see Kalleberg 2009), but some theory more explicitly links neoliberalism to the kinds of worker insecurity examined in this study. According to Pierre Bourdieu (1998:98), for example, “The ultimate basis of this economic order placed under the banner of individual freedom is indeed the structural violence of unemployment, of insecure employment and the fear provoked by the threat of losing employment.” As Bourdieu (1998:83) notes elsewhere, “Objective insecurity gives rise to a generalized subjective insecurity which is now affecting all workers in our highly developed economy.” Guy Standing (2016:Chapter 1) argues neoliberalism promotes job and labor market insecurity—albeit defined differently than in the current study—and other forms of insecurity, primarily among a growing population of the “precariat.”

How, exactly, does neoliberalism foment job and labor market insecurity? From one perspective, neoliberalism represents risks to workers through the “elimination” (Centeno and Cohen 2012) and “creative destruction” (Harvey 2006) of the very mechanisms theorized to provide them with security: namely, certain forms of government support and regulations (see also Kalleberg 2018; Kalleberg and Vallas 2017). In addition, Standing (2016) highlights labor market flexibility and de-unionization as major reasons for heightened worker insecurity. In their review of research on work, Nina Bandelj et al. (2011:816) conclude that “among the many liabilities of neoliberal globalization are increased job security . . . in the developed world.” Much in the way that economic conditions may provide “cues” to workers (Anderson and Pontusson 2007), neoliberalism may likewise offer such cues to workers about their job and labor market security.

Despite strong theoretical contentions about the effects of neoliberalism, they are often “presumed” (Crowley and Hodson 2014:92). There are few quantitative studies that directly measure neoliberalism and examine its relationship to work-related outcomes (Kalleberg 2018; Kalleberg and Vallas 2017; see Williams 2017 for a notable exception). In a novel content analysis of work-related ethnographies, however, Marth Crowley and Randy Hodson (2014) show that what they call neoliberal workplace practices—namely, firings, task organization, absence of unions, and the use of flexible labor, among other factors—are related to greater job insecurity. Other qualitative research paints a picture of workers who are often aware that they do not have a safety net, may not do as well economically as previous generations, and may change employers often (Pugh 2015; Silva 2013), even if they do not call these forces “neoliberalism.” Theories of neoliberalism and a fairly small body of empirical research thus suggest the following counterpart to H1 derived from theories of economic freedom above:

Hypothesis 1a (H1a): Countries with higher levels of economic freedom are expected to have higher levels of job and labor market insecurity, all else equal.

The effects of neoliberalism may also depend on time and place. Neoliberalization, which is the process of adopting neoliberal policies (Harvey 2005, 2006), is likely to foment insecurity if countries undergo relatively rapid changes in a short time. Neoliberalization can be dated at least to the 1970s and 1980s, but some countries such as Chile and Britain adopted neoliberal policies more rapidly than others (Fourcade-Gourinchas and Babb 2002)—and with salient consequences: In the 1990s, for instance, some post-Socialist countries that quickly adopted neoliberal policies (referred to as “shock therapy”) experienced greater increases in poverty than their counterparts (Harvey 2005, 2006). In fact, neoliberalization may explain why worker insecurity remains higher in post-Socialist countries today (Dixon et al. 2013; Gash and Inanc 2013). Theories of neoliberalism suggest that over-time changes in economic freedom represent (greater) neoliberalization and are likely to have an opposite effect than in H2:

Hypothesis 2a (H2a): Countries with greater increases in levels of economic freedom in about the last five years are expected to have higher levels of job and labor market insecurity, all else equal.

Data, Sample, and Measures

To test these hypotheses, I use ESS and EWCS data and macro data noted below. As Brendan Burchell (2009) suggests, the use of different survey datasets serves as a reliability check on the findings. Multiple rounds/waves of each survey dataset are also used, so that altogether I am able to analyze eight years’ worth of data—2002, 2004, 2005, 2006, 2008, 2010 (×2), 2015, and 2016—and during an important period of time: In 2004, 2007, and 2013, the European Union (EU) enlarged to include a number of formerly Socialist countries (e.g., Poland, Bulgaria, Croatia) as well as others (e.g., Cyprus); moreover, the economic crisis occurred in 2007/2008. The number of years examined represents an improvement over previous research (Chung and Mau 2014), but the purpose of this study is not to examine changes in worker insecurity over time for reasons noted above and below. I attempt to construct the individual-level samples in the same way across survey datasets. The individual-level data are weighted using post-stratification weights in the main analyses.

The country-level samples, which overall range from N = 20 (2002 ESS) to N = 35 (2015 EWCS), improve upon the small-N problem in previous research on Europe, but they do not fully solve it. All of the analyses are based on a sample of greater than 15 Level 2 units—a point under which Level 2 coefficient estimates may be quite biased—albeit bias is also a function of the number of Level 2 variables and complexity of the model (Stegmueller 2013). The country-level sample sizes in the EWCS meet Mark L. Bryan and Stephen P. Jenkin’s (2016) guidelines for multilevel linear and logit models, which are, respectively, 25 and 30, depending on their complexity; however, the multilevel ordered logit models I use here are more complex. Much like extant research using the ESS, the country-level sample sizes in the ESS often do not meet these guidelines, except for some (linear) analyses in the 2010 ESS (N = 25). In light of this, I attempt to keep the models parsimonious, and in my interpretations, I give greater credence to the results of basic models and those with larger country-level sample sizes (i.e., the EWCS).

Covering EU and affiliated countries, ESS survey data have been collected bi-annually since 2002. They are based on probability samples of people in each country (ESS 2018). The first five rounds of the ESS are used (–2010)—in addition to the eighth round (2016)—as each has at least one measure of job insecurity or labor market insecurity and sometimes both.² The analyses are limited to respondents aged 15 to 64 in the paid labor force who report working hours in the previous week. Self-employed, missing data, others, and those in the military are excluded. I also exclude France in 2002 and 2004, and Hungary in 2004 (only), due to missing data on self-employment. In line with some other research (Chung and van Oorschot 2011), I exclude Israel and Russia as well. After these exclusions, the ESS country samples range from N = 20 (2002) to N = 27 (2008). The individual-level samples range from n = 14,491 (2002) to n = 18,712 (2008).³

The EWCS surveys workers aged 15 or 16 and above in EU member and other countries, using a multistage, stratified sampling design with interviews conducted face-to-face. Workers are those working at least one hour in the week before the interview. The sample is limited to employees aged 15 to 64. I use the 2005, 2010, and 2015 waves: The 2005 wave includes the first 27 EU member states, plus Turkey, Croatia (now an EU member), Norway, and Switzerland (N = 31). The 2010 wave includes the countries in the 2005 wave, except Switzerland; it adds Albania, Macedonia, Montenegro, and Kosovo, but Kosovo is excluded due to missing Level 2 data and Germany is excluded due to missing Level 1 data (on the education variable I use) (N = 32). The 2015 wave includes the countries in the 2010 wave, except Kosovo; it adds Germany, Switzerland, and Serbia (N = 35) (European Foundation for the Improvement of Living and Working Conditions 2018). After deleting missing values on the individual-level variables, the individual-level samples per wave are n = 21,228 (2005), 27,383 (2010), and 29,389 (2015).

Dependent Variables

This study focuses on job insecurity and labor market insecurity and analyzes them separately: When measures of both job and labor market insecurity are available in the same dataset, they are weakly related to one another (gamma = 0.11 [2004 ESS], 0.15 [2010 ESS], 0.11 [2010 EWCS], and 0.03 [2015 EWCS]). These measures largely, but not entirely, follow Anderson and Pontusson’s (2007) definitions of these concepts and previous research’s operationalization of them; yet, I do not collapse categories of the variables in the main analyses because of the potential loss of information. The measures differ between the ESS and EWCS as well as across rounds/waves of these surveys.

Table 1 presents the question wording of the dependent variables and descriptive statistics for these variables, organized by dependent variable and year. There are limitations to some of these measures. For example, I use the 2008 and 2016 ESS measure to operationalize job insecurity because, on its face, it more closely resembles this than labor market insecurity. Yet, this measure does not exactly correspond to Anderson and Pontusson’s definition; in fact, other research uses it to operationalize “employment insecurity” (Chung and van Oorschot 2011). Regardless, this is a problematic measure, as the question is double-barreled. In addition, although some questions’ wording appears similar over time, there are differences. As Lübke and Erlinghagen (2014:323) indicate, the measure of labor market insecurity in the 2004 ESS (and 2002 ESS) ends with “if you wanted to,” whereas in the 2010 ESS, it ends with “if you had to leave your current job.” Question wording differences preclude a longitudinal analysis of the ESS, as there would be no more than two time points with identical dependent measures. In the EWCS, the exact question is used to measure job insecurity in 2005, 2010, and 2015, but as we will see from the results, there are somewhat different patterns across waves. All dependent variables are coded so that higher values indicate greater insecurity; missing values are deleted.

Table 1.

Dependent Variable Question Wording and Descriptive Statistics: ESS and EWCS.

Dependent Variable and Question Wording	Proportions or x̄	ICC/Pseudo
Perceived job insecurity
“ . . . My job is secure.” (2004 ESS)	—	0.09
“Very true”	0.31
“Quite true”	0.37
“A little true”	0.19
“Not at all true”	0.13
“I might lose my job in the next 6 months” (2005 EWCS)	—	0.09
“Strongly disagree”	0.42
“Tend to disagree”	0.28
“Neither agree nor disagree”	0.12
“Tend to agree”	0.11
“Strongly agree”	0.06
“How likely would you say it is that you will become unemployed in the next 12 months?” (2006 ESS)	—	0.08
“Not at all likely”	0.47
“Not very likely”	0.39
“Likely”	0.10
“Very likely”	0.04
“Using this card, please tell me how likely it is that during the next 12 months you will be unemployed and looking for work for at least four consecutive weeks?” (2008 ESS)	—	0.14
“Not at all likely”	0.34
“Not very likely”	0.39
“Likely”	0.20
“Very likely”	0.07
“ . . . My job is secure.” (2010 ESS)	—	0.16
“Very true”	0.30
“Quite true”	0.32
“A little true”	0.23
“Not at all true”	0.16
“I might lose my job in the next 6 months” (2010 EWCS)	—	0.10
“Strongly disagree”	0.32
“Tend to disagree”	0.30
“Neither agree nor disagree”	0.17
“Tend to agree”	0.14
“Strongly agree”	0.07
“I might lose my job in the next 6 months” (2015 EWCS)	—	0.09
“Strongly disagree”	0.47
“Tend to disagree”	0.20
“Neither agree nor disagree”	0.15
“Tend to agree”	0.10
“Strongly agree”	0.07
“Using this card, please tell me how likely it is that during the next 12 months you will be unemployed and looking for work for at least four consecutive weeks?” (2016 ESS)	—	0.10
“Not at all likely”	0.40
“Not very likely”	0.42
“Likely”	0.13
“Very likely”	0.04
Perceived Labor Market Insecurity
“ . . . how difficult or easy would it be for you . . . to get a similar or better job with another employer if you wanted to?” Extremely easy = 00, extremely difficult = 10 (2002 ESS)	5.70	0.11
“ . . . how difficult or easy would it be for you to get a similar or better job with another employer if you wanted to?” Extremely easy = 00, extremely difficult = 10 (2004 ESS)	5.63	0.10
“ . . . how difficult or easy would it be for you to get a similar or better job with another employer if you had to leave your current job?” Extremely easy = 00, extremely difficult = 10 (2010 ESS)	5.54	0.07
“If I were to lose or quit my current job, it would be easy for me to find a job of similar salary” (2010 EWCS)	—	0.09
“Strongly agree”	0.08
“Tend to agree”	0.24
“Neither agree nor disagree”	0.19
“Tend to disagree”	0.27
“Strongly disagree”	0.21
“If I were to lose or quit my current job, it would be easy for me to find a job of similar salary” (2015 EWCS)	—	0.05
“Strongly agree”	0.14
“Tend to agree”	0.23
“Neither agree nor disagree”	0.19
“Tend to disagree”	0.18
“Strongly disagree”	0.26

Note. Data are weighted. Proportions may not add up to 1 due to rounding. ICC/Pseudo refers to ICC or its Pseudo variant; see text for calculations. ESS = European Social Survey; EWCS = European Working Conditions Survey; ICC = intraclass correlation.

Based on null models of job and labor market insecurity with a randomly varying intercept but no other variables, the levels of both forms of insecurity vary significantly across countries. In the far right-hand column of Table 1 appears the intraclass correlation (ICC) or “pseudo” ICC, which indicates the extent of cross-national variation in the dependent variables.⁴ Overall, the cross-national variation in the dependent variables ranges from five percent (labor market insecurity in the 2015 EWCS) to 16 percent (job insecurity in the 2010 ESS).

Country-level Independent Variables

To account for the cross-national variation in worker insecurity, I include country-level independent variables suggested by theory and prior research. I also discuss potential objections to the variables used and modeling strategy in the sections below.

Economic Freedom and Its Change

To approximate the main concepts of theories of economic freedom, I use the EFW index published by the Fraser Institute (Gwartney et al. 2018). Each country is rated on its economic freedom, the overall index score (0–10) of which includes five major areas: government size, legal system and property rights, sound money, international free trade, and regulation. According to Gwartney et al. (2018: Appendix), government size includes measures of government consumption, transfers and subsidies, government enterprises and investment, and the top marginal tax rate. Legal system and property rights includes nine subareas (i.e., judicial independence, protection of property rights, and legal enforcement of contracts). The sound money area includes measures of money growth, inflation and its standard deviation, and freedom to own foreign bank accounts. Freedom to trade internationally includes subareas and measures of tariffs, regulatory trade barriers, black market exchange rates, and controls of the movement of people and capital. Regulation, as the final area, encompasses labor market, business, and credit market regulations. The underlying data for the index come from multiple sources, such as the International Monetary Fund (IMF), World Bank, World Economic Forum, and United Nations (Gwartney et al. 2018: Appendix). The EFW index is employed in some literature in economics as a measure of “economic freedom” (see J. C. Hall and Lawson 2014 for a review)—and coded as such (i.e., with higher values indicating more “freedom”).

As discussed earlier, I use the EFW index to approximate the main concepts of theories of neoliberalism as well. On its face, this index maps on to characteristics of neoliberalism as espoused in the “Washington Consensus” (De Haan et al. 2006), which is a set of economic principles promulgated by the United States and adopted by international organizations (Williamson 1993). Although not uncritically, research uses the EFW index as a proxy for neoliberalism (Peck 2010:10–12) or “liberalization” (Kwon 2016). As a proxy for neoliberalism in the present study, the EFW index also theoretically encompasses several salient foci of prior research on worker insecurity: government size (i.e., spending), legal system (i.e., corruption), and (de-)regulation (i.e., employment protection). Hence, my use of this index offers a means of developing a more parsimonious model. Furthermore, compared with OECD institutional measures, the index is available for a larger and more diverse set of countries, including those outside of the OECD.

Yet, the index is not without potential disadvantages. First, Jakob De Haan et al. (2006) claim researchers may be hesitant to use the index because of the (libertarian) ideological orientation of the Fraser Institute. They argue that this does not undermine the index’s validity and suggest that “readers who do not like the usage of the EF [economic freedom] terminology” may use other terminology to describe the index (De Haan et al. 2006:160). I follow this suggestion by using terminology rooted in theories of economic freedom and neoliberalism alike. Moreover, De Haan et al. (2006) suggest that the index’s internal reliability could be improved. Analyses of the five areas of index corresponding to the year of the survey data and about five years prior (based on the operationalization below) reveal that Cronbach’s alphas are acceptable across datasets/samples (=.60–.79), except for the 2016 ESS sample (.28–.35). This warrants caution in drawing conclusions from the 2016 ESS in particular. Still, these are empirically driven alphas with items reversed as needed (see Ott 2018), and in fact, factor analyses of the five areas of the index comparable with the above analyses indicate that the “government size” area often loads best on a separate factor; it is often negatively related to other areas (see also Cohen 2011). Supplementary analyses discussed later indicate that the EFW index is “better without size of government” (Ott 2018), as its expected effects are more consistent.

The advantages of using this index outweigh the disadvantages. Specifically, I use the panel data recommended for analyses with a temporal component, as I am examining multiple years of data and changes in economic freedom over time. To test H1 and H1a, which are, respectively, derived from theories of economic freedom and neoliberalism and posit opposite predictions about the levels of worker insecurity, I use the current level of economic freedom (i.e., corresponding to each survey year).⁵ Using the current level is consistent with previous research on perceptions in literature on economic freedom (Gehring 2013; Ovaska and Takashima 2006; Veenhoven 1999). To test H2 and H2a, which posit competing effects of economic freedom changes/neoliberalization on the levels of worker insecurity, I calculate an over-time change variable, Economic Freedom ~5-Year % ∆, as follows: 100 × ([Economic Freedom_t – Economic Freedom_t–5]/Economic Freedom_t–5).⁶ Although my use of an approximate five-year lag is informed by research that uses five-year interval data to operationalize changes in economic freedom over time (Gwartney et al. 1999), a five-year window is short from the perspective of this and other research as well as theory that dates the rise of neoliberalism to the 1970s and 1980s. Unfortunately, the data are in five-year intervals prior to 2000, and extending the temporal window would result in inconsistency beyond what is noted in Footnote 6.

Other Country-level Variables

Based on theories of economic freedom and corresponding to some previous research (Chung and van Oorschot 2011), I also include GDP Per Capita Growth Rate, which is a country’s annual growth in constant local currency and comes from the World Bank (data.worldbank.org/indicator/NY.GDP.PCAP.KD.ZG; last accessed September 2018). Another measure I include is the Unemployment Rate, which represents countries’ current unemployment rates; the data are from the World Bank (data.worldbank.org/indicator/SL.UEM.TOTL.ZS; last accessed September 2018). Theories of economic freedom and some previous research suggest that countries with higher GDP growth rates should have lower levels of worker insecurity; countries with higher unemployment rates should have higher levels of worker insecurity.

Individual-level Independent Variables

Research suggests that several individual-level variables should be associated with job and labor market insecurity (see Keim et al. 2014; Lee et al. 2018). In the analysis, I include part-time based on the OECD’s common definition (<30 hours/week = 1; ≥30 hours/week = 0), occupation based on 1988 International Standard Classification of Occupation (ISCO) codes (dummy variables for Managers, Professionals, Technicians and Associate Professionals [reference], Clerks, Service/Shop/Marketing Sales, Craft/Trade/Agricultural [combined ISCO categories], Operators and Assemblers, and “Elementary Occupations”), Age (in years), Age squared, sex (Female = 1), Education, contract type, and firm size.

Despite my attempt to use variables comparable to one another within and between survey datasets, Education, contract type, and firm size are measured at least somewhat differently: In the ESS, education is the number of years completed; in the EWCS, it refers to levels of education, with dummy variables for Lower secondary/second stage of basic education or less, Upper secondary (=reference), Postsecondary non tertiary, and Tertiary education. The indicator for contract type in the EWCS refers to “Indefinite,” and in the ESS, it is “Unlimited” (=1; otherwise = 0): It is coded this way (with “non-” temporary high) because the other categories do not exactly match. Firm size is measured via a set of dummy variables in all rounds/waves of all surveys (<10; 10–99 [reference]; 100–499; 500+), but in the 2015 EWCS only, another category, Missing data, is included to capture the large number of missing data.⁷

Descriptive statistics for selected independent variables are given in Table A1. Of note is the relatively little country-level variation in Economic Freedom (current year), and in the 2010 data, Economic Freedom ~5-Year % ∆ is negative (ESS) or nearly zero (EWCS). Although not in the table, correlations between the economic freedom measures and other macro-level variables that will be included in the same models are not suggestive of multicollineary, as the “average” r across datasets and country samples ranges from –.36 (Economic Freedom and Unemployment) to .53 (Economic Freedom ~5-Year % ∆ and GDP Growth). Yet, the strong relationship between the latter two variables in the 2004 ESS (r = −.85) is of some concern.

Analytic Strategy, Methods, and Models

My general analytic strategy is to examine each dependent measure and cross-section separately. The analysis makes use of multilevel modeling because the data consist of individuals nested in countries (Raudenbush and Bryk 2002). Hierarchical linear models are used for dependent variables treated as continuous (labor market insecurity measures from the ESS). Hierarchical ordered logit models are used for ordinal dependent variables (all job insecurity measures, plus labor market insecurity measures in the EWCS), with the coefficients reparameterized by multiplying them by –1. All country-level variables are grand-mean centered (i.e., centered around the overall average of all countries); individual-level variables are group-mean centered (i.e., centered around each country’s average). The models are relatively simple in that the coefficients are fixed, with the exception of randomly varying country-level intercepts. I use HLM 7.03 to estimate the models (Bryk, Raudenbush, and Congdon 2013), which importantly “uses the t distribution with degrees of freedom based on the number of groups . . . and . . . should give better inference for the fixed effect parameters” (Bryan and Jenkins 2016:8). Yet, hierarchical linear modeling (HLM) does not offer a means of relaxing the proportional odds assumption for ordered logit models to my knowledge, so I collapsed the ordinal dependent variables into three categories and used the “gllamm” command in Stata to run permutations of the final models (3 and 3a) below that allow the effects of all independent variables to vary across thresholds (i.e., nonproportional odds models; see Rabe-Hesketh and Skrondal 2012): The models as a whole violate the proportional odds assumption—and unemployment does so most consistently, among the country-level variables—but the substantive results are similar to the main analysis (available upon request).

The analysis proceeds by dependent variable—job insecurity followed by labor market insecurity—and is in separate parts: The first part focuses on Economic Freedom (current year) to test H1 and H1a, which are, respectively, derived from theories of economic freedom and neoliberalism and posit competing predictions about worker insecurity. The modeling strategy attempts to follow the causal logic of theories of economic freedom and neoliberalism. In addition to the individual-level variables, randomly varying country intercepts, and thresholds (as necessary) in “Model 0,” Model 1 adds Economic Freedom at the country level. Model 2 adds GDP Growth Rate at the country level, which is assumed to be influenced by economic freedom. Model 3 adds Unemployment Rate, at the country level, which neoliberalism may exacerbate (Bourdieu 1998). The second part of the analysis proceeds in the same way but examines Economic Freedom ~5-Year % ∆ in lieu of Economic Freedom (current year) to test H2 and H2a. As the number of variables in the Level 2 models increases, greater caution should be taken in interpreting the results, particularly in smaller samples (e.g., Bryan and Jenkins 2016).

A potential objection to this modeling strategy is that I do not include measures of economic freedom and freedom change in the same model. As Jakob De Haan and Jan-Egbert Sturm (2000) note—and Chris Doucouilagos and Mehmet Ali Ulubasoglu (2006) reiterate—such models may suffer from endogeneity problems. Some research on growth, often using time series analysis, includes a time-lagged measure of economic freedom and its over-time change (see De Haan and Sturm 2000). Adopting a similar strategy for this study—namely, using economic freedom_t–5 and economic freedom change—is not advisable, though, as these variables are generally highly correlated (the “average” r across datasets/samples = −.69, with a maximum of –.89).

Another set of objections concerns possible omitted variables at the country level. For instance, the relationship between economic freedom and insecurity may vary by economic development (Gehring 2013); economic development is also related to worker insecurity (Hipp 2016). Particularly because (logged) per capita GDP in current U.S. dollars is correlated with the current value of economic freedom (“average” r across datasets/samples = .64; maximum = 0.84 [2004 ESS]), however, I do not include this variable in the main analysis.

Results

Table 2 presents the results of hierarchical ordered logit models of job insecurity.⁸ The results in the top panel of this table speak to the hypothesized effects of economic freedom and neoliberalism (H1 and H1a), both operationalized as the current year of Economic Freedom. The results in the bottom panel speak to the theoretically posited effects of economic freedom change and neoliberalization (H2 and H2a), both operationalized as Economic Freedom ~5-Year % ∆.

Table 2.

Selected Results of Hierarchical Ordered Logit Models of Perceived Job Insecurity: ESS and EWCS.

Model #	Independent Variables & Fit	2004 ESS	2005 EWCS	2006 ESS	2008 ESS	2010 ESS	2010 EWCS	2015 EWCS	2016 ESS
Models With Economic Freedom, Current Year
M1	Economic Freedom, current year	−0.418*(0.169)	−0.707**(0.203)	−0.563**(0.199)	−0.941**(0.274)	−0.594^†(0.322)	−0.866*(0.335)	−0.555*(0.221)	−0.043(0.437)
	Pseudo R²	.226	.288	.277	.324	.128	.173	.156	.001
M2	Economic Freedom, current year	−0.671**(0.225)	−0.549**(0.177)	−0.351*(0.164)	−0.976**(0.292)	−0.581^†(0.319)	−0.905*(0.336)	−0.566*(0.225)	0.030(0.446)
	Per Capita GDP Growth Rate, annual	−0.084(0.054)	0.097**(0.027)	0.126**(0.033)	−0.014(0.041)	−0.056(0.065)	0.032(0.043)	0.005(0.023)	0.071(0.108)
	Pseudo R²	.304	.499	.569	.327	.152	.187	.157	.022
M3	Economic Freedom, current year	−0.495*(0.185)	−0.365^†(0.193)	−0.220(0.147)	−0.804*(0.302)	−0.559^†(0.312)	−0.576^†(0.294)	−0.279(0.276)	0.140(0.420)
	Per Capita GDP Growth Rate, annual	−0.084^†(0.043)	0.087**(0.026)	0.104**(0.029)	−0.005(0.039)	−0.060(0.064)	0.011(0.037)	−0.001(0.023)	0.072(0.100)
	Unemployment Rate, current year	0.080**(0.021)	0.054^†(0.028)	0.080**(0.027)	0.098(0.060)	0.040(0.036)	0.057**(0.015)	0.033(0.020)	0.069^†(0.038)
	Pseudo R²	.573	.555	.695	.389	.195	.436	.220	.154
Models With Economic Freedom ~5-Year % ∆
M1a	Economic Freedom ~5-Year % ∆	0.052*(0.024)	0.023^†(0.012)	0.040*(0.017)	0.069**(0.024)	0.024(0.051)	0.039(0.024)	0.017(0.031)	0.001(0.050)
	Pseudo R²	.182	.100	.207	.237	.008	.082	.009	.000
M2a	Economic Freedom ~5-Year % ∆	0.141**(0.040)	0.006(0.012)	0.008(0.017)	0.095**(0.031)	0.048(0.055)	0.046^†(0.027)	0.022(0.032)	−0.052(0.073)
	Per Capita GDP Growth Rate, annual	−0.178*(0.069)	0.111**(0.033)	0.140**(0.042)	−0.068(0.052)	−0.088(0.075)	−0.028(0.051)	−0.012(0.026)	0.151(0.153)
	Pseudo R²	.365	.341	.480	.283	.061	.090	.015	.046
M3a	Economic Freedom ~5-Year % ∆	0.092*(0.038)	0.003(0.010)	0.008(0.014)	0.089**(0.028)	0.059(0.053)	0.013(0.024)	0.041(0.029)	−0.033(0.069)
	Per Capita GDP Growth Rate, annual	−0.131*(0.061)	0.090**(0.030)	0.102**(0.036)	−0.061(0.047)	−0.098(0.073)	−0.014(0.043)	−0.017(0.023)	0.116(0.145)
	Unemployment Rate, current year	0.071**(0.023)	0.080**(0.026)	0.093**(0.027)	0.143*(0.054)	0.051(0.038)	0.063**(0.017)	0.050**(0.016)	0.064(0.038)
	Pseudo R²	.545	.502	.668	.436	.126	.374	.242	.158
	Country N	23	31	22	27	25	32	35	21

Note. Main entries are coefficients; standard errors are in parentheses. Country-level variables are grand-mean centered. All models include individual-level variables (see Table A1 and notes), which are group-mean centered and weighted, and thresholds. ESS = European Social Survey; EWCS = European Working Conditions Survey; GDP = gross domestic product.

†

p ≤ .10. *p ≤ .05. **p ≤ .01 (two-tailed t tests).

The pattern of results in Model 1 (M1)—in which only Economic Freedom (current year), individual variables, and thresholds (as necessary) are included—largely supports H1 derived from theories of economic freedom: The levels of job insecurity are generally lowest in countries with the highest levels of economic freedom in M1. Despite differences in the measurement of job insecurity, there is a relatively consistent significant negative relationship between Economic Freedom and job insecurity across datasets/years in M1, with the main exception of the 2016 ESS data; the relationship in the 2010 ESS data is only marginally significant (p ≤ .10). As judged by the country-level pseudo R² values, Economic Freedom (current year) accounts for between 0 and 32 percent of the cross-national variation in job insecurity, for an “average” (not shown here and throughout; calculated based on the table) of about 20 percent across years/datasets.⁹ The aforementioned significant and marginally relationships between economic freedom and job insecurity also hold once economic growth—a potential mediating factor, according to theories of economic freedom—is added in Model 2 (M2). Where this pattern begins to break down is in Model 3 (M3), which introduces unemployment: Some of the once-significant coefficients for economic freedom are insignificant (e.g., 2015 EWCS) or only marginally significant (e.g., 2010 EWCS) in this model. Altogether, the final model accounts for about between 15 and 70 percent of the cross-national variation in job insecurity, for an average of about 40 percent across datasets/years. Still, we must interpret these and other results cautiously, as some, especially in the ESS, are based on small Level 2 samples and the models are complex.

Compared with the results in the top panel of the table, the results of Model 1a (M1a) in the bottom panel of the table tell a partly different theoretical story: They are more supportive of a hypothesis drawn from theories of neoliberalism (H2a), positing a positive relationship between neoliberalization, as operationalized by Economic Freedom ~5-Year % ∆, and levels of job insecurity. Yet, the significant and marginally significant effects of this variable in Model 1a are limited to data/years prior to 2010. Moreover, Economic Freedom ~5-Year % ∆ in Model 1a accounts for a smaller proportion of the cross-national variation in job insecurity (Pseudo R² range ~= 0 to 24 % average ~= 10 percent) than Economic Freedom, current year, in Model 1 above. The introduction of economic growth in Model 2a (M2a) reveals that the consistency of the significance and marginal significance of Economic Freedom ~5-Year % ∆ is further limited, which is also the case in the final model (M3a) that controls for unemployment. Model 3a accounts for between about 13 and 67 percent of the cross-national variation in job insecurity, with a slightly lower average (~38 percent) than its counterpart (M3) above.

Do similar substantive results also hold for labor market insecurity? By and large, the answer is yes. The results of hierarchical linear and ordered logit models are displayed in Table 3, with those for Economic Freedom located in the top panel; those for Economic Freedom ~5-Year % ∆ are in the bottom panel. As H1 derived from theories of economic freedom predicts, the findings of M1 indicate that average levels of labor market insecurity are often lower in countries with higher levels of economic freedom across datasets/years, but this relationship is not significant in the 2010 EWCS and only marginally significant in the 2010 ESS. The average country pseudo R² value across datatsets/years in Model 1 is about 21 percent. The above pattern of significance and marginal significance of the coefficients for Economic Freedom persists in M2, but is much less consistent in M3, where Economic Freedom is only significant at conventional levels in the 2004 ESS. Across datasets/years, on average, about 40 percent of the cross-national variation in labor market insecurity is accounted for in the final model (M3).

Table 3.

Selected Results of Hierarchical Linear and Ordered Logit Models of Perceived Labor Market Insecurity: ESS and EWCS.

Model #	Independent Variables & Fit	2002 ESS	2004 ESS	2010 ESS	2010 EWCS^a	2015 EWCS^a
Models With Economic Freedom, Current Year
M1	Economic Freedom, current year	−1.112*(0.389)	−0.763**(0.241)	−0.569^†(0.305)	−0.457(0.331)	−0.507**(0.163)
	Country-Level Pseudo R²	.299	.323	.132	.056	.223
M2	Economic Freedom, current year	−1.081*(0.426)	−1.148**(0.319)	−0.551^†(0.298)	−0.396(0.328)	−0.518**(0.166)
	Per Capita GDP Growth Rate, annual	0.026(0.134)	−0.128(0.080)	−0.076(0.061)	−0.050(0.042)	0.005(0.017)
	Country-Level Pseudo R²	.300	.397	.183	.096	.225
M3	Economic Freedom, current year	−0.805(0.464)	−0.982**(0.304)	−0.505^†(0.260)	−0.143(0.309)	−0.075(0.174)
	Per Capita GDP Growth Rate, annual	0.042(0.130)	−0.127^†(0.070)	−0.083(0.053)	−0.066^†(0.039)	−0.004(0.014)
	Unemployment Rate, current year	0.069(0.053)	0.074*(0.035)	0.084*(0.030)	0.044*(0.016)	0.051**(0.013)
	Country-Level Pseudo R²	.357	.498	.394	.269	.481
Models With Economic Freedom ~5-Year % ∆
M1a	Economic Freedom ~5-Year % ∆	0.130**(0.043)	0.083*(0.036)	0.029(0.048)	0.006(0.023)	0.013(0.024)
	Country-Level Pseudo R²	.308	.199	.015	.002	.009
M2a	Economic Freedom ~5-Year % ∆	0.124*(0.045)	0.187**(0.063)	0.061(0.050)	0.024(0.025)	0.017(0.025)
	Per Capita GDP Growth Rate, annual	0.057(0.128)	−0.208^†(0.109)	−0.115(0.069)	−0.080(0.048)	−0.010(0.020)
	Country-Level Pseudo R²	.315	.307	.114	.084	.016
M3a	Economic Freedom ~5-Year % ∆	0.092^†(0.052)	0.140^†(0.067)	0.081^†(0.043)	0.000(0.024)	0.039*(0.018)
	Per Capita GDP Growth Rate, annual	0.067(0.124)	−0.163(0.107)	−0.133*(0.058)	−0.070(0.043)	−0.015(0.014)
	Unemployment Rate, current year	0.061(0.055)	0.066(0.041)	0.096**(0.030)	0.046**(0.017)	0.059**(0.009)
	Country-Level Pseudo R²	.355	.377	.385	.264	.546
	Country N	20	23	25	32	35

Note. Main entries are coefficients; standard errors are in parentheses. ESS = European Social Survey; EWCS = European Working Conditions Survey; GDP = gross domestic product.

Ordered logit models are used for these data (others are linear) and include thresholds. See Table 2 notes.

†

p ≤ .10. *p ≤ .05. **p ≤ .01 (two-tailed t tests).

Turning to the bottom panel of Table 3, the results of M1a reveal some support for H2a derived from theories of neoliberalism, but the expected positive relationship between neoliberalization—operationalized as Economic Freedom ~5-Year % ∆—is only significant in years/datasets prior to 2010. The average country pseudo R² value across datatsets/years in M1a is about 11 percent. The relationship between Economic Freedom ~5-Year % and labor market insecurity in the aforementioned years/datasets remains significant in M2a, but is only marginally significant in M3a; it is also marginally significant in the 2010 ESS and significant in the 2015 EWCS in M3a. This final model accounts for about 39 percent of the cross-national variation in labor market insecurity, on average, across datasets/years.

Supplementary Analyses

Thus far, the results of the first/basic models indicate some support for one hypothesis each derived from theories of economic freedom and neoliberalism, albeit with variation across survey years/data. Still, unemployment is a part of this story: Not only does it have a fairly consistent relationship to worker insecurity in the main analysis, but this relationship is even more consistent in models in supplementary analyses in which it is the only Level 2 fixed effect; in most instances, too, it accounts for a greater proportion of the cross-national variation in worker insecurity than economic freedom or its change (see Supplemental Table OS3).

In focusing on the effects of economic freedom and its change throughout all of the results, theories of economic freedom appear better equipped to account for the cross-national variation in job and labor market insecurity than theories of neoliberalism, as judged by the (qualitative) consistency of expected relationships, comparisons of pseudo R² values—especially in the EWCS that has larger samples—as well as alternative fit statistics (from mixed models run in Stata: see Supplemental Table OS4). This general conclusion also holds in supplementary analyses in which I excluded the government size area from the EFW index and its approximate five-year percentage change (see Supplemental Table OS5), given concerns about the reliability of the index and prior research indicating that it performs better when the government size area is excluded (Ott 2018). In fact, these results reveal more robust support for H1 derived from theories of economic freedom than in the main analyses, although this permutation of the EFW index has poor reliability in the 2016 ESS and is not significant in this survey dataset.

Finally, given the possible temporal pattern in the effects of neoliberalization in the main analysis, I conducted supplementary analyses of survey waves with identical job insecurity measures and identical (EWCS) or similar (ESS) labor market insecurity measures using the same country-level samples for each comparison (see Supplemental Table OS6). The significant and marginally significant effects of Economic Freedom ~5-Year % on job insecurity levels in the top panel of this table are limited to years prior to 2010 in the ESS, which is more suggestive of a year effect than a country-sample effect in this survey, although measurement differences in Economic Freedom ~5-Year % cannot be completely discounted (see Footnote 6) and bearing in cautions about the 2016 ESS. The results for job insecurity in the EWCS are more mixed. The ESS results in the bottom panel reveal the effects of Economic Freedom ~5-Year % on labor market insecurity levels are again limited to the pre-2010 years, but this can be due to a slight wording change in the dependent variable in 2010. There is no pre-2010 comparison for the EWCS labor market insecurity results, but across waves (the 2010 EWCS results are reproduced from Table 3) and models, only once is Economic Freedom ~5-Year % significant.

Conclusion

Economic and political changes bring with them new challenges to comparative theories in sociology, economics, and political science (Thelen 2012). By deriving and testing hypotheses deduced from theories of neoliberalism and economic freedom, this study attempts to meet such challenges and fill important gaps in research on worker insecurity.

The results of this study offer some support for one hypothesis each derived from theories of economic freedom and neoliberalism. In the simplest models, job and labor market insecurity levels are lower in European countries that are currently more “economically free” as theories of economic freedom anticipate; as theories of neoliberalism would expect, job and labor market insecurity levels tend to be higher in countries that have increased levels of economic freedom, or neoliberalization, in about the past five years, at least in some cases. Given that this study subjected these theories to a strong test across different forms of worker insecurity, years, and datasets and that the effects of economic freedom and its change often offer explanatory power, these theories contribute to literature on worker insecurity that has not focused on them.

Yet, these theories are limited, which was evident in the mixed results of more complex models and the variability of the findings. If neoliberalism and economic freedom are similar concepts and can be approximated in comparable ways, theories of neoliberalism do not appear to account for the cross-national variation in worker insecurity as well as theories of economic freedom. Specifically, the effects of neoliberalization appear to be selective (see Preminger 2016 for a similar point in a different context). For their part, theories of economic freedom need to better specify the mechanisms behind the relationship between economic freedom and worker insecurity levels in light of the association between economic freedom and development as well as the inconsistent effects of growth on worker insecurity levels (on the latter, see also Chung and van Oorschot 2011; Erlinghagen 2008; Lübke and Erlinghagen 2014). Alongside this are theoretical and methodological questions about the “government size” area of the EFW index, which may represent disagreement within each set of theories about how much and what type of intervention constitutes “economic freedom” or “neoliberalism” (De Haan et al. 2006; Jessop 2002; Ott 2018). The economic crisis may present a challenge to both sets of theories as well, if the somewhat different results in some of the survey data after 2008 are any indication.

As suggested above, a theoretical focus on neoliberalization and economic freedom alone is incomplete. Synthesizing disparate literatures suggests another—albeit somewhat speculative—interpretation, including some of the anomalous findings in the post-2008 survey data and the relatively consistent effects of unemployment in the current study and previous research (Dixon et al. 2013; Erlinghagen 2008; Esser and Olsen 2012; Green 2009; Hipp 2016; Lübke and Erlinghagen 2014): We may conceive of an “arc of neoliberalism,” as Miguel A. Centeno and Joseph N. Cohen (2012) call it, as an inverted u-shape figure with the apex representing its height. As countries move toward neoliberalism, particularly in rapid fashion over a short amount of time (i.e., neoliberalization), there is likely to be insecurity among workers. In the short run, insecurity may be mitigated if unemployment is held in check. In the long run, as countries achieve and maintain a certain level of institutional and economic stability, some workers may be less likely to perceive insecurity. This may be due to the “economic freedom” that institutions and policies provide or countries’ general institutional quality (Gwartney and Lawson 2003), the latter interpretation of which also resonates more with institutional research on worker insecurity; alternatively, this may be due to workers’ acceptance of the flexibility necessary in the contemporary economy (Pugh 2015). Although time is an important element of this process (Gwartney et al. 1999), countries’ abilities to fend off full-scale financial crises are as well (Calhoun 2011; Chung and van Oorschot 2011).

As acknowledged within, there are limitations to the data and analysis that bear on the conclusions. Yet, they also offer several directions for future research. Although no causal claims can be made about these data, future research can analyze longitudinal data, including beyond the time period investigated here, to test hypotheses suggested by the interpretation immediately above. Some of the different findings across years of the survey data pertain to the same insecurity measure in this study, but future research in the above vein can extend the current study by systematically testing such differences. Research can also examine whether the findings here exist in a larger set of more economically diverse countries, which may help to parse out the interrelationships among economic freedom, neoliberalization, and economic conditions. This study can partly speak to the reliability of the EFW index in Europe—and, indeed, there is instability—but research can further investigate this in different contexts and perhaps develop alternative measures of neoliberalism and neoliberalization (see Kwon 2016 for a similar point; see Williams 2017 for one such attempt).

Despite these limitations, this study contributes to literature on worker insecurity by testing alternative theories that have often been empirically neglected by research. The findings do not suggest that this neglect is warranted. With revisions and possibly synthesis, these theories can offer useful explanations in research on worker insecurity as well as perhaps in the broader, interdisciplinary literatures on neoliberalism and economic freedom.

Supplemental Material

Online_Supplemental – Supplemental material for Understanding Perceived Worker Insecurity in Europe, 2002–2016: Economic Freedom and Neoliberalism as Alternative Theories?

Supplemental material, Online_Supplemental for Understanding Perceived Worker Insecurity in Europe, 2002–2016: Economic Freedom and Neoliberalism as Alternative Theories? by Jeffrey C. Dixon in Sociological Perspectives

Footnotes

Appendix

Table A1.

Descriptive Statistics for Selected Independent Variables: ESS and EWCS, Various Years.

	2002 ESS	2004 ESS	2006 ESS	2008 ESS	2010 ESS	2016 ESS	2005 EWCS	2010 EWCS	2015 EWCS
Country level
Economic Freedom (current year): Fraser	7.52 (0.51)	7.52 (0.65)	7.54 (0.53)	7.38 (0.47)	7.37 (0.48)	7.63 (0.34)	7.48 (0.47)	7.36 (0.32)	7.50 (0.42)
Economic Freedom ~5-Year % ∆: Calculated from Fraser	3.50 (4.52)	2.27 (4.73)	3.96 (6.47)	1.93 (5.58)	−0.84 (3.20)	2.78 (2.99)	4.38 (8.51)	0.33 (4.68)	1.77 (3.29)
GDP Per Capita Growth Rate (annual): World Bank	1.62 (1.60)	4.12 (2.74)	4.16 (2.58)	0.84 (3.35)	1.59 (2.33)	2.05 (1.42)	3.72 (3.03)	1.59 (2.45)	3.05 (4.08)
Unemployment Rate (current year): World Bank/ILO	6.68 (4.07)	8.08 (4.20)	7.04 (2.85)	6.33 (2.17)	9.87 (4.14)	7.40 (3.64)	8.09 (3.25)	11.07 (5.96)	10.49 (5.75)
Individual level
Part-Time (<30 hours/week = 1; ≥30 hours/week = 0)	0.15 (0.35)	0.11 (0.32)	0.14 (0.34)	0.11 (0.31)	0.11 (0.32)	0.13 (0.32)	0.13 (0.35)	0.13 (0.34)	0.16 (0.37)
Indefinite/Unlimited Contract (=1; otherwise = 0)	0.84 (0.36)	0.81 (0.39)	0.79 (0.40)	0.79 (0.40)	0.82 (0.37)	0.82 (0.37)	0.77 (0.42)	0.78 (0.42)	0.78 (0.41)
Education (in years: ESS only)	12.99 (3.59)	12.87 (3.61)	13.43 (3.69)	13.34 (3.60)	13.58 (3.66)	14.08 (3.53)	—	—
Education (levels: EWCS only)	—	—	—	—	—	—
Lower secondary/second stage of basic education or less	—	—	—	—	—	—	0.19 (0.39)	0.20 (0.41)	0.15 (0.36)
Upper secondary (=reference)	—	—	—	—	—	—	0.42 (0.49)	0.43 (0.49)	0.42 (0.49)
Post secondary nontertiary	—	—	—	—	—	—	0.11 (0.32)	0.05 (0.22)	0.07 (0.26)
Tertiary education, first and second stage	—	—	—	—	—	—	0.28 (0.45)	0.32 (0.47)	0.36 (0.48)
Age (in years)	39.35 (11.20)	39.98 (11.19)	40.00 (11.53)	40.18 (11.51)	41.15 (11.28)	41.74 (11.80)	38.97 (11.35)	39.46 (11.38)	40.67 (11.61)
Female (=1; male = 0)	0.47 (0.50)	0.45 (0.50)	0.48 (0.50)	0.49 (0.51)	0.48 (0.50)	0.50 (0.50)	0.47 (0.50)	0.47 (0.50)	0.49 (0.50)
Ns
Country level	20	23	22	27	25	21	31	33	35
Individual level	14,491	14,974	15,772	18,712	15,876	15,775	21,228	27,383	29,389

Note. Descriptive statistics for age squared as well as the occupation and firm size dummy variable series are not presented due to space limitations. ESS = European Social Survey; EWCS = European Working Conditions Survey; GDP = gross domestic product.

Acknowledgements

A previous version of this paper was presented at the Social Science History Association Conference in Chicago (November 21–24, 2013), and I thank session participants for their comments. In addition, I thank Joseph Nathan Cohen, Andrew S. Fullerton, anonymous Sociological Perspectives reviewers, and the editor for their useful feedback on a previous version of this paper. I also thank Selina Gallo-Cruz and Zeynep Mirza for their helpful discussions about this paper. Moreover, I thank Joshua Hall for answering technical questions about the Economic Freedom of the World index used in this paper.

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 research, authorship, and/or publication of this article: This research was partly supported by an Ardizonne Junior Faculty Excellence Award (Summer 2013) through the College of the Holy Cross. Part of this paper was also completed during the author’s sabbatical, generously provided by the College of the Holy Cross.

Supplemental Material

Supplemental material for this article is available online.

Notes

Author Biography

Jeffrey C. Dixon is an Associate Professor in the Department of Sociology and Anthropology at the College of the Holy Cross. His current research focuses on worker insecurity, nonstandard work, neoliberalism, and the intersection among these phenomena. His previous research has appeared in such journals as Social Forces, Public Opinion Quarterly, European Sociological Review, and The British Journal of Sociology.

References

Anderson

Christopher J.

Pontusson

Jonas

. 2007. “Workers, Worries and Welfare States: Social Protection and Job Insecurity in 15 OECD Countries.” European Journal of Political Research 46(2):211–35.

Balz

Anne

. 2017. “Cross-national Variations in the Security Gap: Perceived Job Insecurity among Temporary and Permanent Employees and Employment Protection Legislation.” European Sociological Review 33(5):675–92.

Bandelj

Nina

Shorette

Kristen

Sowers

Elizabeth

. 2011. “Work and Neoliberal Globalization: A Polanyian Synthesis.” Sociology Compass 5(9):807–23.

Barnett

Clive

. 2010. “Publics and Markets. What’s Wrong with Neoliberalism?” Pp. 269–96 in The SAGE Handbook of Social Geography, edited by Smith

Susan J.

Pain

Rachel

Marston

Sallie A.

Jones III

John P

. London, England: Sage.

Beck

Ulrich

. 2000. The Brave New World of Work. Malden, MA: Blackwell.

Birch

Kean

. 2015. “Neoliberalism: The Whys and Wherefores . . . and Future Directions.” Sociology Compass 9(7):571–84.

Bourdieu

Pierre

. 1998. Acts of Resistance: Against the Tyranny of the Market. Translated by Nice

. New York: The Free Press.

Bryan

Mark L.

Jenkins

Stephen P

. 2016. “Multilevel Modelling of Country Effects: A Cautionary Tale.” European Sociological Review 32(1):3–22.

Bryk

Anthony S.

Raudenbush

Stephen W.

Congdon

Richard

. 2013. “HLM for Windows.” Version 7.03. Scientific Software International, Chicago, IL.

10.

Burchell

Brendan

. 2009. “Flexicurity as a Moderator of the Relationship between Job Insecurity and Psychological Well-being.” Cambridge Journal of Regions, Economy and Society 2:365–78.

11.

Calhoun

Craig

. 2011. “Series Introduction: From the Current Crisis to Possible Futures.” Pp. 9–42 in Business As Usual: The Roots of the Global Financial Meltdown, edited by Calhoun

Craig

Derluguian

Georgi

. New York: Social Science Research Council.

12.

Campbell

John L.

Pedersen

Ove K

. 2001. “The Rise of Neoliberalism and Institutional Analysis.” Pp. 1–24 in The Rise of Neoliberalism and Institutional Analysis, edited by Campbell

J. L.

Pedersen

O. K

. Princeton: Princeton University Press.

13.

Centeno

Miguel A.

Cohen

Joseph N

. 2012. “The Arc of Neoliberalism.” Annual Review of Sociology 38:317–40.

14.

Cheng

Grand H.-L.

Chan

Darius K.-S

. 2008. “Who Suffers More from Job Insecurity? A Meta-analytic Review.” Applied Psychology: An International Review 57(2):272–303.

15.

Chung

Heejung

. 2016. “Dualization and Subjective Employment Insecurity: Explaining the Subjective Employment Insecurity Divide between Permanent and Temporary Workers across 23 European Countries.” Economic and Industrial Democracy. Published electronically July 11. doi:10.1177/0143831X16656411

16.

Chung

Heejung

Mau

Steffen

. 2014. “Editorial: Subjective Insecurity and the Role of Institutions.” Journal of European Social Policy 24(4):303–18.

17.

Chung

Heejung

van Oorschot

Wim

. 2011. “Institutions versus Market Forces: Explaining the Employment Insecurity of European Individuals during (the Beginning of) the Financial Crisis.” Journal of European Social Policy 21(4):287–301.

18.

Clark

Andrew

Postel-Vinay

Fabien

. 2009. “Job Security and Job Protection.” Oxford Economic Papers 61(2):207–39.

19.

Cohen

Joseph N.

2011. “Is ‘Economic Freedom’ the Same as Free Market Capitalism? A Decompositional Analysis of the Economic Freedom of the World Index.” MPRA Paper No.33676. Retrieved April 2019 (http://mpra.ub.uni-muenchen.de/33676/).

20.

Crowley

Marth

Hodson

Randy

. 2014. “Neoliberalism at Work.” Social Currents 1(1):91–108.

21.

De Haan

Jakob

Lundström

Susanna

Sturm

Jan-Egbert

. 2006. “Market-oriented Institutions and Policies and Economic Growth: A Critical Survey.” Journal of Economic Surveys 20(2):157–91.

22.

De Haan

Jakob

Sturm

Jan-Egbert

. 2000. “On the Relationship between Economic Freedom and Economic Growth.” European Journal of Political Economy 16(2):215–41.

23.

Dickerson

Andy

Green

Francis

. 2012. “Fears and Realisations of Employment Insecurity.” Labour Economics 19(2):198–210.

24.

Dixon

Jeffrey C.

Fullerton

Andrew S.

Robertson

Dwanna L

. 2013. “Cross-national Differences in Workers’ Perceived Job, Labour Market, and Employment Insecurity in Europe: Empirical Tests and Theoretical Extensions.” European Sociological Review 29(5):1053–67.

25.

Doucouilagos

Chris

Ulubasoglu

Mehmet Ali

. 2006. “Economic Freedom and Economic Growth: Does Specification Make a Difference?” European Journal of Political Economy 22:60–81.

26.

Duménil

Gérard

Lévy

Dominique

. 2011. The Crisis of Neoliberalism. Cambridge, MA: Harvard University Press.

27.

Erlinghagen

Marcel

. 2008. “Self-perceived Job Insecurity and Social Context: A Multi-level Analysis of 17 European Countries.” European Sociological Review 24(2):183–97.

28.

Esping-Andersen

Gøsta

. 1990. The Three Worlds of Welfare Capitalism. Princeton, NJ: Princeton University Press.

29.

Esser

Ingrid

Olsen

Karen M

. 2012. “Perceived Job Quality: Autonomy and Job Security within a Multi-level Framework.” European Sociological Review 28(4):443–54.

30.

European Foundation for the Improvement of Living and Working Conditions. 2018. “European Working Conditions Survey Integrated Data File, 1991-2015” [Data collection], 7th ed., SN: 7363. UK Data Service [distributor]. doi:10.5255/UKDA-SN-7363-7.

31.

European Social Survey. 2018. “European Social Survey Database.” Retrieved September 2018 (http://www.europeansocialsurvey.org/).

32.

Fevre

Ralph

. 2007. “Employment Insecurity and Social Theory: The Power of Nightmares.” Work, Employment and Society 21(3):517–35.

33.

Foucault

Michel

. 2008. The Birth of Biopolitics: Lectures at the Collège De France, 1978-79. Edited by Senellart

Michel

, Translated by Burchell

Graham

. New York: Palgrave-MacMillan.

34.

Fourcade-Gourinchas

Marion

Babb

Sarah L

. 2002. “The Rebirth of the Liberal Creed: Paths to Neoliberalism in Four Countries.” American Journal of Sociology 108(3):533–79.

35.

Gallie

Ducan

. 2013. “Economic Crisis, Quality of Work, and Social Integration: Issues and Context.” Pp. 1–29 in Economic Crisis, Quality of Work and Social Integration: The European Experience, edited by Gallie

Duncan

. Oxford, England: Oxford University Press.

36.

Gash

Vanessa

Inanc

Hande

. 2013. “Insecurity and the Peripheral Workforce.” Pp. 142–168 in Economic Crisis, Quality of Work and Social Integration: The European Experience, edited by Gallie

Duncan

. Oxford, England: Oxford University Press.

37.

Gehring

Kai

. 2013. “Who Benefits from Economic Freedom? Unraveling the Effect of Economic Freedom on Subjective Well-being.” World Development 50:74–90.

38.

Green

Francis

. 2009. “Subjective Employment Insecurity around the World.” Cambridge Journal of Regions, Economy and Society 2(3):343–63.

39.

Gwartney

James D.

Lawson

Robert A

. 2003. “The Concept and Measurement of Economic Freedom.” European Journal of Political Economy 19:405–30.

40.

Gwartney

James D.

Lawson

Robert A.

Hall

Joshua

Murphy

Ryan

. 2018. “Economic Freedom of the World: 2018 Annual Report.” Fraser Institute. Retrieved October 19, 2018 (www.fraserinstitute.org/studies/economic-freedom).

41.

Gwartney

James D.

Lawson

Robert A.

Holcombe

Randall G

. 1999. “Economic Freedom and the Environment for Economic Growth.” Journal of Institutional and Theoretical Economics 155(4):643–63.

42.

Hall

Joshua C.

Lawson

Robert A

. 2014. “Economic Freedom of the World: An Accounting of the Literature.” Contemporary Economic Policy 32(1):1–19.

43.

Hall

Peter A.

Soskice

David

. 2001. “An Introduction to Varieties of Capitalism.” Pp. 1–70 in Varieties of Capitalism, edited by Hall

Peter A.

Soskice

David

. New York: Oxford University Press.

44.

Harvey

David

. 2005. A Brief History of Neoliberalism. New York: Oxford University Press.

45.

Harvey

David

. 2006. “Neo-liberalism as Creative Destruction.” Geographic Annuals 88:145–58.

46.

Hipp

Lena

. 2016. “Insecure Times? Workers’ Perceived Job and Labor Market Security in 23 OECD Countries.” Social Science Research 60:1–14.

47.

Jessop

Bob

. 2002. “Liberalism, Neoliberalism, and Urban Governance: A State-theoretical Perspective.” Antipode 34(3):452–72.

48.

Kalleberg

Arne L.

2009. “Precarious Work, Insecure Workers: Employment Relations in Transition.” American Sociological Review 74:1–22.

49.

Kalleberg

Arne L.

2018. Precarious Lives: Job Insecurity and Well-being in Rich Democracies. Cambridge, England: Polity Press.

50.

Kalleberg

Arne L.

Vallas

Steven P

. 2017. “Probing Precarious Work: Theory, Research, and Politics.” Pp. 1–30 in Precarious Work. Research in the Sociology of Work, vol. 31, edited by Kalleberg

Arne L

. Bingley, England: Emerald.

51.

Keim

Alaina C.

Landis

Ronald S.

Pierce

Charles A.

Earnest

David R

. 2014. “Why Do Employees Worry about Their Jobs? A Meta-analytic Review of Predictors of Job Insecurity.” Journal of Occupational Health Psychology 19(3):269–90.

52.

Korpi

Walter

. 2006. “Power Resources and Employer-centered Approaches in Explanations of Welfare States and Varieties of Capitalism: Protagonists, Consenters, and Antagonists.” World Politics 58(2):167–206.

53.

Kwon

Roy

. 2016. “Can We Have Our Cake and Eat It, Too? Liberalization, Economic Growth, and Income Inequality in Advanced Industrial Societies.” Social Forces 95(2):469–502.

54.

László

Krisztina D.

Pikhart

Hynek

Kopp

Mária S.

Bobak

Martin

Pajak

Andrzej

Malyutina

Sofia

Salavecz

Gyöngyvér

Marmot

Michael

. 2010. “Job Insecurity and Health: A Study of 16 European Countries.” Social Science Medicine 70(6):867–74.

55.

Lee

Cynthia

Huang

Guo-Hua

Ashford

Susan J

. 2018. “Job Insecurity and the Changing Workplace: Recent Developments and the Future Trends in Job Insecurity Research.” Annual Review of Organization Psychology and Organizational Behavior 5:335–59.

56.

Lübke

Christiane

Erlinghagen

Marce

. 2014. “Self-perceived Job Insecurity across Europe Over Time: Does Changing Context Matter?” Journal of European Social Policy 24(4):319–36.

57.

Mai

Quan

. 2017. “Precarious Work in Europe: Assessing Cross-national Differences and Institutional Determinants of Work Precarity in 32 Countries.” Pp. 273–306 in Precarious Work. Research in the Sociology of Work, vol. 31, Kalleberg

Arne L.

Vallas

Steven P

. Bingley, England: Emerald.

58.

Miller

Terry

Kim

Anthony B.

Roberts

James M

. 2018. 2018 Index of Economic Freedom. Washington, DC: The Heritage Foundation.

59.

North

Douglass C.

1990. Institutions, Institutional Change and Economic Performance. Cambridge, England: Cambridge University Press.

60.

Ott

Jan

. 2018. “Measuring Economic Freedom: Better without Size of Government.” Social Indicators Research 135:479–98.

61.

Ovaska

Tomi

Takashima

Rya

. 2006. “Economic Policy and the Level of Self-perceived Well-being.” The Journal of Socio-Economics 35(2):308–25.

62.

Peck

Jamie

. 2010. Constructions of Neoliberal Reason. Oxford, England: Oxford University Press.

63.

Pitlik

Hans

Rode

Martin

. 2016. “Free to Choose? Economic Freedom, Relative Income, and Life Control Perceptions.” International Journal of Wellbeing 6(1):81–100.

64.

Preminger

Jonathan

. 2016. “The Contradictory Effects of Neoliberalization on Labour Relations: The Health and Social Work Sectors.” Economic and Industrial Democracy 37(4):644–64.

65.

Pugh

Allison J.

2015. The Tumbleweed Society: Working and Caring in an Age of Insecurity. Oxford, England: Oxford University Press.

66.

Rabe-Hesketh

Sophia

Skrondal

Anders

. 2012. Multilevel and Longitudinal Modeling Using Stata, Vol. II: Categorical Responses, Counts, and Survival. 2nd ed. College Station, TX: Stata Press.

67.

Raudenbush

Stephen W.

Bryk

Anthony S

. 2002. Hierarchical Linear Models: Applications and Data Analysis Methods. 2nd ed. Thousand Oaks, CA: Sage.

68.

Silva

Jennifer M.

2013. Coming Up Short: Working-class Adulthood in Age of Uncertainty. Oxford, England: Oxford University Press.

69.

Standing

Guy

. 2016. The Precariat: The New Dangerous Class. London, England: Bloomsbury.

70.

Stegmueller

Daniel

. 2013. “How Many Countries for Multilevel Modeling? A Comparison of Frequentist and Bayesian Approaches.” American Journal of Political Science 57(3):748–61.

71.

Thelen

Kathleen

. 2012. “Varieties of Capitalism: Trajectories of Liberalization and the New Politics of Social Solidarity.” Annual Review of Political Science 15:137–59.

72.

Veenhoven

Ruut

. 1999. “Quality-of-Life in Individualistic Society: A Comparison of 43 Nations in the Early 1990’s.” Social Indicators Research 48:157–86.

73.

Williams

Colin C.

2017. “Tacking Employment in the Informal Economy: A Critical Evaluation of the Neoliberal Policy Approach.” Economic and Industrial Democracy 38(1):145–69.

74.

Williamson

John

. 1993. “Democracy and the ‘Washington Consensus.’” World Development 21(8):1329–36.

75.

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 10(2):166–86.

76.

Wright

Erik O.

2000. “Working-class Power, Capitalist-Class Interests, and Class Compromise.” American Journal of Sociology 105(4):957–1002.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.62 MB