Paper - Web view · 2010-02-20Word count = 7022 *Full paper ... As Beck states, ... and...
Transcript of Paper - Web view · 2010-02-20Word count = 7022 *Full paper ... As Beck states, ... and...
International Differences in Perceptions 1
International Differences in Perceptions of Temporary Work: The Role of Relativity*
Lindsey M. KingUniversity of North Carolina at Chapel Hill
Word count = 7022
*Full paper submitted to the International Labour Process Conference, March 15 – 17, 2010, Rutgers, NJDirect all correspondence to Lindsey M. King, Department of Sociology, CB #3210 Hamilton Hall, University of North Carolina, Chapel Hill, North Carolina 27599-3210 ([email protected]).
International Differences in Perceptions 2
Abstract
This research project explores the relative desirability of unemployment and temporary
work to understand how workers evaluate their career options. The decision-making
process is proposed to be influenced by welfare state and labor market structures:
employment protection legislation (EPL), prevalence of temporary employment, passive
labor market policies (LMP), and the unemployment rate. The first set of analyses test
whether individual welfare state components hold greater explanatory power than clusters
of characteristics as described by welfare regimes. Using data from the International
Social Survey Program (ISSP), country-level explanations are tested against the
predictive power of individual work values and job characteristics of currently-employed
workers. Overall, the valuation of job security significantly predicts workers’ willingness
to accept temporary employment, although the strength of the relationship varies between
countries. Perceptions of job security and employability reduce the likelihood that
workers will choose temporary work over unemployment, while the presence and
valuation of interesting work increases the likelihood. The preference for temporary work
over unemployment is explained somewhat by welfare regimes, but it appears that
workers’ values and perceptions of their current jobs hold far more explanatory power.
International Differences in Perceptions 3
International Differences in Perceptions of Temporary Work:
The Role of Relativity
The rise of precarious employment means that workers must often face an
unpleasant reality of choosing between less-than-ideal employment options. If “[b]eing
employed on a fixed term contract at some point in the course of a career is…the rule
rather than the exception” (Auer & Cazes 2003b: 46), understanding how workers
perceive temporary jobs as an employment option is of critical importance. If temporary
jobs are viewed as (1) a “trap”—part of an inescapable cycle of precarious employment
—rather than a bridge to permanent work (Booth, Francesconi, & Frank 2002), (2) “bad”
jobs in comparison to permanent employment (Kalleberg, Reskin & Hudson 2000), or (3)
an employment situation workers choose out of necessity rather than a desire for
flexibility (Feldman, Doerpinghaus, & Turnley 1995), workers may evaluate these jobs
unfavorably.
To comprehend how workers evaluate their options, decisions must be analyzed
with consideration of their relative attractiveness to workers (Gallie & Alm 2000).
Generally, unemployment is cast as the ultimate undesirable employment outcome, yet
for some workers the drawbacks of temporary work may outweigh the drawbacks of
unemployment. In unemployment, there exists a strong possibility for skill degradation
which increases exponentially as the length of unemployment increases (Gangl 2006).
Yet many countries offer unemployment benefits with a fairly high replacement rate,
reducing at least the wage-loss aspect of unemployment. Workers may underestimate the
length of time they will be unemployed, believing they will move quickly into a new job
and therefore believing the risks of unemployment to be minimal (Feinberg 1977). Thus,
International Differences in Perceptions 4
many workers may opt for unemployment if they lose their current job, believing a
temporary job to be an disagreeable and unnecessary step.
Although precarious forms of employment have undoubtedly gained a larger
share of the labor force since the 1980s (Booth, Dolado, & Frank 2002a), nations have
developed vastly different responses to precarity and risk. As Beck states, “All Western
countries are similarly affected by the informalization and individualization of paid work.
But this epochal change is perceived and valued differently in different cultures.” (2000:
111). How nations handle the risks inherent in labor markets is central to understanding
how workers evaluate their employment trajectories.
In this paper, I explore how currently-employed workers weigh the relative
desirability of unemployment and temporary work. When forced to select one or the
other, which will workers choose? The extent to which workers view temporary work as
a riskier prospect than unemployment indicates the degree to which temporary jobs are
viewed as “bad.” The decision-making process is proposed to be influenced by welfare
state and labor market structures: unemployment insurance, legal protections for
temporary workers, and the like. The explanatory value of welfare state structures will be
tested against the predictive power of individual value systems and current job
characteristics.
WELFARE STATE AS RISK MITIGATOR
Welfare states frame labor markets: Employment trajectories are strongly
determined by the legal and political framework in which labor markets are embedded.
Allocation of people to jobs can be facilitated or hindered by the institutional structures,
and workers move through labor markets bounded to various degrees by systems of rules
International Differences in Perceptions 5
governing their mobility. The interconnectedness of welfare states and labor markets
means that labor markets can only be understood in terms of the overarching political
agenda (Esping-Andersen 1990).
The welfare state is, at its core, a system of risk management for its citizens.
Edwards and Glover (2001) provide a succinct summary of the function of the welfare
state:
The traditional model for extensive or developed welfare states is an institutional nation-state response to coping with unforeseen but broadly predictable consequences, based on actuarial principles and collectively shared rights and responsibilities. It is founded on a solidaristic and state-centered response to the meeting of risks encountered within a typical lifecourse on the basis of broad social obligation as part of citizenship, together with the creation of mutual security. (P. 3)
In this conceptualization, strong welfare states make ample provisions for the risks
people are likely to encounter during the course of their lives. For workers, some of these
risks include unemployment, involuntary job loss, skill obsolescence, insufficient wages,
or insufficient work. Weak welfare states leave risk management to individuals or
employers; concertive responses to labor market failures or fluctuations are minimal.
WELFARE REGIMES
Welfare regimes categorize countries according to the level of decommodification
—the extent to which citizens are reliant on paid employment for their material well-
being—and stratification experienced by its citizens (Arts 2002). Esping-Andersen’s
(1990) original conceptualization of welfare regimes contained three categories: liberal,
social democratic, and conservative. Liberal regimes contain deregulated labor markets
and high citizen commodification: Limited social protection means that citizens are
reliant on an uninterrupted career of paid employment for income maintenance. Social-
International Differences in Perceptions 6
democratic regimes emphasize decommodification and equality. Labor market
protections are enjoyed by a broad spectrum of workers who are not exclusively
dependent on employers for income. Benefits are universal and redistribution high.
Conservative regimes offer moderate decommodification, but benefits are tightly linked
to occupational status (Arts 2002). Workers with an occupational affiliation receive
strong protections, and conservative regimes provide ample income security to cushion
the impact of labor market turbulence.
The limited scope of Esping-Andersen’s original formulation omits many
countries; repeated attempts have been made to expand the typology’s purview beyond
three worlds of welfare capitalism. Other countries that show distinct similarities can be
grouped into the Mediterranean, post-communist, and productivist regimes.
Characterized by strong worker protections and moderate-to-high decommodification,
Mediterranean regimes display high labor market stratification. Certain classes of
workers receive extremely high protection, lessening their reliance on the state for
income provision, but other workers are effectively shut out of the labor market and
depend on social welfare programs for income (Dolado, García-Serrano, & Jimeno 2002).
Still left with the residue of a centrally-planned economy, post-communist
welfare regimes have historically contained labor markets entirely dominated by the
state: “[U]nder the full employment policy it was the responsibility of the state to
provide employment to all able-bodied persons of working age” (Cazes & Nesporova
2003: 7). Employment protection has since been softened, although economic limitations
prevent the state from being able to supplement increased labor market turbulence with
International Differences in Perceptions 7
income protection. The lingering effects of communism mean that low stratification
remains a priority.
In productivist regimes, the company takes over functions usually performed by
the state. Holliday (2000) describes the focus and welfare strategy of productivist
regimes:
Here, social policy is strictly subordinate to the overriding policy objective of economic growth. Everything else flows from this: minimal social rights with extensions linked to productive activity, reinforcement of the position of productive elements in society, and state-market-family relationships directed towards growth. (P. 708)
Decommodification is therefore low and stratification high. Those without access to the
social safety nets offered by employers have little recourse for income maintenance. Full-
time, tenured employment is therefore highly desirable as citizen welfare depends on
employer magnanimity. Social structure is shaped more by norms than legislation, a
feature unique to the productivist regime type (Passet 2003).
Having outlined the major regime types under analysis, I shall now elaborate on
the structure of temporary work and unemployment in each regime.
TEMPORARY WORK AND THE WELFARE STATE
Employment protection legislation (EPL) governs the conditions under which
employers may fire individual workers, lay off groups of workers, or employ temporary
workers. “The primary task of EPL is to give more employment and income security to
workers, both in their current jobs and in the case of redundancy” (Cazes & Nesporova
2003: 3), but legislation may have unintended effects. Strong EPL for permanent workers
may cause employers to rely more heavily on temporary employees as a means of
circumventing strict restrictions on the hiring and firing of permanent employees. If
International Differences in Perceptions 8
permanent workers cannot be easily fired when employers desire or require flexibility,
the converse implication is that employers will be reluctant to hire permanent workers,
preferring instead to deploy temporary workers (Employment in Europe 2006).
The most notable instance of this perverse effect of EPL is Spain. Although
temporary work was originally implemented to reduce high unemployment rates, EPL for
permanent workers remained high. Creating flexibility in one arena while maintaining
rigidity in another resulted in over-reliance on temporary workers: Over one-third of
Spain’s labor force is currently on a temporary contract (Polavieja 2003). Despite
attempts to reduce the disparity, Spain retains a two-tier labor market in which the market
for permanent work is difficult to penetrate (Polavieja 2003).
Figure 1 summarizes the proportion of temporary workers in the labor force as a
function of employment protection legislation. Liberal regimes (represented by the
United States, the United Kingdom, Australia, Canada, and Ireland) cluster in the lower-
left corner of the scale, indicating that this regime type tends to be low on both EPL and
temporary work. Mediterranean regimes (Spain and Portugal) exhibit the opposite
pattern: EPL is extremely high, as is the proportion of temporary workers. The social
democratic regimes (Sweden, Norway, Finland, and Denmark) are centrally located on
the scale, although their use of temporary workers is generally lower than their levels of
EPL would suggest. Post-communist regimes (Hungary and the Czech Republic) are
moderate to low on EPL and display a correspondingly low incidence of temporary work.
Although productivist regimes (Japan and South Korea) have moderate employment
protection legislation, the prevalence of temporary work is higher than expected. In South
Korea, for example, a massive 36% of workers are on fixed-term contracts. Conservative
International Differences in Perceptions 9
regimes (Germany, France, Belgium, and Switzerland) offer moderate to high levels of
employment protection, and the incidence of temporary work is similarly moderate to
high.
Figure 1. Strength of Employment Protection Legislation and the Proportion of Temporary Workers in the Labor Force, 2005
0 0.5 1 1.5 2 2.5 3 3.5 40
5
10
15
20
25
30
35
40
Australia
Germany
UKUS
Hungary
Ireland
Norway
Sweden
Czech Republic
Canada
Japan
Spain
France
Portugal
Denmark
Switzerland
Belgium
Finland
South Korea
Employment Protection Legislation
Tem
-po
-ra
ry
Wor
ker
s as
a Pr
o-
por-
tion
of
the
Civ
il-
ian
Labo
r Fo
rce
Notes: Data come from the Organisation for Economic Co-operation and Development (OECD). Employment protection legislation is a weighted average of regulations covering regular workers, temporary workers, and collective dismissals. The scale ranges from 0 (least restrictive) to 6 (most restrictive).
UNEMPLOYMENT AND THE WELFARE STATE
“While unemployment is not the only risk temporary workers face…temporary
workers are also exposed to repeat spells of temporary work,…unemployment remains a
clear indicator of temporary work’s inability to integrate workers” (Gash 2008: 651). To
International Differences in Perceptions 10
the extent that temporary work becomes a “trap” of endless spells of temporary contracts
and unemployment rather than a “bridge” into permanent work (Booth et al. 2002),
workers will avoid temporary jobs. Workers’ ability to search for alternate employment
depends heavily on their ability to maintain an income during spells of unemployment.
For this reason, the amount of state spending on passive labor market protections (LMP),
such as unemployment benefits and early retirement, is associated with high
unemployment rates partly because workers have the luxury of seeking a job that fits
their skills and expectations (Gangl 2008; Nickell 1997). In countries with weak passive
LMP, unemployment is a highly undesirable prospect for workers due to the high
likelihood of poverty (Gallie & Paugam 2000). Low replacement rates encourage workers
to find new employment quickly, regardless of how well the new job matches their skill
set.
Figure 2 compares passive labor market policies to the unemployment rate. With
the exception of Switzerland, conservative regimes provide strong unemployment
protection to workers and unemployment rates in these countries are among the highest
of all regime types. Mediterranean regimes also tend to have high unemployment rates,
although their spending on passive labor market policies is slightly less than that of
conservative regimes. Social democratic regimes vary widely in their spending on passive
labor market policies, from the substantial expenditures of Denmark to the more modest
spending levels of Norway. Unemployment is similarly varied: Norway and Demark
enjoy extremely low unemployment rates, while the unemployment rates of Finland and
Sweden are higher than expected. Liberal regimes, once again, cluster at the low end of
both scales: expenditures on passive labor market policies are exceedingly low and are
International Differences in Perceptions 11
Figure 2. Passive Labor Market Policies and the Unemployment Rate, 2005
0 0.5 1 1.5 2 2.5 30
2
4
6
8
10
12
Australia
Germany
UK US
Hungary
Ireland
Norway
SwedenCzech Republic
Canada
Japan
SpainFrance
Portugal
DenmarkSwitzerland
BelgiumFinland
South Korea
Passive Labor Market Policies
U n e m p l o y m e n t R a t e
Notes: Data come from the Organisation for Economic Co-operation and Development (OECD). Passive labor market policies measure government expenditures as a percent of GDP spent on unemployment income maintenance and early retirement schemes.
matched with correspondingly low unemployment. Productivist regimes resemble liberal
regimes in their minimal spending on passive benefits such as unemployment insurance.
Productivist regimes also parallel liberal regimes in exhibiting low levels of
unemployment. Spending on passive labor market policies in post-communist regimes
compares to levels in liberal and productivist regimes but unemployment is much higher
than other regimes with similar spending levels.
International Differences in Perceptions 12
Hypotheses
From the outline of welfare regimes and welfare state structures described above,
two competing hypotheses are presented. The first hypothesis proposes that the
aforementioned components of labor markets—employment protection legislation,
temporary workers as a proportion of the labor force, passive labor market policies, and
the unemployment rate—individually affect workers’ willingness to accept temporary
employment. The competing hypothesis proposes that clusters of labor market
characteristics, as embodied by welfare regime typologies, are better able to predict
workers’ acceptance of temporary employment.
WORK VALUES AND PERCEIVED JOB CHARACTERISTICS
Work Values: The importance of job security
Work values encapsulate the desired characteristics of a work situation (Judge &
Bretz 1992; Kalleberg 1977). As defined by Kalleberg and Reynolds (2003), work values
encompass:
1) the extent to which people are involved in work, or the centrality of work to their lives; 2) the importance people place on various job facets, or their ‘conceptions of the desirable’ with regard to their work activity; and 3) the extent to which people work in the particular arrangement voluntarily or involuntarily.” (P. 425)
The “conceptions of the desirable” pertain to workers’ attraction to or revulsion from
temporary work: to the extent that temporary employment matches desired work
characteristics, workers should seek out, or at least not actively avoid, this type of
employment arrangement. To the degree that the precarious nature of temporary work
fails to fit workers’ ideals, these types of jobs should be avoided when possible.
International Differences in Perceptions 13
Work values are hypothesized to affect the willingness to accept temporary work
thusly: Workers who value the extrinsic rewards of work—“ rewards provided by the
organization for the purpose of facilitating or motivating task performance” (Mottaz
1985: 366) such as job security, promotion prospects, and pay—should show an aversion
to temporary work. Conversely, workers who value the intrinsic rewards of work—
rewards “ derived from the content of the task itself” (Mottaz 1985: 366) should prefer
temporary work over no job at all because work itself is seen as a rewarding undertaking.
The relationship between work values and willingness to accept temporary work
may differ across countries. Specifically, the relationship between job security valuation,
an extrinsic work reward, and willingness to accept temporary employment may vary.
Denmark, for instance, is well known for its system of “flexicurity:” flexible work
combined with high levels of social protection (Madsen 2008). The combination removes
the uncertainty usually associated with temporary work, giving rise to labor markets
characterized by high flexibility. Because job security in Denmark is not associated with
“good” jobs, the link between job security valuation and willingness to accept temporary
work may be weak.
Conversely and as previously noted, Spain exhibits a two-tier labor market in
which it is difficult for temporary workers to transition into permanent work (Polavieja
2003). A similar scene unfolds in Japan: Permanent workers enjoy a host of protections
provided by their employers, while temporary and other non-standard workers are
relatively deprived (Passet 2003). The divide between “good” and “bad” jobs being
determined by permanent or temporary employment status, it seems that job security
International Differences in Perceptions 14
valuation would exhibit a strong relationship with amenability to temporary work in these
cases.
The hypothesis derived from these examples is broad, and designed to test if job
security valuation does indeed vary in the strength of its relationship to workers’
willingness to accept temporary employment.
Job Characteristics and Aversion to Temporary Employment
De Witte and Näswall (2003) differentiate between objective and subjective job
insecurity. Objective insecurity is a state of precariousness deriving from a time-limited
contract or employment with an uncertain future. Subjective insecurity describes
workers’ perceptions that job loss is possible or imminent. Thus, temporary workers, who
are employed in an objectively insecure type of work, also display high levels of
subjective job insecurity. Job insecurity is not solely the domain of temporary workers,
however: Any worker may be concerned about the future of their job, and the incidence
of job insecurity even among permanent workers is well documented (Green 2006;
Bockerman 2004; De Cuyper & De Witte 2005).
Hypotheses regarding perceived job characteristics mirror those of job values:
workers who perceive their jobs to be secure should show a lower inclination toward
temporary work than workers who feel insecure in their jobs. Also, workers who feel
their jobs are interesting should exhibit a stronger desire to remain employed, even in
temporary work, than enter unemployment. A related construct is employability, “a form
of work specific adaptability that enables workers to identify and realize career
opportunities” (Fugate, Kinicki, & Ashforth 2004: 16). Both job security and
employability imply control over one’s employment outcomes. In the case of job
International Differences in Perceptions 15
security, workers feel that their future with their current employer is certain and
predictable; in the case of employability, workers feel they would be able to easily find
another job were the current job to be involuntarily lost or voluntarily abandoned. Both
employability and job security indicate that the worker believes himself or herself to be in
a strong labor market position (De Cuyper et al. 2008). These workers should exhibit a
preference for unemployment, a seemingly low-risk endeavor for a highly employable
worker, over temporary work.
METHOD
Data
Individual-level data.
Individual-level data come from the International Social Survey Program (ISSP),
a cross-national survey designed to capture attitudes on various topics. The ISSP includes
countries with a broad range of political and economic configurations and is thus ideally
suited for cross-national research. One of the primary goals of the ISSP is to make cross-
national surveys as comparable as possible. This by no means eliminates the possibility
that concepts are understood differently across nations (Hult, 2005), but this confounding
factor is given special attention in the drafting of the survey, reducing the chance that any
finding of international differences is merely a product of research design.
This research project will focus exclusively on the 2005 Work Orientations
Module. Respondents are asked about their preferences for various employment
situations and job characteristics; and are also asked to evaluate their current work
situation. The analytic sample is restricted to non-agricultural workers ages 18 to 65 who
are working for pay at the time of the survey. Workers who are employed less than part-
International Differences in Perceptions 16
time, here defined as less than ten hours a week, are omitted. A total 12,822 cases are
eligible for inclusion in the analysis.
Country-level data.
The Organisation for Economic Cooperation and Development (OECD) compiles
detailed information on welfare state components. All information on the welfare state
components is extracted from the OECD Employment and Labor Market Statistics
database. Although the ISSP contains more countries than are included in the present
study, data were unavailable on employment policies and labor force characteristics for
many countries. Countries for which the OECD lacked the requisite data were omitted,
leaving nineteen countries in the sample: Australia, Belgium, Canada, the Czech
Republic, Denmark, Finland. France, Germany, Great Britain, Hungary, Ireland, Japan,
Norway, Portugal, South Korea, Spain, Sweden, Switzerland, and the United States.
Dependent Variable
Preference for temporary work over unemployment.
Workers’ preference for temporary work over unemployment is measure by a
single item. Respondents are asked their level of agreement with the statement “In order
to avoid unemployment I would be willing to accept temporary employment.” Response
categories range from “strongly disagree” to “strongly agree,” with higher scores
indicating a higher willingness to accept temporary work rather than enter
unemployment.
International Differences in Perceptions 17
Independent Variables
Country-level variables.
As previously described, welfare regimes are classified according to the expanded
“Three Worlds” typology developed by Esping-Andersen (1990) and elaborated by Arts
(2007). The liberal welfare regime type is represented by Australia, Canada, Great
Britain, Ireland and the United States. Denmark, Norway, Sweden, and Finland represent
the social democratic regime type. Conservative regimes include Germany, Switzerland,
France, and Belgium. Hungary and the Czech Republic represent the post-communist
regime type. Spain and Portugal exemplify the Mediterranean regime type in the dataset.
The productivist regime type is represented here by South Korea and Japan.
The OECD distills measures of employment legislation into a single-item scale
indicating the overall strictness of EPL. The scale contains three components: “(i)
employment protection of regular workers against individual dismissal; (ii) specific
requirements for collective dismissals; and (iii) regulation of temporary forms of
employment” (OECD N.d.: 1). The scale ranges from zero to six: Low scores correspond
to weak employment protection, whereas high scores indicate stringent employment laws.
The prevalence of temporary work is measured as a proportion of the civilian
labor force, as is the unemployment rate. Passive labor market policies measure public
expenditures on programs such as unemployment insurance and early retirement as a
percentage of GDP.
Individual-level variables: work values.
Measures of work values encompass both intrinsic and extrinsic characteristics.
Extrinsic work values assess the importance placed on job security, advancement, and
International Differences in Perceptions 18
high wages. Intrinsic work values include the importance of interesting work,
independent work, and flexible scheduling. For each value, respondents are asked to
indicate how important each aspect is to them. The scale ranges from 1 to 5, with 1
signaling “very unimportant” and 5 indicating “very important.” Respondents are also
asked to indicate whether they would prefer to work full-time or at most part-time.
Individual-level variables: Perceived job characteristics.
Perceived job characteristics assess the same dimensions as the job value
measures outlined above, with the exception of preferences for full-time versus part-time
work. Respondents are asked to indicate how strongly they agree that their job possesses
the aforementioned characteristics. Scores range from 1, which indicates strong
disagreement, to 5, which represents strong agreement.
To assess perceived external labor market prospects, workers are asked, “How
easy or difficult do you think it would be for you to find a job at least as good as your
current one?” Scores are reverse coded such that higher scores correspond to a greater
perceived ease in finding an equivalent job. Table 1 provides the means for job values,
perceived job characteristics, and perceptions of employability, averaged across
countries.
Control variables: Demographics and job characteristics.
Analyses control for gender, age, marital status, presence of a working spouse,
educational level, occupation, union status, employment type (private sector dependent,
public sector dependent, or self-employed), and work status (full-time or part-time). Each
has been shown to impact workers’ attitudes (Neil & Snizek 1987; Boyce et al. 2007;
Henson & Krasas-Rogers; Feldman et al. 1995; Tarnai et al. 1995; Mottaz 1985).
International Differences in Perceptions 19
Table 1. Grand Means and Standard Deviations of Job Values, Perceived Job Characteristics, and Perceived Employability, 2005 ISSP Work Orientations Module
Mean (standard deviation)
Job values (1=Not important at all; 5=Very important)
Job security 4.48 (.703)High income 4.01 (.773)Good opportunities for
advancement3.82 (.958)
An interesting job 4.47 (.626)A job that allows one to work
independently4.08 (.827)
A job that allows someone to decide their days and times of work
3.71 (.970)
Preference for full-time over part-time work (proportion)
0.63
Perceived job characteristics (1=Strongly disagree; 5=strongly agree)
My job is secure 3.61 (1.137)My income is high 2.69 (1.056)My opportunities for advancement
are high2.66 (1.079)
My job is interesting 3.88 (.971)I can work independently 3.87 (1.046)
Freedom to choose schedule (1=Worker has no discretion; 3=Employee has full discretion)
1.64 (.679)
Difficulty in finding a job at least as good as the current one (1=Very difficult; 5=Very easy)
2.62 (1.130)
N=13073
Analytical Strategy
The data necessitate hierarchical generalized linear modeling to account for
individuals nested within countries. Due to the nature of the response variable—a five-
category agreement scale—a nonlinear model must be fitted to account for an ordinal
International Differences in Perceptions 20
dependent variable. Ordinal regression employs a cumulative probability model to
account for ordered response categories (Raudenbush & Bryk 2002). A key assumption
of this model is proportional odds: Odds ratios for each response category are similarly
affected by the regressors (Raudenbush & Bryk 2002). The multilevel model for
proportional odds is:
log [ P ijc
1−Pijc ]=γ c+x 'ij β+z ' ij T θ j(c=1, …C−1)
in which C represents the number of ordered response categories with C – 1 thresholds γc
and z ' ij T θ j represents a vector of standardized Level-2 random effects (Hedeker 2008).
This particular model results in regression coefficients that have the correct coefficient
but the opposite sign. The cumulative probabilities conditional on Level-2 random effects
are given as:
Pr (Y ij=c|θ )=P ijc−Pij ,c−1
Between-country variability is modeled using restricted penalized quasi-
likelihood. PQL approximates maximum likelihood estimates for the Level-1 random
coefficients and the Level-2 fixed effects and variance (O’Connell et al. 2008). With
PQL, a deviance statistic that might be used to compare models is not available, and the
non-nested structure of many of the models invalidates likelihood ratio tests. Model fit
will be assessed by comparing the amount of between-country variance accounted for by
each model. While such informality is obviously less than ideal, it does provide a way to
compare the propriety of non-nested models. The general Level-2 model specification is:
βqj=γq 0+∑s=1
sq
γ qsW sj+¿uqj ¿
(1)
(2)
(3)
International Differences in Perceptions 21
Where:q is the number of level-1 coefficientsj is the number of level-2 coefficientsβqj is a set of random level-2 coefficients j across q level-1 predictors,γ00 is the level-2 overall intercept,γq0 yields the level-2 regression slopes across q level-1 predictors,W sj denotes the summed effects of coefficient set W which contains j level-2 predictors, performed for s number of level-1 coefficient (where s = 1,…, Sq), multiplied by the fixed level-1 coefficients γqs, uqj is the level-2 random effect, andu0 j is the level-2 intercept for unit j.
Analyses proceed in two stages: In the first stage, the welfare regime model is
compared to the welfare components model. Comparisons are conducted using a random-
intercept-only model as the baseline. In the second stage, the country-level model that
best fits the data will be incorporated into a model that includes individual-level
variables.
RESULTS
Descriptive Analysis
For each country and regime type, levels of willingness to accept temporary work
to avoid unemployment in presented in Table 2. Workers in liberal regimes show the
strongest agreement that they would accept temporary work rather than be unemployed,
with workers in Ireland and the United States displaying the highest level of agreement of
all countries. Overall, 72% of workers in liberal regimes agree that they would take a
temporary job to avert a spell of unemployment. Conservative regimes follow closely in
their willingness to accept temp work, with an average 69% of workers agreeing that they
would take temporary work over unemployment. Workers in social democratic regimes
show less willingness than workers in liberal or conservative regimes to accept temporary
employment. Sixty-three percent of workers in social democratic regimes agree that they
International Differences in Perceptions 22
would take temporary work to avoid unemployment. Slightly more than half of workers
in Mediterranean and post-communist regimes agree that, to avoid unemployment, they
would take a temporary job. Workers in productivist regimes evidence, by far, the lowest
levels of agreeableness to temporary work.
Table 2. Percentage Distribution of Responses to the Question “In Order to Avoid Unemployment I Would Be Willing to Accept Temporary Employment,” by Country and Welfare Regime Type
Welfare regime type Country
Strongly
disagree
Disagree
Neither
agree nor
disagree Agree
Strongly
agree
Average who agree or strongly agree by regime
type
Liberal
Australia 3.44 11.65 11.85 58.66 14.39
71.98*
Canada 5.70 13.50 17.30 48.95 14.56Great Britain 3.43 18.31 16.02 52.86 9.38
United States 3.46 9.41 8.76 54.70 23.68
Ireland 2.49 9.56 5.20 59.04 23.70
Conservative
Germany 5.08 9.31 13.12 53.88 18.62
69.29*France 13.18 14.08 15.39 40.04 17.30Switzerland 2.54 9.86 7.47 64.39 15.74
Flanders 4.03 10.99 17.80 57.86 9.32
Social democratic
Norway 4.17 11.72 15.63 55.60 12.89
62.59*Sweden 4.75 11.87 19.69 50.28 13.41Denmark 12.27 9.14 16.09 37.50 25.00
Finland 8.30 19.06 16.98 38.87 16.79Mediterranean
Spain 11.95 21.82 19.92 26.57 19.73 55.09*Portugal 8.67 11.15 16.31 44.48 19.40
Post-communist
Hungary 18.49 17.03 18.49 31.63 14.3654.97*Czech
Republic
3.23 11.56 21.26 54.93 9.01
Productivist Japan 36.19 11.26 17.96 16.89 17.69 40.32*
International Differences in Perceptions 23
South Korea 11.93 23.02 19.00 33.43 12.62
Total 7.88 12.97 15.24 47.40 16.50 63.90* Within-regime chi-squared test for country differences is significant at p<.01N = 12822
A series of Pearson’s chi-squared tests were performed within each regime type to
assess whether countries showed significant differences in their levels of willingness to
accept temporary work. Tests indicate significant differences in amenability to temporary
work within welfare regime type. Because there exists a high level of within-regime
variation, the next step is to explore the welfare state components that may contribute to
attitudes toward temporary work and unemployment.
Welfare Components Model Compared to Welfare Regime Model
Table 3 compares the welfare regimes and welfare components models (Models 2
and 3, respectively) to a baseline level-2 model containing only an intercept (Model 1).
For clarity of presentation, only the coefficients are provided. The explanatory power of
the welfare components model is poor: None of the individual welfare components
significantly affect the relative attractiveness of temporary work and unemployment. The
welfare-components model explains 13% of between-country variance, a small
percentage relative to the greater explanatory power of Model 2, which is able to account
for half of the between-country variance in levels of agreeableness toward temporary
employment compared to unemployment. Because the welfare-regime model provides a
better fit for the data, this model is preferred as the level-2 predictor.
International Differences in Perceptions 24
Table 3. Regression of Workers’ Preference for Temporary Work on Welfare Regimes and Welfare Components, HLM Ordered Logit Models
Model 1 Model 2 Model 3Intercept 2.462** 2.804** 2.463**Welfare state components
Employment protection legislation -0.140Passive labor market policies 0.227Unemployment rate -0.006Proportion temporary workers in labor force
-0.020
Welfare state regimes (Comparison: Liberal regime)
Conservative -0.144Social democratic -0.276Post-communist -0.650*Productivist -1.260**Mediterranean -0.484
Threshold 2 1.157** 1.158** 1.158**Threshold 3 1.951** 1.952** 1.952**Threshold 4 4.208** 4.210** 4.210**Variance components
Intercept (Between-country variance) 0.226** 0.112** 0.196**Variance explained 50.4% 13.3%N (Level 1) 12818 12813 12814
* p <= .05; ** p <= .01
Individual Values
In this stage of analysis, individual-level predictors and random slopes are
successively included into the model to test their predictive power. Model 4, the baseline
International Differences in Perceptions 25
model, contains level-1 predictors and level-2 random intercepts. Models 5 through 7
account for welfare regime type after a baseline of level-1 effects is established. Table 4
compares the findings of the various models.
Looking first at the baseline Model 4, several job values significantly impact
respondents’ agreement that they would prefer a temporary job over unemployment. In
particular, workers who value the extrinsic rewards of job security and high income show
less amenability to temp work than workers for whom less importance is placed on
extrinsic rewards. Intrinsic job values exhibit the opposite pattern: As the value of
Table 4. Regression of Workers’ Preference for Temporary Work on Job Values, Perceived Job Characteristics, Perceived Employability, and Welfare Regimes, HLM Ordered Logit Models
Model 4
Model 5
Model 6
Model 7
Intercept -2.874** -3.204** -3.217** -3.216**Job values
Job security 0.100** 0.100** 0.067 0.078High income 0.112** 0.110** 0.123** 0.125**Advancement prospects 0.032 0.032 0.030 0.029Interesting work -0.184** -0.183** -0.173** -0.169**Independent work -0.057* -0.056* -0.051* -0.050Flexibility 0.006 0.006 0.007 0.008Preference for full-time work (Compared to part-time or less)
-0.005 -0.006 -0.001 -0.001
Perceived job characteristicsJob security 0.054** 0.054** 0.055* 0.055**High income -0.019 -0.019 -0.021 -0.021Advancement prospects 0.047* 0.047* 0.046* 0.046*Interesting work -0.072** -0.071** -0.071** -0.071**Independent work 0.002 0.004 0.002 0.002Flexible hours -0.017 -0.017 -0.015 -0.015Ease in finding a good job 0.034* 0.036** 0.035* 0.035*
Individual characteristicsGender (Male=1) 0.342** 0.342** 0.344** 0.345**
Job characteristicsUnion member 0.106* 0.108* 0.104* 0.102*
Welfare state regimes (Comparison: Liberal regime)
Conservative 0.191 0.151 0.197Social democratic 0.269 0.120 0.296
International Differences in Perceptions 26
Post-communist 0.511 0.586 0.529Productivist 1.359** 1.436** 1.340**Mediterranean 0.350 0.683* 0.404
Slope coefficientsConservative 0.056Social democratic 0.182Post-communist -0.121Productivist -0.151Mediterranean -0.399*
Threshold 2 1.209** 1.210** 1.212** 1.212**Threshold 3 2.020** 2.021** 2.025** 2.025**Threshold 4 4.378** 4.380** 4.390** 4.392**Variance components
Intercept (Between-country variance)
0.232** 0.114** 0.122** 0.111**
Value of job security 0.037** 0.017**Variance explained 50.9% 63.4% 59.5%df (Level 1) 10858 10853 10853 10848 Controls: Age, marital/cohabitation status, presence of a working spouse/partner, educational attainment, sector (private or public), dependent employment status (dependent or self-employed), full-time/part-time status, occupation* p <= .05; ** p <= .01interesting and independent work rewards rises, workers are more likely to agree that
they would accept temporary employment as a means of avoiding unemployment.
Willingness to accept temporary employment also appears to be a function of the
characteristics of one’s current job. Workers who feel their job security and advancement
prospects are high are less likely than workers who perceive that their jobs possess fewer
of such extrinsic rewards to prefer temporary work over unemployment. The current level
of intrinsic rewards matters only insofar as workers higher ratings of interesting work are
associated with a higher likelihood of agreeing that temporary work would be preferable
to unemployment.
A high evaluation of one’s external labor market prospects decreases agreement
that temporary work is a preferable alternative to unemployment. As the perceived ease
with which a comparable job could be found increases, willingness to accept temporary
employment over unemployment decreases. In sum, intrinsic values and rewards,
International Differences in Perceptions 27
particularly the value and presence of independent work in one’s job, tend to predispose
workers toward temporary employment. Conversely, the presence and valuation of
extrinsic rewards show the opposite effect, with higher rewards and values associated
with a lower inclination toward temporary work rather than unemployment.
Models 5 through 7 add various country level differences to the level-1 model.
Model 5 tests the significance of welfare regime in the presence of individual-level
indicators. Even after accounting for workers’ values, subjective job characteristics, and
objective job characteristics, the welfare regimes model still explains half of the between-
country variance. Although there is a great deal of variation within each country, there
remains significant variation between countries.
Model 6 explores the role of random slopes in between-country variation.
Allowing variation in the slopes of job security valuation increases the proportion of
explained between-country variance to 63.4%. It was hypothesized that the value placed
on job security may function differently in various countries. This does indeed appear to
be the case. The influence of job security valuation on the willingness to accept
temporary work varies between countries, although the model does not allow statements
on the nature of the difference. Interestingly, the coefficient for job security valuation is
no longer significant after accounting for random slopes, but this may be attributable to
the structure of the model: At the individual level, regressions include 10,853 cases.
Calculating random slopes for each country converts job security valuation into a level-2
variable, dramatically reducing the number of cases to 13. Thirteen cases may simply be
too few to determine significance if the effect is small.
International Differences in Perceptions 28
Model 7 attempts to establish if this difference can be accounted for by welfare
regimes. The relationship between job security valuation and temporary work inclination
is significantly stronger in Mediterranean regimes than in liberal regimes, but none of the
other regime types differ from liberal regimes in the explanatory power of random slopes.
Welfare regimes do a poor job of explaining why the influence of job security valuation
differs between countries. Thus, Model 7, which accounts for 59.5% of between-country
variance, is rejected in favor of Model 6, which is able to account for a greater proportion
of between-country variance.
According to Model 6, the post-communist regime is no longer significantly
different from the liberal regime, as it was in Model 2, although the Mediterranean
regime now displays significance. The productivist regime still exhibits the greatest
difference from the liberal regime, with workers who are considerably more averse to
temporary work even after accounting for work values and job characteristics.
DISCUSSION
Individual-level hypotheses were supported: Extrinsic work rewards and the value
placed on them significantly related to workers’ relative preference for temporary work
over unemployment. Workers who value either job security or high income show a lower
willingness to accept temporary work as a means of avoiding unemployment than
workers for whom extrinsic rewards are less present or important. Workers who believe
they could easily find a good job also display an aversion to temporary work. The value
and presence of interesting work exhibited the opposite effect: Workers who valued or
possessed interesting work showed a higher preference for temporary work to escape
unemployment.
International Differences in Perceptions 29
Workers who value intrinsic job rewards may be more inclined to accept
temporary work rather than be unemployed because work itself is a perceived as valuable
pursuit. This finding may also be interpreted to mean that workers with interesting jobs
have access to temporary jobs of good quality. Several studies have noted that workers in
highly skilled jobs may prefer to operate in the external labor market. Although contract
jobs carry the risk of finite and fixed duration, such jobs may provide opportunities for
skill enhancement, self-determination, and income leveraging (Kunda, Barley, & Evans
2002; O’Mahoney & Bechky, 2006).
The finding that job security, both its valuation and its presence, are associated
with a decreased willingness to accept temporary work to avoid unemployment merits
further investigation. Given that many workers place a high premium on job security,
their avoidance of temporary work could be interpreted to mean either that (1) workers
are inflexible when it comes to time-limited employment or (2) temporary jobs are
perceived as comparatively bad in relation to permanent employment. In reality, a
combination of both factors is likely.
Perceived job security and employability both reduce the likelihood that workers
will express a preference for temporary work over unemployment. This finding in
particular carries the most serious implications for workers. If temporary work is
available but viewed as undesirable, workers may choose unemployment instead, thereby
leading to wasted human capital and creating an unnecessary burden on labor market
protection programs. Few workers anticipate being unemployed for months, let alone
years. If people systematically overestimate their chances in the external labor market
and underestimate the length of time they will be unemployed, workers may choose an
International Differences in Perceptions 30
employment outcome that carries far worse repercussions—skill degradation and income
loss being the most salient possibilities—than they originally anticipated.
An obvious limitation of the present study is the lack of information about
whether respondents currently work in a temporary job. The ISSP does not assess
temporary or permanent status of survey respondents. Measures of perceived job
characteristics, particularly measures of perceived job security, should partly compensate
for this deficiency, as temporary workers tend to feel less secure in their jobs than
permanent employees (DeCuyper 2008; De Witte & Näswall 2003).
Temporary workers’ representation in the ISSP should be higher in countries
containing a high proportion of temporary workers (South Korea, Spain, and Portugal). If
workers currently employed in temporary contracts were more amenable to temporary
work, then countries with a high proportion of temporary workers should show greater
inclination toward temporary work. Instead, the opposite is found. Workers in
Mediterranean and productivist regimes show a stronger aversion to temporary work than
workers in liberal regimes. Although temporary work is prevalent in other regimes—the
social democratic and conservative regimes in particular—the Mediterranean and
productivist regimes contain two-tier labor markets divided along the lines of permanent
versus temporary status. This feature is not present to the same degree in other welfare
regimes. In addition to the relative quality of unemployment compared to temporary
employment, the relative quality of permanent employment also seems to figure heavily
in workers’ evaluations of temporary jobs. In sum, it seems that career researchers would
do well to take a cue from the workers themselves and consider the role of relativity in
decision making.
International Differences in Perceptions 31
International Differences in Perceptions 32
References
Arts, Wil and John Gelissen. 2002. “Three Worlds of Welfare Capitalism or More? A
State-of-the-Art Report.” Journal of European Social Policy 12(2): 137–158.
Auer, Peter and Sandrine Cazes. 2003b. “The Resilience of the Long-term Employment
Relationship.” Pp. 22–58 in Employment Stability in an Age of Flexibility: Evidence
from Industrialized Countries, edited by P. Auer and S. Cazes. Geneva, Switzerland:
International Labour Office.
Beck, Ulrich. 2000. The Brave New World of Work. Cambridge, UK: Polity Press.
Bockerman, Petri . 2004. "Perception of Job Instability in Europe." Social Indicators
Research 67(3):283–314.
Booth, Alison L., Juan J. Dolado, and Jeff Frank. 2002a. “Symposium on Temporary
Work: Introduction.” Economic Journal 112(480): F181–F188.
Booth, Alison L., Marco Francesconi, and Jeff Frank. 2002b. “Temporary Jobs: Stepping
Stones or Dead Ends?” Economic Journal 112(480): F189–F213.
Boyce, Anthony S., Ann Marie Ryan, Anna L. Imus, and Frederick P. Morgeson. 2007.
“Temporary Worker, Permanent Loser? A Model of the Stigmatization of Temporary
Workers.” Journal of Management 33(1): 5–29.
Cazes, Sandrine and Alena Nesporova. 2003. “Employment Protection Legislation (EPL)
and Its Effects on Labour Market Performance.” Paper presented at the High-Level
Tripartite Conference on Social Dialogue, Valletta, Malta.
De Cuyper, Nele and Hans De Witte. 2005. “Job Insecurity: Mediator or Moderator of
the Relationship Between Type of Contract and Various Outcomes?” SA Journal of
Industrial Psychology 31(4): 79–86.
International Differences in Perceptions 33
De Cuyper, Nele, Claudia Bernhard-Oettel, Erik Berntson, Hans De Witte, and Barbara
Alarco. 2008. “Employability and Employees’ Well-being: Mediation by Job
Insecurity.” Applied Psychology: An International Review 57(3): 488–509.
De Witte, Hans, and Katharina Naswall . 2003. "'Objective' Vs 'Subjective' Job
Insecurity: Consequences of Temporary Work for Job Satisfaction and Organizational
Commitment in Four European Countries." Economic and Industrial Democracy 24
(2): 149–188.
Dolado, Juan J., Carlos García-Serrano and Juan F. Jimeno. 2002. “Drawing Lessons
from the Boom of Temporary Jobs in Spain.” Economic Journal 112(480): F270–
F295.
Edwards, Rosalind and Judith Glover. 2001. Risk and Citizenship: Key Issues in Welfare.
London, UK: Routledge.
European Commission. 2006. Employment in Europe 2006. Luxembourg: Office for
Official Publications of the European Communities.
Esping-Andersen, Gøsta. 1990. The Three Worlds of Welfare Capitalism. Princeton, NJ:
Princeton University Press.
Feldman, Daniel C., Doerpinghaus, Helen I., and William H. Turnley. 1995. “Employee
Reactions to Temporary Jobs.”Journal of Managerial Issues 7(2): 127–141.
Feinberg, Robert M. 1977. “Risk Aversion, Risk, and the Duration of Unemployment.”
Review of Economics and Statistics 59(3): 264–271.
Fugate, Mel, Angelo J. Kinicki, and Blake E. Ashforth. 2004. “Employability: A Psycho-
social Construct, Its Dimensions, and Applications.” Journal of Vocational Behavior
65(1): 14–38.
International Differences in Perceptions 34
Gallie, Duncan and Susanne Alm. 2000. “Unemployment, Gender, and Attitudes to
Work.” Pp. 109–133 in Welfare Regimes and the Experience of Unemployment in
Europe, edited by D. Gallie and S. Paugam. Oxford University Press.
Gallie, Duncan and Serge Paugam. 2000. “The Experience of Unemployment in Europe.”
Pp. 1–24 in Welfare Regimes and the Experience of Unemployment in Europe, edited
by D. Gallie and S. Paugam. Oxford University Press.
Gangl, Markus. 2006. “Scar Effects of Unemployment: An Assessment of Institutional
Complementarities.” American Sociological Review 71(6): 986–1013.
Gangl, Markus. 2008. “Unemployment and Worker Career Prospects: A Cross-national
Perspective.” Pp. 169–194 in Flexibility and Employment Security in Europe: Labour
Markets in Transition, edited by R. Muffels. Cheltenham, UK: Edward Elgar.
Gash, Vanessa. 2008. “Bridge or Trap? Temporary Workers’ Transitions to
Unemployment and the Standard Employment Contract.” European Sociological
Review 24(5): 651–668.
Green, Francis. 2006. Demanding Work: The Paradox of Job Quality in the Affluent
Economy. Princeton, NJ: Princeton University Press.
Hedeker, Donald. 2008. “Multilevel Models for Ordinal and Nominal Variables.” Pp.
237–274 in Handbook of Multilevel Analysis, edited by J. de Leeuw and E. Meijer.
New York: Springer Publishers.
Henson, Kevin D. and Jackie Krasas-Rogers. 2001. “’Why Marcia You’ve Changed!’
Male Clerical Temporary Workers Doing Masculinity in a Feminized Occupation.”
Gender and Society 15(2): 218–238.
International Differences in Perceptions 35
Holliday, Ian. 2000. “Productivist Welfare Capitalism: Social Policy in East Asia.”
Political Studies 48(4): 706–723.
Hult, Carl. 2005. “Organizational Commitment and Person-environment Fit in Six
Western countries.” Organization Studies 26(2): 249–270.
Judge, Timothy A. and Robert D. Bretz, Jr. 1992. “Effects of Work Values on Job Choice
Decisions.” Journal of Applied Psychology 77(3): 261–271.
Kalleberg, Arne L. 1977. “Work Values and Job Rewards: A Theory of Job Satisfaction.”
American Sociological Review 42(1): 124–143.
Kalleberg, Arne L., Barbara F. Reskin, and Ken Hudson. 2000. “Bad Jobs in America:
Standard and Nonstandard Employment Relations and Job Quality in the United
States.” American Sociological Review 65(2): 256–278.
Kalleberg, Arne L. and Jeremy Reynolds. 2003. “Work Attitudes and Nonstandard Work
Arrangements in the United States, Japan, and Europe.” Pp. 423–476 in Nonstandard
Work in Developed Economies: Causes and Consequences, edited by S. Houseman
and M. Osawa. Kalamazoo, MI: W.E. Upjohn Institute for Employment Research.
Kunda, Gideon, Stephen R. Barley, and James Evans. 2002. “Why Do Contractors
Contract? The Experience of Highly Skilled Technical Professionals in a Contingent
Labor Market.” Industrial and Labor Relations Review 55(2): 234–261.
Madsen, Per K. 2008. “The Danish Road to ‘Flexicurity’: Where Are We Compared to
Others? And How Did We Get There?” Pp. 341–362 in Flexibility and Employment
Security in Europe: Labour Markets in Transition, edited by R. Muffels. Cheltenham,
UK: Edward Elgar.
International Differences in Perceptions 36
Mottaz, Clifford J. 1985. “The Relative Importance of Intrinsic and Extrinsic Rewards as
Determinants of Work Satisfaction.” Sociological Quarterly 26(3): 365–385.
Neil, Cecily C. and William E. Snizek. 1987. “Work Values, Job Characteristics, and
Gender.” Sociological Perspectives 30(3): 245–265.
Nickell, Stephen. 1997. “Unemployment and Labor Market Rigidities: Europe Versus
North America.” Journal of Economic Perspectives 11(3): 55–74.
O’Connell, Ann A. Jessica Goldstein, H. Jane Rogers and C. Y. Joanne Peng. 2008.
“Multilevel Logistic Models for Dichotomous and Ordinal Data.” Pp. 199–242 in
Multilevel Modeling of Educational Data, edited by A. A. O’Connell and D. B.
McCoach. Charlotte, NC: Information Age Publishing.
O'Mahony, Siobhan and Beth A. Bechky. 2006. "Stretchwork: Managing the Career
Progression Paradox in External Labor Markets". Academy of Management Journal
49(5): 918–941.
Organisation for Economic Co-operation and Development. N.d. “Calculation of
Summary Indicators of EPL Strictness.” Retrieved March 24, 2009
(www.oecd.org/dataoecd/7/22/38940931.pdf).
Passet, Olivier. 2003. “Stability and Change: Japan’s Employment System Under
Pressure.” Pp. 159–214 in Employment Stability in an Age of Flexibility: Evidence
from Industrialized Countries, edited by P. Auer and S. Cazes. Geneva, Switzerland:
International Labour Office.
Polavieja, Javier G. 2003. “Temporary Contracts and Labour Market Segmentation in
Spain: An Employment-rent Approach.” European Sociological Review 19(5): 501–
517.
International Differences in Perceptions 37
Raudenbush, Stephen W. and Anthony S. Bryk. 2002. Hierarchical Linear Models:
Applications and Data Analysis Methods. Thousand Oaks, CA: Sage Publications.
Tarnai, Christian, Holger Grimm, Dirk John, and Rainer Watermann. 1995. “Work
Values in European Comparison: School Education and Work Orientation in Nine
Countries.” Tertium Comparationis: Journal für Internationale Bildungsforschung
1(2): 139–163.
Vardaman, James M., David G. Allen, Robert W. Renn, and Karen R. Moffitt. 2008.
“Should I Stay or Should I Go? The Role of Risk in Employee Turnover Decisions.”
Human Relations 61(11): 1531–1563.