University of Southern California Dissertations and Theses

Essays on the economics of subjective well-being in transition countries

(USC Thesis Other)

Essays on the economics of subjective well-being in transition countries

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content ESSAYS ON THE ECONOMICS OF SUBJECTIVE WELL-BEING
IN TRANSITION COUNTRIES
by
Laura Angelescu
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulllment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(ECONOMICS)
August 2010
Copyright 2010 Laura Angelescu
Acknowledgments
My gratitude goes to my advisor, Richard A. Easterlin, who introduced me to the eld of subjective
well-being and generously guided my steps over the years. I am also thankful to the members
of my guidance committee and my dissertation committee, Juan Carrillo, John McArdle, John
Strauss, and Guofu Tan, for their valuable advice. I also received helpful advice from participants
at various conferences.
My ocemates throughout the years, Anke Plagnol, Onnicha Sawangfa, Olga Shemiakina,
Malgorzata Switek, and Jacqueline Zweig, have shared helpful comments and have been great
friends.
I am grateful to my parents, Valeria and Manuel Angelescu, and to my husband, John McVey,
who have been there for me every step of the way, providing support and increasing my own
happiness.
ii
AB
Table of Contents
Acknowledgments aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa ii
List Of Tables aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa iv
List Of Figures aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa vi
Abstract aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa vii
Chapter 1: aSubjective Well-Being and the Transition: A Brief Overview aaaaaaaaaaaaaaaaaa 1
Chapter 2: aTransition at Work: A Comparison of Job Satisfaction in Eastern and Western
aaaaaaaaaaa Europe aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa 7
aa 2.1 aIntroduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
aa 2.2 aData and Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
aaa 2.2.1 aData . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
aaa 2.2.2 aMethodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
aa 2.3 aFindings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
aaa 2.3.1 aJob Satisfaction in the 1990s . . . . . . . . . . . . . . . . . . . . . . . . 27
aaa 2.3.2 aJob Satisfaction in the New Millennium . . . . . . . . . . . . . . . . . 38
aa 2.4 aSummary and Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
Chapter 3: aLife Satisfaction and the Economic Transition in Poland 48
aa 3.1 aIntroduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
aaa 3.1.1 aThe transition in Poland . . . . . . . . . . . . . . . . . . . . . . . . . . 50
aa 3.2 aData and methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
aa 3.3 aFindings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
aa 3.4 aSummary and Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
Chapter 4: aSummary aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa 80
Bibliography aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa 91
Appendix A
aa Chapter 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
aa A.1 aWorld Values Survey (WVS) and International Social Survey
aaaaaaaaaaaaaaaa Programme (ISSP) \Work Orientations" module . . . . . . . . . . . . . . . 98
Appendix B
aa Chapter 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
aa B.1 aWorld Values Survey (WVS) and Eurobarometer (EB) data . . . . . . . . . 103
iii
List Of Tables
2.1 Descriptive Statistics for Key Variables in the World Values Survey, wave 2 and
wave 4, transition countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.2 Descriptive Statistics for Key Variables in the World Values Survey, wave 2 and
wave 4, non-transition countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.3 Descriptive Statistics for Key Variables in the International Social Survey Pro-
gramme, 1997 and 2005, transition countries . . . . . . . . . . . . . . . . . . . . . . 18
2.4 Descriptive Statistics for Key Variables in the International Social Survey Pro-
gramme, 1997 and 2005, non-transition countries . . . . . . . . . . . . . . . . . . . 19
2.5 Mean job satisfaction by region and gender, 1990 and 1999 . . . . . . . . . . . . . 28
2.6 Changes in job satisfaction by country, 1990-1999 . . . . . . . . . . . . . . . . . . . 29
2.7 Ordinary least squares regressions of job satisfaction in Eastern and Western Eu-
rope, 1990-1999 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
2.8 Ordinary least squares regressions of job satisfaction with interactions between the
transition country dummy and specied independent variables, 1990-1999 . . . . . 32
2.9 Country xed-eects regressions of job satisfaction with interactions between time
and specied independent variables, 1990-1999 . . . . . . . . . . . . . . . . . . . . 33
2.10 Ordinary least squares regressions of job satisfaction on region, 1990 . . . . . . . . 34
2.11 Ordinary least squares regressions of job satisfaction on region, 1999 . . . . . . . . 35
2.12 Country xed-eects regressions of job satisfaction on time, 1990-1999, transition
countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
2.13 Job satisfaction means and standard errors by year of birth cohort, 1990 and 1999 38
2.14 Mean job satisfaction by region and gender, 1997 and 2005 . . . . . . . . . . . . . 39
2.15 Changes in job satisfaction by country, 1997-2005 . . . . . . . . . . . . . . . . . . . 40
iv
2.16 Ordinary least squares regressions of job satisfaction with interactions between the
transition country dummy and specied independent variables, 1997-2005 . . . . . 41
2.17 Ordinary least squares regressions of job satisfaction on region, 1997 . . . . . . . . 42
2.18 Ordinary least squares regressions of job satisfaction on region, 2005 . . . . . . . . 43
2.19 Country xed-eects regressions of job satisfaction on time, 1997-2005, transition
countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
2.20 Job satisfaction means and standard errors by year of birth cohort, 1997 and 2005 45
3.1 Mean life satisfaction by dataset and survey date, on original scale and on 1-10
scale, 1989-2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
3.2 Selected macroeconomic indicators for Poland, 1989-2009 . . . . . . . . . . . . . . 62
3.3 WVS sample description by wave . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
3.4 Life satisfaction means and standard errors by birth cohort and wave . . . . . . . . 65
3.5 EB sample description by year . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
3.6 Ordinary least square regressions of life satisfaction on selected macroeconomic
indicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
3.7 Ordinary least squares and year of birth xed-eects regressions of life satisfaction
{ WVS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
3.8 Ordinary least square regressions of life satisfaction on cohort { WVS . . . . . . . 73
3.9 The impact of employment status on life satisfaction by birth cohort and wave {
WVS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
3.10 Ordinary least squares regressions of life satisfaction . . . . . . . . . . . . . . . . . 76
3.11 Year of birth xed-eects regressions of life satisfaction . . . . . . . . . . . . . . . . 77
v
List Of Figures
2.1 GDP per capita index (1989=100) and absolute GDP per capita in transition and
non-transition countries for the WVS and ISSP groups of countries, respectively,
1989-2006 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.2 The unemployment rate in transition and non-transition countries for the WVS
and ISSP groups of countries, respectively, 1989-2006 . . . . . . . . . . . . . . . . . 21
3.1 Life satisfaction in Poland between 1989 and 2009 . . . . . . . . . . . . . . . . . . 68
3.2 GDP per capita, Gini coecient, unemployment and in
ation rates in Poland be-
tween 1989 and 2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
vi
Abstract
The two essays in this dissertation share in common an attempt to study the transition process
in terms of its eect on people's subjective well-being, in addition to the objective conditions
traditionally emphasized in economics. The rst essay focuses on the particular area of job
satisfaction, while the second analyzes overall life satisfaction.
As the transition progresses is there a convergence in terms of job satisfaction between Eastern
and Western Europe? I analyze the level of satisfaction with work and its determinants in each
of the two regions for the decade and a half following the fall of communism, using data from the
World Values Survey and the International Social Survey Programme \Work Orientations" mod-
ule. Job satisfaction in transition countries is signicantly lower than in the West. These countries
experience a signicant decrease in satisfaction with work between 1990 and 1999, followed by a
signicant increase by 2005. In non-transition countries, there is no signicant change throughout
this interval. As a result, the job satisfaction gap between East and West rst expands and then
shrinks. This gap is mainly the result of dierences in macroeconomic conditions between the two
regions. Not everyone in Eastern Europe is aected the same way by the transition, young and
more educated, more skilled individuals being among the winners of the process.
Since 1989 Poland has been considered a leader in economic reform, but did the process of
transition from a planned economy to a free market model make its people happier? Using data
from the World Values Survey and the Eurobarometer, I nd a collapse followed by recovery in
life satisfaction. Despite the fact that GDP per capita quickly recovers to pre-transition levels,
higher unemployment and involuntary early retirement take their toll on the happiness of the
vii
Polish people, which is one of the reasons why the recovery in terms of subjective well-being takes
longer than the economic recovery. The eventual recovery of life satisfaction is also made possible
by birth cohort replacement { new generations appear to be better adjusted to the new society
and better equipped to deal with the negative side eects of transition.
viii
Chapter 1
Subjective Well-Being and the Transition: A Brief
Overview
The transition from communism to capitalism in Central and Eastern Europe and the former
Soviet Union is one of the largest natural experiments in economic reform. At the end of the
1980s and beginning of the 1990s, about 25 planned or communist economies undertook a major
transformation from centrally planned to market-based economies, and from single-party regimes
to liberal democracies (Campos and Horv ath, 2006; Delikta s and Balcilar, 2005). This also marked
the end of the Cold War. The revolutions of 1989 were made possible by the lifting of the Soviet
imperial constraint. While resistance and dissent existed in Central and Eastern Europe (CEE)
before 1989, for example in the form of the Solidarity movement in Poland, until Gorbachev came
to power in the Soviet Union, democracy was hardly considered a possibility. His decision not to
back the communist regimes in CEE with military force created the right context for the Soviet
bloc to implode rapidly and peacefully. A unique feature of this transition is the fact that it did
not propose a new model of society, but rather it saw the imitation of existing Western models as
the quickest path to democracy and prosperity (Rupnik, 2000).
At the onset of the transition, the countries of CEE had a lot of similarities { their income per
person was in the middle income range from a global perspective and relatively evenly distributed,
heavy industry and large monopolistic rms dominated the industry, their international trade
1
was shaped by state agreements rather than market considerations (Fischer and Gelb, 1991).
At the same time, however, there were important dierences. Countries like Czechoslovakia,
East Germany, the Soviet Union, Bulgaria, and Romania started o with relatively centralized
economies. By contrast, Yugoslavia had long been the most decentralized of the socialist countries,
while Hungary and Poland had also made important progress pre-1989 toward market-based
distribution of products and greater autonomy for rms. In fact, the change in regime in Poland
and Hungary had been almost a decade in the making and took the form of repeated confrontation
between the governing elites and social movements in Poland or of negotiations between rival
elites in Hungary. At the other extreme, Bulgaria and Romania experienced a very short and
comparatively violent collapse of a much more authoritarian form of communism (Oe, 1997;
Fischer et al., 1996).
Shifts from one type of society to another also occurred in other parts of the world and at
other times { West Germany, Italy, and Japan in the 1940s; Spain and Portugal in the 1970s; some
Latin American countries in the 1970s and 1980s; South Korea and Taiwan in the 1980s; China
since the late 1970s; Vietnam since the late 1980s (Balcerowicz, 1994). However, a number of
features make the transition in CEE unique. Its scope is exceptionally large, involving changes in
both the political and economic systems. The fairly quick mass democratization made it possible
to implement comprehensive economic reforms under a democratic regime. The transition in CEE
was also exceptional through its lack of violence, with the only case of violent transition taking
place in Romania.
From the beginning of the transition it was quite clear where reforms were needed. Fischer et al.
(1996, p.46) sum up these six areas of action: \macroeconomic stabilization; price liberalization;
trade liberalization and current account convertibility; enterprise reform (especially privatization);
the creation of a social safety net; and the development of the institutional and legal framework
for a market economy (including the creation of a market-based nancial system)". When it
came to the speed and sequence of these various reforms, however, much less agreement existed.
2
The debate over the speed of transition mostly took the form of \big bang" or \shock therapy"
approach versus gradualism. The former advocates sizing the moment of political opportunity
and implementing major changes as quickly as possible, while the latter is considered a way to
minimize disruption, output and job losses (Havrylyshyn, 2006). Poland is the typical example
of a big bang approach to the transition, and Czechoslovakia, Bulgaria, and Romania followed its
lead in 1991. Hungary, on the other hand, adopted a gradualist program in January 1991 (Bruno,
1992).
The impact that the various reforms had on the economies of CEE has been extensively studied
and a good summary of the challenges of the rst decade of transition is provided in Campos and
Coricelli (2002, table 13). The most notable fact remains the collapse of output, with only two out
of 25 countries surpassing their 1989 real GDP level by 1999. Important changes also occurred
in the labor market, such as a signicant increase in the
ows out of the labor force and very
rapid increases in unemployment, although from a very low starting level. Privatization was the
most radical transformation policy and also the most challenging one. By 2000 the private sector
share in GDP was of at least 60 per cent in all of the transition countries except Slovenia. The
eect of privatization on economic growth, however, is hard to determine, given that some of the
fastest growing economies { Poland, Slovenia { have been among the slowest to privatize (Svejnar,
2002). One undeniable characteristic of the transition process has been the unexpectedly high
social costs. While the increases in unemployment and income inequality were not a surprise, the
rise in the mortality rate and the decline in school enrollment rates were not expected.
When asking whether the transition has been a success or a failure, Kornai (2006) argues
that the answer depends on the perspective. In the context of world history, the transition has
been a success story because it established a capitalist economic system within a historically
brief time frame, and in a peaceful manner. From the perspective of everyday life, however, the
transition has been a disappointment for the many people faced with deep economic troubles. The
experience of the latter has been made worse by the fact that the unusually good performance of
3
Western economies in the 1990s raised the bar for perceptions of economic success, by the initial
underestimation of the costs associated with the transition, by the many questionable choices
made by policymakers (Svejnar, 2002). It is also important to point out that the performance of
the various transition countries has been far from homogeneous. The Central European countries
{ the Czech Republic, Hungary, Slovakia, Slovenia, and Poland { have made the most progress
in transitioning from planned to market economies. The Baltic states { Latvia, Lithuania, and
Estonia { are also among the success stories. At the other extreme are the rest of the former
Soviet republics, which have suered the most from the transformation process (G orniak, 2000).
In addition to the extensive literature studying the transition process by looking at economic,
social, or political indicators, there are also a number of empirical studies of life satisfaction
in transition countries. Some papers look at a specic country over varying time periods. Thus,
Namazie and Sanfey (2001) look at Kyrgyzstan and nd a positive association between satisfaction
and income, and a lower level of satisfaction among older people, the unemployed, and those who
are divorced. Frijters et al. (2004a,b) nd the same positive association with income in East
Germany between 1991 and 2001, while Easterlin and Plagnol (2008) show that by the late 1990s
life satisfaction in the former GDR had recovered to about its 1990 level. A number of studies
look at life satisfaction in Russia { some (Frijters et al., 2006; Senik, 2004) nd that changes in
income explain an important part of changes in life satisfaction, while Graham and Pettinato
(2002) focus on relative income dierences and nd that they are particularly important for the
life satisfaction of those in the middle of the income distribution. Others (Graham et al., 2004;
Saris, 2001; Veenhoven, 2001) report declines in life satisfaction in Russia during the transition,
and generally nd high levels of unhappiness. In Hungary life satisfaction decreases between the
early and the late 1990s and entrepreneurs appear to be among the winners of the transition
(Lelkes, 2006a,b).
4
When comparing transition and non-transition, developed countries, the latter consistently
display higher levels of life satisfaction. The impact that demographic and socio-economic vari-
ables have on subjective well-being however, is strikingly similar in the two groups of countries
(Blanch
ower, 2001; Green and Tsitsianis, 2005; Hayo, 2007; Sanfey and Teksoz, 2007; Tsai, 2009).
Dierences in life satisfaction across Eastern European countries are typically explained by dif-
ferences in macroeconomic indicators such as GDP or unemployment, and not by individual-level
explanatory variables (Hayo, 2007). As far as the life satisfaction trend in transition countries
is regarded, Sanfey and Teksoz (2007) nd a decrease until the mid-1990s, followed by improve-
ments. The decrease during the rst years of transition is consistent with the ndings of Hayo
and Seifert (2003) regarding subjective economic well-being { in 7 out of 10 transition countries
there is a decline between 1991 and 1995 in the proportion of people saying that their economic
situation is satisfactory or very satisfactory. Similar to Sanfey and Teksoz (2007), Easterlin (2009)
nds a collapse followed by recovery in life satisfaction, but he points out that the recovery in
happiness falls short of the recovery in GDP.
The two essays in this dissertation share in common an attempt to study the transition pro-
cess in terms of its eect on people's subjective well-being, in addition to the objective conditions
traditionally emphasized in economics. The rst essay deals with the particular area of job satis-
faction, while the second focuses on overall life satisfaction The main objectives are to determine
the trend of subjective well-being starting very soon after the collapse of communism, to see what
the determinants of this trend are, and who won and who lost during the transition.
As the transition progresses is there a convergence in terms of job satisfaction and its determi-
nants between Eastern and Western Europe? I nd that throughout the rst decade of transition,
job satisfaction in the post-communist economies is signicantly lower than in the West (Chapter
2). In transition countries, there is a signicant decrease in satisfaction with work between 1990
and 1999, followed by a signicant increase by 2005. In non-transition countries, there is no sig-
nicant change throughout this interval. As a result, the job satisfaction gap between East and
5
West rst expands and then shrinks. This gap is mainly the result of dierences in macroeconomic
conditions between the two regions. Not everyone in Eastern Europe is aected the same way by
the transition, young and more educated, more skilled individuals being among the winners of
the process.
The second essay focuses on overall life satisfaction during the transition in Poland. Just like
for job satisfaction, I nd a collapse, followed by recovery in life satisfaction, which is consistent
with what the studies listed before nd. Despite the fact that GDP per capita quickly recovers to
pre-transition levels, higher unemployment and early retirement take their toll on the happiness
of the Polish people, which is one of the reasons why the recovery in terms of subjective well-being
lags behind the economic recovery. An important role in the eventual recovery of life satisfaction
is played by birth cohort replacement { new generations appear to be better adjusted to the new
society and better equipped to deal with the negative side eects of transition. As they replace
older people in the population, life satisfaction is more likely to keep up with the pace of economic
improvements. This is consistent with the higher degree of adaptation among younger generations
that Alesina and Fuchs-Sch undeln (2007) nd in East Germany.
6
Chapter 2
Transition at Work: A Comparison of Job Satisfaction in
Eastern and Western Europe
2.1 Introduction
This chapter compares the level, trend, and determinants of job satisfaction between transition
and non-transition countries in Europe. Transition countries are those countries in Central and
Eastern Europe (CEE) that changed from a planned economy to a Western free market model
at the end of the 1980s or the beginning of the 1990s. Non-transition countries are those of
Western Europe. Due to geographic proximity, CEE countries most likely compare themselves
with Western European countries, as an example of the capitalist model
1
. Presumably, half a
century of communism has left deep scars in terms of job satisfaction so a signicant dierence
between the two regions right after the fall of communism would not be surprising. However, as the
transition to a market economy in CEE progresses, one might expect convergence between East
and West. Looking at the two groups of countries during the rst decade and a half of transition, I
nd that a full convergence in terms of job satisfaction has not yet been achieved. In fact, Eastern
Europe is even further behind the West in 1999 compared with 1990 and only between 1997 and
1
Throughout the paper, \Eastern Europe" and \transition countries" will be used interchangeably, and so will
\Western Europe" and \non-transition countries".
7
2005 there is some convergence. In transition countries, more educated and skilled people are
among the winners of the transformation. People under 30 years old also improve their relative
satisfaction compared with other age groups, while the relative satisfaction with work of people
between 40 and 49 years old gets worse.
Job satisfaction can be dened as \the individual's response to a specic question designed to
elicit his feelings about the job as a whole". This denition assumes that \the individual has some
feelings about the entire job and can respond rationally to a well-constructed question about it"
(Hamermesh, 1977, p. 54).
The concept of job satisfaction has been developed within sociology and industrial psychology
(Blauner, 1964; Herzberg et al., 1957), as well as within the eld of organizational behavior
(Spector, 1997). Having long studied satisfaction with work, psychologists have looked into the
psychometric properties of single-item job satisfaction measures such as the ones used here and
found them to have high reliability, signicant validity, and substantial predictability (Dolbier
et al., 2005; Wanous et al., 1997). It has also been found that there is a positive correlation with
occupational status and a weak positive correlation with earnings. The relationship between job
satisfaction and age is a little less clear { some studies nd a linear, increasing, relationship with
age (Brush et al., 1987), while others nd a curvilinear, U-shaped relationship (Zeitz, 1990). One
explanation for these seemingly contradictory results might have to do with gender { Clark et al.
(1996) nd a linear relationship for women, while for men the curvilinear relationship is only
apparent if the age distribution starts as early as the late teens. In addition to age, a birth cohort
eect has been found, with older cohorts typically more satised at work than newer ones. J urges
(2003) nds that for West Germany there is a succession of increasingly dissatised cohorts of
workers, with intrinsically more satised cohorts leaving the labor market and being replaced by
less satised cohorts. Various hypothesis have been advanced in order to explain this cohort eect.
In general, older workers report a closer match between what they have and what they want in
terms of job conditions and also a higher salary. It is dicult though to assess to what extent
8
this closer match is due to older cohorts having more skill and better jobs, and to what extent
it is their lower, more realistic expectations that explain their higher job satisfaction (White and
Spector, 1987; Wright and Hamilton, 1978).
Starting with Hamermesh (1977, 2001), economists have been increasingly interested in study-
ing satisfaction with work. One relationship that has been fairly extensively studied is that
between job satisfaction and quit rates (Akerlof et al., 1988; Clark, 1997; Freeman, 1978), with
the nding that an increase in satisfaction with work lowers the probability that an employee
will subsequently quit. Moreover, low satisfaction typically leads to higher absenteeism (Vroom,
1964) and labor turnover rates and is thus costly to society. It is not surprising then that high
organizational performance is related to high satisfaction. Psychologists have pointed out the
link between measures of job satisfaction or employee engagement, on the one hand, and rm
performance re
ected by measures such as protability, productivity, turnover or absenteeism, on
the other (Harter et al., 2002; Judge et al., 2001). They have also found that job satisfaction
aects mental health or physical ailments (Locke, 1976). Because a direct and positive relation-
ship exists in the service industries between employee satisfaction and customer satisfaction, job
satisfaction becomes even more important in the context of the shift from manufacturing towards
these industries (Rogers et al., 1994). Furthermore, performance of professional and knowledge
workers is often dicult to measure and this makes indirect measures, such as employee satisfac-
tion, more useful (Sousa-Poza and Sousa-Poza, 2000b). At the same time, job satisfaction has
been linked to a number of specic aspects of the workplace, such as mode of supervision, physical
work conditions, and so forth. Because these variables are not generally measured in large data
les, this makes satisfaction a potential proxy for such unobserved objective factors (Freeman,
1978). If the individuals' responses to job satisfaction questions were purely idiosyncratic, then
such relationships as the ones above would not be so consistently found (Clark, forthcoming).
The ability of job satisfaction to in
uence the economic outcomes listed above makes it an
important variable to analyze. But what are the determinants of satisfaction with work? One can
9
view expressed job satisfaction as a mental mapping of all the objective and subjective charac-
teristics of the job into an index of satisfaction (Hamermesh, 2001). Therefore, satisfaction with
work does not depend solely on the respondent's objective circumstances, such as his/ her salary
or job type, but also on subjective factors, such as the his/ her psychological state, aspirations,
willingness to voice discontent, the hypothetical alternatives to which the current job is compared,
or the relative importance for each individual of the various objective measures. For example,
if job security is what one values, then this is a signal that more emphasis should be placed on
dealing with unemployment. The way objective job characteristics are perceived is important
because job satisfaction tends to follow these perceptions. Because it re
ects both objective and
subjective factors, job satisfaction is more complex than standard economic variables (Freeman,
1978).
When ranking various life domains, individuals often place job towards the top of the ranking
(Clark, forthcoming). It is therefore not surprising that satisfaction with work, together with
family, nances and health, is one of the most important predictors of overall life satisfaction
(Argyle, 1989; Judge and Watanabe, 1993). Furthermore, individuals spend more time at work
than they spend doing almost anything else (Clark, forthcoming). The feeling among economic
theorists, however, has generally been that subjective outcomes describing work cannot be linked
to any underlying concept of utility and that, even if they could, their subjective nature makes
them too noisy to be useful (Hamermesh, 2001; L evy-Garboua and Montmarquette, 2004). Such
a reluctance towards dealing with subjective variables { variables that measure \what people say"
rather than \what people do" (Freeman, 1978) { comes as a surprise for non-economists (Sousa-
Poza and Sousa-Poza, 2000a,b) given that well-being is arguably the central economic variable
driving individuals' decisions.
My focus is on studying job satisfaction from an empirical perspective. Although subjec-
tive variables have to be treated cautiously, answers to questions about satisfaction with work
can convey important information about economic activity, as the previous literature review has
10
shown. Job satisfaction is therefore a variable worth studying, not as a replacement for objective
measures, but as an added dimension.
Labor relations in communist countries were dierent from those in free market economies
and the dierence between the two systems started with the educational process. In the East,
the young were schooled for the communist system through restrictions that channeled them
towards certain study paths and denied them others. The same process continued after school
using forms of intervention ranging from channeling toward a place of work to compulsory posting
to one (Kornai, 1992). Furthermore, the state set wages, prices, and enterprise budgets in ways
that ensured no open unemployment, but produced low real wages and narrow skill and sectoral
pay dierences. Such policies led to inecient allocations of labor and a preference towards
egalitarianism (Blanch
ower and Freeman, 1997). Such a system also gave employees an incentive
to shirk because they could be sure to get their wages without any great eort. With a political
system embodying the power of the working class, whether they believed this or not, workers
were encouraged to act more openly against their bosses and resist them silently, thanks to this
illusion of power. At the same time, however, while trade unions were prevalent, they were also
characterized by a complete lack of independence (Kornai, 1992). In terms of satisfaction with
work, the eect of socialist policies is more ambiguous { if one values job security most, then they
have a positive eect, while if one looks to be rewarded commensurately to his/ her skills, then
they have a negative eect.
Capitalist countries are facing their own set of problems. Due to the globalization process,
jobs become more scarce as unskilled work moves away from the developed countries into the
poorer ones, with lower wage costs. Paid work is becoming more and more a privilege, a basic
status-forming activity of Western civilization (Vecern k, 2003). This increased risk of losing one's
job has stressful consequences on people's lives (Beck, 1992, 2000). As a result, one typically nds
a fall in job satisfaction in Western countries such as Britain or Germany (Green and Tsitsianis,
11
2005). A higher intensity of work eort and declining task discretion are some other possible
explanations for such declining trends.
The transformation of political and economic systems that was brought about by the tran-
sition of the CEE countries from a planned economy to a free market model means that these
countries are now also faced with these challenges of capitalism, while the in
uence of the half
century of communism continues to be felt. The expectations of the population concerning social
protection remain high, the pressures put on work performance and job mobility continue to be
weak, and deciencies in work habits and a reluctance towards
exibility persist in large sections
of the labor force (Vecern k, 2003). Because of the persistence of poor working conditions in the
former communist countries, in which workers have long been deprived of normal market modes of
responding to such conditions, having neither the 'exit' option of nding employment outside the
state-run sector nor the 'voice' option of forming free trade unions, employees in these countries
are expected to show relatively low levels of job satisfaction at the beginning of the transition
process (Blanch
ower and Freeman, 1997).
It is therefore interesting to see to what extent there are some systematic dierences between
the workers in the former communist societies of CEE and the workers in the West in terms of
the level and determinants of their job satisfaction. If such dierences exist, do they get narrower
as the Central and Eastern European countries make progress in their transition to a market
economy? I nd that job satisfaction in transition countries actually decreases during the 1990s
which makes them fall even further behind Western Europe, where no signicant change in level
occurs. This decrease is mainly the result of birth cohort replacement, younger cohorts having
lower levels of satisfaction with work. However, the job satisfaction of younger people relative to
older people gradually improves. Together with the improvements at the macroeconomic level, this
leads to an increase in job satisfaction in Eastern Europe between 1997 and 2005 and a narrowing
of the gap with the West. Are such cross-country comparisons meaningful or do they re
ect
cultural dierences in the way people rate their experiences? Looking at overall life satisfaction,
12
Bolle and Kemp (2009) nd that such national dierences in rated life satisfaction are real rather
than re
ecting dierences in how satisfaction is rated, and therefore comparisons of international
averages of life satisfaction are meaningful.
I will next present the two datasets and the methodology employed in my empirical analysis.
The ndings are then presented separately for the 1990-1999 and 1997-2005 time interval. The
last section sums up the analysis.
2.2 Data and Methodology
2.2.1 Data
The data come from two dierent sources: the World Values Survey (WVS, 2009) and the Inter-
national Social Survey Programme (ISSP) \Work Orientations" module. The WVS is a multi-
country survey that covers people's attitudes toward a broad range of issues, such as economics
and politics, family, work, or religious values. So far three of its waves asked a question on job
satisfaction: wave 1 (1981-1984), wave 2 (1989-1993), and wave 4 (1999-2004), but only the last
two included transition countries so they will be the only ones used in the current analysis.
Although quite a few CEE countries were surveyed in wave 2 of the WVS, I will focus on
those in which the survey was carried out soon enough after the change in regime for the data
to roughly re
ect the pre-transition situation. Following the methodology described in Easterlin
(2009, p. 132), the group of countries for which early transition observations are available includes
Belarus, Estonia, Hungary, Latvia, Lithuania, Poland, and the Russian Federation. Belarus will
be excluded from the analysis because no job satisfaction question was asked in wave 2, leaving
me with six transition countries. For Western Europe, the exact date of the wave 2 survey was not
as crucial due to the fact that no regime change occurred in this region. Furthermore, the survey
dates for the various countries spanned a shorter period of time than was the case in Eastern
Europe, being carried out mostly throughout 1990. The non-transition countries included in the
13
analysis are Austria, Belgium, Denmark, Finland, France, Great Britain, Iceland, Ireland, Italy,
Netherlands, Portugal, Spain, and Sweden. In all nineteen countries { six in Eastern Europe and
thirteen in Western Europe { the wave 4 survey was carried out between 1998 and 2000. This
means that the WVS provides me with data roughly at the beginning of the transition process
and a decade later. For simplicity, the wave 2 survey will be dated 1990 and the wave 4 survey
will be referred to as 1999.
In order to bring the analysis of job satisfaction and its determinants more up to date, the
WVS will be supplemented with data from the ISSP \Work Orientations" surveys, which took
place in 1989, 1997, and 2005
2
, and gathered varied objective and subjective job information.
The country coverage is much less extensive than the WVS, focusing on OECD members, and
it tends to change from one survey to the next. Only one transition country { Hungary { and
one Western European country { Great Britain { were included in all three waves. Therefore,
I will focus on the 1997 and 2005 surveys. There were ve Eastern European and ve Western
European countries included in both of these surveys. The transition group includes Bulgaria,
Czech Republic, Hungary, the Russian Federation, and Slovenia. The Western European group
is represented by Great Britain, Denmark, Portugal, Spain, and Sweden.
The sample was reduced to include only employed respondents between 18 and 65 years old.
Because the job satisfaction question was typically asked only of those with a job, I focus on
employed individuals. The age restriction is meant to ensure that the respondents belong to age
categories for which it is typical to be involved in the labor market and not still be in school or
retired. Both surveys use nationally representative data.
Appendix A.1 lists the variables that I use in the job satisfaction analysis. Some of the
variables are recoded to ensure a more meaningful analysis, either by grouping certain categories
together, or by using indicator variables. As much as possible, the variables in the two surveys
2
Details regarding the questionnaire, sampling, and data collection are available in the Study Monitor-
ing Report for each ISSP wave. That for 2005, for example, is available via the following web page:
http://www.gesis.org/en/services/data/survey-data/issp/modules-studyoverview/work-orientations/2005/
14
are coded in a similar way. However, the scale for the job satisfaction question diers. In the
WVS job satisfaction is assessed on a scale from 1 to 10: \Overall, how satised or dissatised
are you with your job?", with 1 = \Dissatised" and 10 = \Satised". In the ISSP, the scale
is 1 to 7: \How satised are you in your (main) job?", with 1 = \completely dissatised" and
7 = \completely satised". Although an income variable is available, it is quite an imperfect
one. In the WVS it is usually coded as a ten-step variable. While in some cases these steps have
income brackets attached to them, in others they represent just a subjective ladder on which the
respondents place themselves. For simplicity, I assume this variable to represent ten dierent
perceived income brackets. In the ISSP, income is coded on dierent scales from country to
country. In order to have a consistent variable across countries, I use income quintiles to measure
the respondent's relative income. With regards to subjective variables, in the WVS respondents
are asked to assess the degree of freedom of decision they have at work. In the ISSP survey, they
rate their jobs on a larger number of dimensions which are described in Appendix A.1. In addition
to these individual level variables, I also use some macroeconomic indicators: GDP per capita,
the unemployment rate, and in
ation. Even if the analysis of job satisfaction focuses on those
employed at the time of the survey, the general unemployment rate is an indication of the risk of
losing their jobs and is therefore expected to have an eect on their satisfaction with work.
Tables 2.1 and 2.2 present descriptive statistics for the WVS variables used in the analysis for
each of the two dates covered here for Eastern Europe and Western Europe, respectively. In the
case of job satisfaction the descriptive statistics are presented at the regional level, as well as for
each of the individual countries. Simply by looking at the sample means, one sees that between
the early and the late 1990s, Eastern Europe actually falls behind in terms of the level of job
satisfaction. If at the onset of the transition there are a few Western countries { France, Italy,
Spain { with mean job satisfaction below that of some of the former communist economies, by 1999
all of the transition countries are below every single country in Western Europe. The individual
characteristics of the population are fairly stable in the West, with many more changes happening
15
Table 2.1: Descriptive Statistics for Key Variables in the World Values Survey, wave 2 and wave
4, transition countries
Wave 2: 1989-1991 Wave 4: 1998-2000
Variable Number Sample Standard Number Sample Standard
of obs. mean deviation of obs. mean deviation
Job satisfaction 5,061 6.74 2.37 3,739 6.53 2.46
Estonia 754 6.70 2.33 561 6.67 2.25
Hungary 551 7.34 2.26 419 6.85 2.31
Latvia 660 6.46 2.45 445 6.74 2.33
Lithuania 665 6.89 2.27 515 6.84 2.45
Poland 1,131 7.12 2.18 499 6.62 2.34
Russian Federation 1,300 6.26 2.49 1,300 6.17 2.64
Individual characteristics
Male 5,059 0.49 0.50 3,739 0.52 0.50
Age 5,061 39.26 11.30 3,739 38.98 10.92
ACE 4,313 18.81 3.12 3,722 19.74 2.85
Married 5,054 0.76 0.43 3,728 0.64 0.48
Single 5,054 0.13 0.34 3,728 0.18 0.39
Div./separated/widowed 5,054 0.11 0.31 3,728 0.18 0.38
Full time 5,061 0.85 0.36 3,739 0.86 0.35
Part time 5,061 0.10 0.30 3,739 0.09 0.29
Self employed 5,061 0.05 0.22 3,739 0.05 0.22
White collar 3,431 0.45 0.50 3,707 0.45 0.50
Union 4,492 0.59 0.49 3,739 0.22 0.42
Income bracket 4,831 4.88 1.92 3,468 5.36 2.51
Subjective variable
Freedom of decision 4,945 6.03 2.84 3,661 5.54 2.93
in the East during this rst decade of transition. The education level increases in both regions.
Although still higher than in Western Europe, the percentage of married people in transition
countries decreases by twelve percentage points. Part time jobs and self employment are more
common in the West than in the East and there is not much change in this respect in any region
during the decade covered here. Union membership is considerably lower in 1999 compared with
1990 in both regions, but the decrease is really drastic in Eastern Europe. During communism,
trade unions acted as political \transmission belts" and not as independent representatives of the
workers, thus producing demotivated, demoralized and unsatised workers (Lange and Georgellis,
2007). It is not surprising then, that once given a choice, Eastern European employees prefer to
withdraw from labor unions. Finally, while the perceived freedom of decision on the job increases
in Western Europe, it actually goes down in the transition countries. This is quite surprising
16
Table 2.2: Descriptive Statistics for Key Variables in the World Values Survey, wave 2 and wave
4, non-transition countries
Wave 2: 1989-1991 Wave 4: 1998-2000
Variable Number Sample Standard Number Sample Standard
of obs. mean deviation of obs. mean deviation
Job satisfaction 10,146 7.52 1.99 8,408 7.56 1.85
Austria 698 8.00 1.75 759 7.72 2.03
Belgium 1,294 7.72 1.81 852 7.61 1.80
Denmark 642 8.23 1.66 623 8.05 1.79
Finland 424 7.55 2.03 504 7.73 1.50
France 461 6.79 1.97 773 7.11 1.90
Great Britain 837 7.36 2.09 479 7.33 1.88
Iceland 549 7.83 1.74 714 7.86 1.56
Ireland 527 7.76 2.01 522 7.77 2.00
Italy 1,036 7.27 2.10 972 7.31 2.01
The Netherlands 435 7.49 1.67 631 7.55 1.34
Portugal 614 7.34 2.30 438 7.63 1.98
Spain 1,862 7.07 2.08 495 7.32 1.90
Sweden 767 7.92 1.84 646 7.32 1.85
Individual characteristics
Male 10,132 0.61 0.49 8,405 0.57 0.50
Age 10,146 37.81 11.88 8,408 38.94 11.29
ACE 9,783 17.89 3.79 8,227 18.70 3.70
Married 10,132 0.69 0.46 8,351 0.59 0.49
Single 10,132 0.24 0.43 8,351 0.31 0.46
Div./separated/widowed 10,132 0.07 0.26 8,351 0.11 0.31
Full time 10,146 0.78 0.41 8,408 0.75 0.44
Part time 10,146 0.11 0.31 8,408 0.15 0.35
Self employed 10,146 0.11 0.32 8,408 0.11 0.31
White collar 9,428 0.55 0.50 8,157 0.58 0.49
Union 10,146 0.26 0.44 8,408 0.30 0.46
Income bracket 8,463 5.88 2.67 6,810 6.07 2.41
Subjective variable
Freedom of decision 10,091 6.89 2.56 8,379 7.06 2.34
given that the lack of freedom was one of the main negative features of the communist regimes
that was supposed to end with the switch to capitalism. In fact, through the many functions it
fullled (cf. Kornai, 1992, pp. 221-222) the rm could become a cell of totalitarian power, not
just a scene of work. At the same time, however, with a political system embodying the power
of the working class, during communism employees may have be given the illusion that they had
more control than they actually did, and this could explain the decrease in perceived freedom of
decision during the post-communist years.
17
Table 2.3: Descriptive Statistics for Key Variables in the International Social Survey Programme,
1997 and 2005, transition countries
1997 2005
Variable Number Sample Standard Number Sample Standard
of obs. mean deviation of obs. mean deviation
Job satisfaction 2,923 4.92 1.26 2,998 5.06 1.25
Bulgaria 459 4.96 1.17 477 5.22 1.23
Czech Republic 514 5.13 1.14 659 5.02 1.06
Hungary 624 4.78 1.16 439 5.10 1.20
Russia 812 4.86 1.48 925 4.95 1.44
Slovenia 514 4.97 1.13 498 5.08 1.16
Individual characteristics
Male 2,923 0.54 0.50 2,998 0.53 0.50
Age 2,923 38.63 10.33 2,998 39.76 11.19
Less than high school 2,921 0.38 0.48 2,947 0.31 0.46
High school 2,921 0.48 0.50 2,947 0.53 0.50
University 2,921 0.14 0.35 2,947 0.16 0.37
Married 2,916 0.72 0.45 2,972 0.63 0.48
Single 2,916 0.16 0.37 2,972 0.24 0.42
Div./separated/widowed 2,916 0.12 0.32 2,972 0.14 0.34
Full time 2,923 0.84 0.37 2,998 0.86 0.34
Part time 2,923 0.06 0.23 2,998 0.05 0.23
Self employed 2,923 0.10 0.30 2,998 0.08 0.27
Public sector 2,879 0.59 0.49 2,942 0.36 0.48
White collar 2,761 0.55 0.50 2,800 0.56 0.50
Union 2,923 0.40 0.49 2,923 0.23 0.42
Income quintile 2,399 2.87 1.39 2,181 2.88 1.44
Subjective variables
Job security 2,887 3.44 1.23 2,956 3.54 1.11
Pay 2,901 2.50 1.09 2,978 2.70 1.13
Promotion 2,844 2.38 1.06 2,956 2.57 1.11
Interesting job 2,885 3.58 1.08 2,974 3.56 1.06
Independent work 2,879 3.63 1.16 2,970 3.51 1.15
Help others 2,848 3.67 1.11 2,950 3.62 1.07
Useful to society 2,851 4.01 0.88 2,949 3.90 0.91
Exhausted after work 2,910 2.56 0.90 2,984 2.53 0.93
Physical work 2,906 3.45 1.33 2,980 3.45 1.32
Stressful work 2,892 2.99 1.15 2,967 2.93 1.18
Dangerous conditions 2,870 3.61 1.38 2,947 3.72 1.35
Similar descriptive statistics for the ISSP variables are shown in table 2.3 for Eastern Europe,
and table 2.4 for Western Europe. Between 1997 and 2005, there is a recovery in job satisfaction
among the Eastern European countries. Combined with the lack of change in Western Europe,
this leads to some convergence between the two regions, although not enough for the former
communist countries to catch up. Even in 2005, four of the ve Western countries included in
the analysis are above each of their transition counterparts. The education level improves in
18
both regions. The percentage of married people continues to decrease considerably in Eastern
Europe, so that by 2005 it is actually lower than in the West. I nd a similar decrease in union
membership, which is now much lower in the East. In transition countries there is also a decrease
in the percentage of people employed in the public sector of almost thirty percentage points.
This is in line with the transformations involved in the transition process, such as privatization.
Unfortunately, a similar variable is not available in the WVS.
Table 2.4: Descriptive Statistics for Key Variables in the International Social Survey Programme,
1997 and 2005, non-transition countries
1997 2005
Variable Number Sample Standard Number Sample Standard
of obs. mean deviation of obs. mean deviation
Job satisfaction 3,193 5.31 1.17 3,972 5.31 1.12
Great Britain 528 5.12 1.23 460 5.25 1.27
Denmark 622 5.70 1.07 1,110 5.51 1.11
Portugal 857 5.17 1.28 1,025 5.26 1.07
Spain 395 5.40 1.09 556 5.24 1.10
Sweden 791 5.25 1.03 821 5.18 1.08
Individual characteristics
Male 3,193 0.54 0.50 3,972 0.52 0.50
Age 3,193 40.45 11.61 3,972 41.66 11.73
Less than high school 2,373 0.55 0.50 3,917 0.46 0.50
High school 2,373 0.28 0.45 3,917 0.36 0.48
University 2,373 0.17 0.37 3,917 0.18 0.39
Married 3,186 0.65 0.48 3,962 0.67 0.47
Single 3,186 0.28 0.45 3,962 0.24 0.43
Div./separated/widowed 3,186 0.07 0.26 3,962 0.10 0.29
Full time 3,193 0.72 0.45 3,972 0.76 0.43
Part time 3,193 0.15 0.36 3,972 0.13 0.33
Self employed 3,193 0.13 0.34 3,972 0.12 0.32
Public sector 3,175 0.33 0.47 3,903 0.32 0.47
White collar 2,202 0.61 0.49 3,746 0.68 0.46
Union 3,193 0.49 0.50 3,925 0.48 0.50
Income quintile 2,495 2.62 1.47 3,337 2.67 1.39
Subjective variables
Job security 3,152 3.62 1.28 3,920 3.73 1.14
Pay 3,173 2.54 1.14 3,937 2.69 1.12
Promotion 3,120 2.57 1.16 3,887 2.74 1.12
Interesting job 3,179 3.99 1.00 3,941 3.91 1.00
Independent work 3,175 4.00 1.10 3,946 3.90 1.10
Help others 3,163 3.92 1.09 3,940 3.82 1.08
Useful to society 3,129 4.01 1.02 3,927 3.91 1.02
Exhausted after work 3,182 2.64 0.92 3,956 2.66 0.89
Physical work 3,182 3.40 1.27 3,951 3.43 1.26
Stressful work 3,180 2.78 0.99 3,953 2.80 0.98
Dangerous conditions 3,167 3.87 1.21 3,930 3.97 1.17
19
Figure 2.1: GDP per capita index (1989=100) and absolute GDP per capita in transition and
non-transition countries for the WVS and ISSP groups of countries, respectively, 1989-2006
Source: World Bank (2009) World Development Indicators data in PPP constant 2005 international dollars,
except Lithuania (1989), Poland (1989), Slovenia (1989), and Czech Republic (1989), extrapolated using the
Economic Commission for Europe (2003).
In terms of macroeconomic conditions, gures 2.1 and 2.2 show the GDP per capita and the
unemployment rate for Eastern and Western Europe over the 1989-2006 time interval, separately
for the countries covered in the WVS and for the countries covered in the ISSP survey. The
trends are quite similar for the countries in the two dierent surveys, which supports the idea of
complementing the analysis based on the WVS with that using the ISSP. Figure 2.1 graphs the
GDP per capita both as an index (1989=100) and in absolute value. In Western Europe there is a
fairly consistent increase, so that in 1999 GDP is almost 20 percentage points higher than in 1990,
followed, by 2006, by an increase of another 20 percentage points. In Eastern Europe, the rst
few years of transition bring a collapse in GDP. Despite the fact that a recovery was underway
in the second half of the 1990s, the GDP index in 1999 is still almost 10 percentage points lower
than the 1990 value. Following a steep increase, the GDP index in Eastern Europe by 2005 is
20
Figure 2.2: The unemployment rate in transition and non-transition countries for the WVS and
ISSP groups of countries, respectively, 1989-2006
Source: International Labor Organization (2009) KILMS database, except Hungary (1990, 1991), Latvia
(1992-1995), Lithuania (1992, 1993), Poland (1990, 1991), Bulgaria (1990, 1991, 1992), and Slovenia (1990, 1991),
from Economic Commission for Europe (2003).
Note: When the unemployment rate for one country is missing in a certain year, the regional average is obtained
using the remaining countries in the respective region. The following rates are missing: Iceland (1989, 1990),
Hungary (1989), Latvia (1989, 1990, 1991), Lithuania (1989, 1990, 1991), Poland (1989), Russia (1989, 1990,
1991), Slovenia (1989, 1990), Bulgaria (1989), Czech Republic (1989), and Slovenia (1989).
around 140, that is, almost the same as in Western Europe. This means that the GDP in the
two regions increases by a similar percentage between 1989 and 2006. However, while the change
in Western Europe is fairly steady, the transition countries go through a much more tumultuous
period, with a collapse followed by recovery. Furthermore, the absolute value of GDP in Eastern
Europe is much lower than in the West, the gap being fairly consistently $10,000 or more.
Figure 2.2 shows the unemployment rate series. In Western Europe, the mid-1990s are marked
by an increase in unemployment but by 1999 the unemployment rate is already lower than in 1989.
The situation continues to improve until 2006. Just like in the case of GDP, in Eastern Europe
during the rst few years of transition the unemployment conditions become dramatically worse.
21
However, while the recovery in terms of GDP has almost been achieved by 1999, the increase
in unemployment continues until around 2000. As a consequence, in 1999 the unemployment
rate in Eastern Europe is in the double digits. An improvement follows in the 2000s, so that
in 2006 the rate is similar to that in Western Europe. Looking at the overall change in the
unemployment rate between 1989 and 2006, while in Western Europe there is an improvement, in
Eastern Europe the rate is still higher in 2006 compared with the pre-transition situation. Such
dramatic increases in unemployment, from virtually zero to double digits, even if followed by a
partial recovery, inevitably have a negative eect on the job satisfaction of the workers in the
transition countries through the decrease in job security. In fact, Clark (forthcoming) shows that
an employed individual in a high unemployment area is less likely to report high levels of job
satisfaction than one living in a low unemployment area.
Although the country coverage of the WVS and the ISSP is dierent, the two groups repre-
senting each region are quite similar at the macroeconomic level, as gures 2.1 and 2.2 showed
for GDP and unemployment. Therefore, the conclusions based on these two dierent surveys can
reasonably complement each other for an analysis that covers a longer time period. The main
downside of these two datasets is that they are not longitudinal studies which makes it dicult
to deal with reverse causation problems. For example, it may be that an interesting job is what
makes an employee satised, but it is also possible that a satised employee is more likely to
consider his job interesting. Or it may be that being a union member makes one less satised,
but it can also be the case that workers joined the union because they were dissatised with their
jobs.
2.2.2 Methodology
The analysis will rst focus on job satisfaction in Eastern and Western Europe during the 1990s,
using the data of the WVS. Are there any signicant dierences in job satisfaction at the onset of
the transition process? How do the trends in job satisfaction during the rst decade of transition
22
compare between the two regions? Are there signs of convergence? To the extent that there are
dierences in the level and trend of job satisfaction between the two groups of countries, are these
due to dierences in the objective macroeconomic conditions in the two regions? How about
the impact of other objective circumstances, at the individual level? What role do subjective
perceptions play? Are these determinants of job satisfaction the same in transition and non-
transition countries? Do they dier over time? In order to get a more updated view, I then
extend the analysis to the 1997 - 2005 time interval, using data from the ISSP, with similar
questions being asked.
As far as the determinants of job satisfaction are concerned, it is reasonable to think that
macroeconomic conditions have a strong impact on people's job satisfaction. Also important are
individual circumstances as captured in the survey data. Finally, it is possible that subjective
perceptions of what one's job has to oer also in
uence the way people assess their level of job
satisfaction beyond what objective conditions would predict. Therefore, in a simplied manner,
a job satisfaction function could be described as follows:
JS
it
=f(macroeconomic conditions
ct
; individual circumstances
it
;
job characteristics rating
it
)
(2.1)
where JS stands for job satisfaction, i indexes individuals, t indexes time, and c indexes countries.
Based on the survey data from the WVS and the ISSP and on the outside information regard-
ing country level variables, this can be written as a model of the following form:
JS
it
=f(Y
ct
; X
it
; Z
it
) +
ict
(2.2)
where Y = a vector of macroeconomic indicators, consisting of log GDP per capita, unemployment
rate, and in
ation*10
1
; X = a vector of individual controls, consisting of gender, age, education,
23
marital status, employment level, union membership, sector, occupation type, and income; Z = a
vector of job characteristics, including freedom of decision on the job, pay and other job aspects.
This model can be implemented through regression analysis, which in this case is carried
out using ordinary least squares (OLS) regressions.
3
In all the regressions the standard errors
are adjusted to allow for clusters in the error term within countries. This makes the standard
errors substantially larger, and therefore the coecients are less likely to be signicant. Of
course, labor market outcomes are likely to dier among former communist countries, as well as
among Western countries. Despite any national dierences though, the hypothesis here is that
the communist experience is suciently similar to leave an identiable common legacy aecting
outcomes and views of the labor market in these countries (Blanch
ower and Freeman, 1997).
4
In order to account for any heterogeneity within regions though, country xed-eects regressions
are used where appropriate. This approach also removes the unobserved characteristics that are
constant across individuals from the same country, thus removing the estimation bias caused by
individual-invariant country characteristics.
The analysis rst focuses on the general determinants of job satisfaction. In order to do this, I
look at the overall sample, including both Eastern and Western Europe, and both surveys available
in each dataset:
JS
it
=
0
+
1
T
c
+
2
W
t
+
0
1
Y
ct
+
0
1
X
it
+
0
1
Z
it
+
ict
: (2.3)
The above equation does not show whether the determinants of job satisfaction have a dier-
ential eect in the two groups of countries. This can be investigated by interacting the transition
3
Because the answers to the job satisfaction questions take on ordered, discrete values, an ordered logit regression
would be recommended for this type of analysis. The results, however, are quantitatively similar to those from the
OLS specications, not surprising given that the answer scales are fairly wide { 1 to 10 in the WVS and 1 to 7 in
the ISSP. Therefore, only the OLS results are reported in the paper due to their easier interpretation.
4
A series of F-tests were carried out to compare the overall regional results with the outcomes for the individual
countries within each group, and enough homogeneity was found to support the grouping of countries in Eastern
and Western Europe.
24
country dummy, T
c
, with the variables of interest, according to the following equations:
JS
it
=
0
+
3
T
c
W
t
+
4
T
c
Y
ct
+
0
1
Y
ct
+
0
1
X
it
+
0
1
Z
it
+
ict
; (2.4)
for Y
ct
2 Y
ct
,
JS
it
=
0
+
3
T
c
W
t
+
5
T
c
X
it
+
0
1
Y
ct
+
0
1
X
it
+
0
1
Z
it
+
ict
; (2.5)
for X
it
2 X
it
,
JS
it
=
0
+
3
T
c
W
t
+
6
T
c
Z
it
+
0
1
Y
ct
+
0
1
X
it
+
0
1
Z
it
+
ict
; (2.6)
for Z
it
2 Z
it
. If the coecient for the interaction term of a certain variable with the transi-
tion country dummy is signicantly dierent from zero, then the eect of that variable on job
satisfaction is dierent in transition and non-transition countries. The expectation, however, is
that these coecients are not usually dierent from zero so that dierences in satisfaction with
work between the two groups of countries are mainly the result of dierences in the level of the
determinants and not in their nature.
It is also possible that the determinants of job satisfaction change over time. While this is
less likely in Western Europe, such changes are expected in the East as a result of the transition
process. Furthermore, if a change is indeed found in the relative level of job satisfaction of various
demographic groups, it can be indicative of who the winners and losers of the transition process
are. Interacting W
t
with the various determinants of job satisfaction allows me to see if any
signicant changes occurred over time:
JS
it
=
0
+
2
W
t
+
7
W
t
Y
ct
+
0
1
Y
ct
+
0
1
X
it
+
0
1
Z
it
+
c
+
ict
; (2.7)
for Y
ct
2 Y
ct
,
25
JS
it
=
0
+
2
W
t
+
8
W
t
X
it
+
0
1
Y
ct
+
0
1
X
it
+
0
1
Z
it
+
c
+
ict
; (2.8)
for X
it
2 X
it
,
JS
it
=
0
+
2
W
t
+
9
W
t
Z
it
+
0
1
Y
ct
+
0
1
X
it
+
0
1
Z
it
+
c
+
ict
; (2.9)
forZ
it
2 Z
it
. These equations are estimated for transition and non-transition countries separately,
using country xed-eects regressions.
The next step is to see the extent to which these determinants are able to explain the dierence
in the level of job satisfaction between transition and non-transition countries. This is done at
each date for which surveys are available. Versions of the following equation are estimated:
JS
it
=
0
+
1
T
c
+
0
1
Y
ct
+
0
1
X
it
+
0
1
Z
it
+
ict
; (2.10)
for each t2f1990; 1999g for the WVS, and each t2f1997; 2005g for the ISSP data. T
c
is an
indicator variable equal to one if the respondent lives in a transition country. If
1
is signicantly
dierent from zero then the level of job satisfaction is dierent in Eastern and Western Europe.
The various country level and individual level controls will be added to the regression incrementally
in order to see what the role of each group of variables is in explaining any dierences in satisfaction
with work between regions. This role is measured through the impact that the added regression
controls have on
1
.
The job satisfaction trend in transition and non-transition countries for each of the two time
intervals { the 1990s, using the WVS, and 1997 to 2005, using the ISSP survey { as well as its
determinants are analyzed using the following equation:
JS
it
=
0
+
2
W
t
+
0
1
Y
ct
+
0
1
X
it
+
0
1
Z
it
+
c
+
ict
; (2.11)
26
estimated for Eastern and Western Europe separately. W
t
is a dummy variable equal to one for
the latter date in the survey and zero for the earlier date. Therefore, if
2
is signicantly dierent
from zero then job satisfaction displays a signicant trend in the respective region. Country
xed-eects will be used and they are denoted by
c
. The extent to which the trend is due to
various country or individual level variables is captured in the impact that controlling for these
variables has on
2
. It is possible, however, that the job satisfaction trend is not only the result
of changes in circumstances, but also of cohort replacement. If the new generations entering the
sample are intrinsically more or less satised with their jobs, irrespective of their circumstances,
this can in
uence the results. In order to further examine this hypothesis, I also include year of
birth controls in my regressions.
The next section details the results obtained through the empirical implementation of the
equations above.
2.3 Findings
2.3.1 Job Satisfaction in the 1990s
The rst part of the analysis focuses on job satisfaction during the 1990s, using the data of the
WVS. The purpose is to look at how job satisfaction in Eastern Europe is aected by the transition
process and to compare this with the situation in Western Europe. The earliest survey available
dates from around 1990. Although the transformation in the former communist countries had in
many cases started before 1990, this survey can still serve as a rough re
ection of the starting
point of what was expected to be a convergence towards Western standards. Table 2.5 shows
the mean job satisfaction in the two regions at the two dates available, 1990 and 1999, for the
overall population, as well as for women and men separately. The dierences in job satisfaction
level between East and West are calculated in column (3). Job satisfaction in Western Europe
is signicantly higher than in the former communist countries both in 1990 and in 1999. In
27
Table 2.5: Mean job satisfaction by region and gender, 1990 and 1999
Eastern Western Dierence
Europe Europe (1)-(2)
(1) (2) (3)
All
1990 6.74 7.52 -0.78**
(0.03) (0.02) (0.04)
1999 6.53 7.56 -1.02**
(0.04) (0.02) (0.05)
Dierence -0.21** 0.03 -0.24**
(0.05) (0.03) (0.06)
Women
1990 6.71 7.48 -0.77**
(0.05) (0.03) (0.06)
1999 6.47 7.51 -1.03**
(0.06) (0.03) (0.07)
Dierence -0.24** 0.02 -0.27**
(0.08) (0.05) (0.09)
Men
1990 6.77 7.55 -0.78**
(0.05) (0.03) (0.05)
1999 6.59 7.59 -1.00**
(0.06) (0.03) (0.07)
Dierence -0.18* 0.05 -0.23**
(0.07) (0.04) (0.08)
Signicance levels: ** p<0.01, * p<0.05, + p<0.10.
fact, the dierence is even bigger ten years into the transition than it was at the beginning of
the process { 1.02 compared with 0.78 { and this gap increase is statistically signicant. This
means that by 1999, when asked to assess their job satisfaction on a scale from one to ten,
people in Eastern Europe would on average pick a value about one point lower than their Western
European counterparts. This considerable dierence shows that the process of transition did not
automatically bring an improvement in one of the most important life domains, work. This lack
of convergence is mostly due to a statistically signicant decrease in job satisfaction in Eastern
Europe from 6.74 to 6.53. In Western Europe there is a slight, but statistically insignicant
increase in job satisfaction from 7.52 to 7.56. Table 2.5 also shows the levels and changes in
satisfaction with work for women and men separately. The trends are very similar to those for the
overall population in both regions, while the levels are slightly higher for men than for women.
28
The two genders are therefore similar enough to justify a unitary analysis of satisfaction with
work for the overall population.
Table 2.6: Changes in job satisfaction by country, 1990-1999
(1) (2) (3) (4)
Eastern Europe Western Europe
Hungary (-) ** (-) ** Austria (-) * (-) **
Estonia 0 0 Belgium 0 0
Latvia (+)+ (+)+ Denmark (-)+ (-)*
Lithuania 0 0 Finland 0 0
Poland (-) ** (-) ** France (+)** (+)*
Russia 0 (-) * Great Britain 0 0
Iceland 0 0
Ireland 0 0
Italy 0 0
The Netherlands 0 0
Portugal (+)+ (+)+
Spain (+)* (+)*
Sweden (-) ** (-) **
The columns are based on the signs of OLS coecients. Columns (1)
and (3) use no controls. Columns (3) and (4) control for gender, age,
and education. Signicance based on robust standard errors.
Signicance levels: ** p<0.01, * p<0.05, + p<0.10.
Table 2.6 looks at the job satisfaction trend in the various countries within each region, indi-
vidually. Columns (1) and (3) use no controls, making them the equivalent of the regional results
shown in table 2.5. Columns (2) and (4) take into account any demographic changes that might
have occurred in the sample between the two dates. Some heterogeneity at the regional level is
apparent. In general, the former Soviet republics seem to maintain their levels of job satisfaction
better than Hungary and Poland, but they also start o at lower levels as table 2.1 showed. In
Western Europe, over half of the countries show no signicant change in satisfaction with work,
while the rest are equally divided between increasing and decreasing trends. This lack of complete
homogeneity within each region supports the inclusion of country xed-eects whenever possible.
The general determinants of job satisfaction are analyzed in table 2.7. All countries and both
surveys are included. The rst two columns focus on objective variables, while the last column
adds perceived freedom of decision to the picture. As expected, macroeconomic conditions have a
29
Table 2.7: Ordinary least squares regressions of job satisfaction in Eastern and Western Europe,
1990-1999
(1) (2) (3)
Coe. Coe. Coe.
Variable (t-stat) (t-stat) (t-stat)
Ln GDP per capita 0.469+ 0.531 0.493+
(1.86) (1.69) (1.89)
Unemployment rate -0.031** -0.039** -0.018+
(-3.40) (-4.29) (-2.02)
In
ation*10
1
-0.009** -0.011** -0.008**
(-5.40) (-7.63) (-5.77)
Male 0.018 0.071+ -0.074+
(0.54) (1.84) (-1.74)
Age 30 - 39 -0.041 -0.018 -0.095*
(-1.20) (-0.44) (-2.24)
Age 40 - 49 0.077 0.052 -0.033
(1.52) (0.92) (-0.64)
Age 50 - 65 0.316** 0.342** 0.179**
(8.03) (7.21) (4.36)
Age completed education 0.012* -0.004 -0.018**
(2.12) (-0.87) (-4.01)
Single -0.267** -0.216** -0.163**
(-5.21) (-4.81) (-3.38)
Divorced/ separated/ widowed -0.184** -0.124* -0.128*
(-3.75) (-2.48) (-2.54)
Part time -0.219** -0.176** -0.147*
(-3.51) (-2.91) (-2.41)
Self employed 0.375** 0.329** -0.329**
(5.52) (4.08) (-3.27)
Belong to labor unions -0.000 0.024 0.007
(-0.00) (0.33) (0.11)
White collar 0.232** 0.015
(5.48) (0.38)
Income bracket 0.062** 0.026+
(3.74) (1.88)
Freedom of decision at work 0.326**
(28.08)
Constant 2.919 2.117 0.873
(1.14) (0.66) (0.32)
Observations 25,470 19,648 19,490
R-squared 0.060 0.070 0.210
All countries and both waves included. The omitted categories are age
18-29, married, and full time. Transition and year dummy variables
were also included. T-statistics are adjusted for clustering at country
level. Signicance levels: ** p<0.01, * p<0.05, + p<0.10.
30
statistically signicant in
uence, with GDP per capita having a positive impact on job satisfaction,
and unemployment and in
ation a negative one. The job satisfaction of men is held back by their
occupation and income because controlling for these in column (2) makes them signicantly more
satised than women, as opposed to the not signicant dierence found in column (1). Perceived
freedom of decision on the other hand, helps them because the coecients become negative in
column (3). People in the oldest group, 50 to 65 years old, are the most satised with work. There
is also an education premium in column (1) but it dissipates when occupation and income are
accounted for, and it actually becomes negative when subjective perceptions are included. This
means that education pays o through the income and freedom of decision it typically leads to.
Married people have higher job satisfaction than unmarried ones, but the direction of causality
here is less clear. In terms of employment level, part time jobs are associated with a lower
satisfaction, potentially because many people work part time only due to a lack of availability of
full time jobs. The self employed are the most satised, with subjective perceptions accounting for
their increased satisfaction because the coecient becomes negative in the last column. Similarly,
freedom of decision explains why people with white collar occupations are more satised at work.
Income and freedom of decision turn out to be the two main determinants of job satisfaction at
the individual level, both having a positive eect.
Is the impact of these factors on job satisfaction dierent in Eastern and Western Europe?
In order to answer this question, I add interaction terms between the transition country dummy
and the dierent explanatory variables. The expectation is that most of the dierence in the
level of satisfaction with work between the two regions is due to dierences in circumstances
and not to their dierential impact on job satisfaction and therefore, the coecients on the
interaction terms should typically not be signicant. Some dierences, however, do arise (table
2.8). Because unemployment decreases job satisfaction, while its interaction term has a positive
sign, this means that a higher unemployment rate is less likely to decrease job satisfaction in
Eastern Europe compared with the West { column (1). This could be the result of the fact
31
Table 2.8: Ordinary least squares regressions of job satisfaction with interactions between the
transition country dummy and specied independent variables, 1990-1999
(1) (2) (3) (4)
Coe. Coe. Coe. Coe.
Variable (t-stat) (t-stat) (t-stat) (t-stat)
Transition*unemployment rate 0.088*
(2.72)
Transition*age 30-39 -0.148*
(-2.17)
Transition*age 40-49 -0.263*
(-2.16)
Transition*age 50-65 -0.046
(-0.74)
Transition*white collar 0.190+
(1.82)
Transition*income bracket 0.077**
(3.28)
Observations 25,470 25,470 19,648 19,648
R-squared 0.062 0.060 0.070 0.072
The regression controls include transition and year dummies, ln GDP per
capita, unemployment, in
ation, gender, age group, marital status, employment
level, union membership. In columns (3) and (4) occupation and income bracket
are also controlled for. T-statistics are adjusted for clustering at country
level. Signicance levels: ** p<0.01, * p<0.05, + p<0.10.
that transition countries start o with lower levels of job satisfaction in 1990 despite lower levels
of unemployment, which is the opposite relationship to what one usually expects. The relative
satisfaction with work of people between 30 and 49 years old compared with the reference age
group, those 18 to 29, is lower in transition countries than in non-transition ones, since the
coecients for their interaction with the transition country dummy is negative and signicant.
Finally, the positive impact of occupation and income on job satisfaction { columns (3) and (4)
{ is stronger in transition countries. It will be interesting to see to what extent these dierences
between East and West get narrower as the transition progresses into the new millennium.
Within each group of countries, it is also possible to see changes in the impact of various
circumstances on job satisfaction over time. Indeed, such signicant interactions with time do
arise, as shown in table 2.9, especially for transition countries. The relative position of older age
groups compared with those between 18 and 29 years old worsens in transition countries { column
32
Table 2.9: Country xed-eects regressions of job satisfaction with interactions between time and
specied independent variables, 1990-1999
Western
Eastern Europe Europe
(1) (2) (3) (4) (5) (6)
Coe. Coe. Coe. Coe. Coe. Coe.
Variable (t-stat) (t-stat) (t-stat) (t-stat) (t-stat) (t-stat)
1999*age 30-39 -0.269+ -0.188*
(-1.71) (-2.23)
1999*age 40-49 -0.589** -0.328**
(-3.69) (-3.76)
1999*age 50-65 -0.631** -0.345**
(-3.60) (-3.76)
1999*age completed education 0.133**
(6.46)
1999*white collar 0.607**
(4.45)
1999*income bracket 0.109**
(2.99)
1999*freedom of decision at work 0.035+
(1.69)
Observations 7,546 7,546 5,651 5,651 7,377 17,924
R-squared 0.030 0.034 0.059 0.058 0.179 0.045
The regression controls include year, ln GDP per capita, unemployment,
in
ation, gender, age group, marital status, employment level, union membership.
In columns (3) and (4) occupation and income bracket are also controlled for.
Signicance levels: ** p<0.01, * p<0.05, + p<0.10.
(1) { pointing to young people as winners of the transition process. A similar improvement for
young people also occurs in Western Europe, but the magnitude is lower. In Eastern Europe, the
likelihood that a white collar job and a higher income lead to a higher level of job satisfaction
increases between 1990 and 1999. The education premium also increases, which is not surprising
given that the capitalist economy is more likely to reward education through a better job and a
higher income than was the case during communism. Additionally, in column (5) I nd that the
impact of perceived freedom of decision at work on satisfaction becomes stronger. To sum up,
young and more educated people fare quite well during the transition, at least compared with
older and less educated individuals, and the role of subjective perceptions in determining job
satisfaction increases during the 1990s.
33
Table 2.10: Ordinary least squares regressions of job satisfaction on region, 1990
Subjective
Objective conditions perceptions
(1) (2) (3) (4) (5)
Coe. Coe. Coe. Coe. Coe.
Variable (t-stat) (t-stat) (t-stat) (t-stat) (t-stat)
Transition country -0.778** -0.539 -0.434 -0.299 -0.493**
dummy (-3.68) (-1.27) (-1.27) (-0.97) (-3.20)
Observations 15,207 15,207 13,590 9,771 15,036
R-squared 0.029 0.040 0.055 0.060 0.193
Country level controls no yes yes yes no
Individual level controls no no yes yes no
Occupation and income no no no yes no
Freedom of decision no no no no yes
Number of countries 19 19 19 19 19
Reference group: Western Europe. Country level controls include ln GDP per capita,
unemployment rate, in
ation. Individual level controls include gender, age group,
age completed education, marital status, employment level union membership.
T-statistics are adjusted for clustering at country level.
Signicance levels: ** p<0.01, * p<0.05, + p<0.10.
The next step is to look at how the level of job satisfaction diers between transition and non-
transition countries and to nd what accounts for any dierence. This is done through ordinary
least square regressions separately for 1990 (table 2.10) and 1999 (table 2.11). Column (1) in each
table replicates the increasing gap found in table 2.5, transition countries having lower levels of
job satisfaction. Columns (2) through (4) look at the role of objective conditions in explaining
the gap, while column (5) focuses on freedom of decision. The potential determinants are added
gradually in order to isolate the role of each set. Among the individual level controls, occupation
and income are added in a separate step because they have a smaller number of observations. This
way I can assess the role of the other objective determinants without the diminished sample size.
The addition of subjective variables among the regression controls can produce exogeneity issues.
As a result, I incorporate them in the analysis separately from the objective variables. The role
of each set of determinants is measured through the impact on the coecient for the transition
country dummy compared with its magnitude in column (1). Dierences in objective conditions
are accountable for the dierence in job satisfaction between Eastern and Western Europe at both
34
dates because the coecients in column (4) are no longer statistically dierent from zero in either
table. Among these determinants, macroeconomic conditions have by far the most explanatory
power. In fact, in 1990 (table 2.10), the coecient in column (2) is no longer signicant, while in
1999 (table 2.11) almost 70 per cent of the gap between the two groups of countries is explained,
but at this latter date the dierence remains statistically signicant. Occupation and income
appear to have the most explanatory power among objective individual circumstances. What
matters more at the individual level are dierences in terms of perceived freedom of decision at
work. Not only do subjective perceptions explain some of the job satisfaction gap, although at
both dates the coecient on the transition country dummy is still signicant when only freedom of
decision at work is controlled for in column (5), but the R-squared values are bigger than when any
other set of variables is included among the explanatory variables. This conrms that subjective
perceptions generally have a strong impact in determining one's level of job satisfaction.
Table 2.11: Ordinary least squares regressions of job satisfaction on region, 1999
Subjective
Objective conditions perceptions
(1) (2) (3) (4) (5)
Coe. Coe. Coe. Coe. Coe.
Variable (t-stat) (t-stat) (t-stat) (t-stat) (t-stat)
Transition country -1.022** -0.368* -0.502** -0.065 -0.534**
dummy (-6.15) (-2.49) (-2.95) (-0.21) (-3.52)
Observations 12,147 12,147 11,880 9,877 12,040
R-squared 0.050 0.059 0.072 0.088 0.210
Country level controls no yes yes yes no
Individual level controls no no yes yes no
Occupation and income no no no yes no
Freedom of decision no no no no yes
Number of countries 19 19 19 19 19
Reference group: Western Europe. Country level controls include ln GDP per capita,
unemployment rate, in
ation. Individual level controls include gender, age group,
age completed education, marital status, employment level union membership.
T-statistics are adjusted for clustering at country level.
Signicance levels: ** p<0.01, * p<0.05, + p<0.10.
What role do these same objective and subjective circumstances play in explaining the job
satisfaction trend in Eastern Europe? I attempt to answer this question in panel A of table 2.12.
35
The reference year is 1990, so the coecient on the dummy variable for 1999 re
ects the change
between the two dates. Because country xed-eects are used, the coecients in column (1) are
slightly dierent from the mean dierences shown in table 2.5. However, the same downward trend
is found as in table 2.5. The various potential determinants are added in the same order as in the
previous two tables. Accounting for macroeconomic conditions in column (2) makes the downward
trend even more pronounced. Indeed, as gures 2.1 and 2.2 showed, after the economic collapse
that quickly followed the onset of the transition process, by 1999 the situation had improved
considerably. The much lower coecient in column (2) (-0.758) compared with that in column
(1) (-0.138) re
ects the fact that had it not been for the positive macroeconomic developments
in transition countries after the initial collapse, the job satisfaction downward trend would have
been even more pronounced. A similar impact, although of a lower magnitude, is found for the
individual level circumstances, when occupation and income are included in column (4). When
the eect of the other individual level characteristics is isolated in column (3), the coecient is
no longer signicant, meaning that these mostly demographic variables account for the lower job
satisfaction in 1999 compared with 1990. Freedom of decision also makes the time coecient
virtually zero { column (5). This could mean that people in transition countries might still
feel unsure about the new system and this explains why the recovery in satisfaction with work
lags behind the economic recovery in Eastern Europe. What else can explain the failure of job
satisfaction in transition countries to recover commensurately with the economic recovery? If
younger cohort are in fact intrinsically less satised than older ones, as J urges (2003) found in
Germany, then this could be an explanation. Indeed, adding controls for year of birth in the form
of dummy variables for each separate year in panel B of table 2.12 makes the time coecient no
longer signicant in column (1), when no other controls are used.
36
Table 2.12: Country xed-eects regressions of job satisfaction on time, 1990-1999, transition
countries
Subjective
Objective conditions perceptions
(1) (2) (3) (4) (5)
Coe. Coe. Coe. Coe. Coe.
Variable (t-stat) (t-stat) (t-stat) (t-stat) (t-stat)
Panel A: Without year of birth controls
Year 1999 -0.138* -0.758* -0.096 -0.243** -0.007
(-2.55) (-2.30) (-1.48) (-2.94) (-0.14)
Observations 8,800 8,800 7,546 5,651 8,606
R-squared 0.019 0.021 0.026 0.054 0.166
Panel B: With year of birth controls
Year 1999 -0.020 -0.628+ 0.114 -0.061 0.070
(-0.34) (-1.90) (1.15) (-0.51) (1.28)
Observations 8,800 8,800 7,546 5,651 8,606
R-squared 0.023 0.025 0.027 0.055 0.168
Country level controls no yes no no no
Individual level controls no no yes yes no
Occupation and income no no no yes no
Freedom of decision no no no no yes
Number of countries 6 6 6 6 6
Reference year: 1990. Country level controls include ln GDP per capita, unemployment
rate, in
ation. Individual level controls include gender, age group, age completed
education, marital status, employment level, union membership.
Signicance levels: ** p<0.01, * p<0.05, + p<0.10.
The cohort eect becomes even more obvious in table 2.13 when I divide the sample in six
separate birth cohorts and I look at their mean job satisfaction in 1990 and 1999. In both Eastern
and Western Europe there is a fairly clear succession of less satised cohorts. If I only look at
the middle four birth cohorts which are surveyed both in 1990 and in 1999, the decrease in job
satisfaction in Eastern Europe is no longer statistically signicant, while in Western Europe there
is an increase, although not a statistically signicant one. What drives the overall decrease in the
East and causes a lack of change in the West is the replacement of the most satised cohort with
the least satised, youngest, one between the two dates.
To sum up, during the rst decade of transition, dierences in macroeconomic conditions play
quite an important role in explaining the dierence in the level of job satisfaction between transi-
tion and non-transition countries, but the role of subjective perceptions becomes more important
37
Table 2.13: Job satisfaction means and standard errors by year of birth cohort, 1990 and 1999
Eastern Europe Western Europe
1990 1999 1990 1999
Mean Mean Mean Mean
Birth year (st. error) (st. error) (st. error) (st. error)
1924-1933 7.341 7.897
(0.12) (0.07)
1934-1943 7.022 6.690 7.765 7.886
(0.07) (0.16) (0.05) (0.07)
1944-1953 6.674 6.549 7.667 7.665
(0.06) (0.08) (0.04) (0.04)
1954-1963 6.635 6.388 7.363 7.537
(0.06) (0.07) (0.04) (0.04)
1963-1973 6.377 6.733 7.214 7.480
(0.09) (0.08) (0.05) (0.04)
1974-1981 6.369 7.369
(0.12) (0.06)
over time as well. The fact that job satisfaction in Eastern Europe in 1999 is still signicantly
lower than in 1990 is mainly the results of cohort replacement with \newer" cohorts intrinsically
less satised than \older" ones, which cancels the positive eect of the economic recovery. Over
time, however, the relative job satisfaction of young people improves, especially in transition
countries, as the interactions of age with time showed. Even so, in Western Europe people under
30 remain the least satised group. The positive eect of education, white collar occupation,
and income also increases in Eastern Europe, pointing to young and more educated people as the
winners of the rst decade of transition.
2.3.2 Job Satisfaction in the New Millennium
The main downside of the WVS is that there is no job satisfaction information beyond 1999.
Therefore, I use the ISSP \Work Orientations" survey to extend the analysis to the 1997 to 2005
time interval. The scale for the job satisfaction question is dierent, being one to seven instead
of one to ten as in the WVS. There is some overlap with the time coverage of the WVS, but
it is short enough to allow the ISSP analysis to be interpreted as an extrapolation of the WVS
data had it extended to 2005. One way to ensure this is to look at the one transition country {
38
Hungary { and three Western European countries { Great Britain, Netherlands, and Italy { which
were included in both the 1989 and the 1997 ISSP surveys. In terms of mean job satisfaction,
I nd a similar trend with the WVS: decrease in the transition country, although not quite as
dramatic as in the WVS, and overall lack of change in the Western countries. I will focus my
analysis from now on on those countries that were surveyed in 1997 and 2005: Bulgaria, Czech
Republic, Hungary, the Russian Federation, Slovenia, Great Britain, Denmark, Portugal, Spain,
and Sweden. They are dierent from the countries studied in the WVS, but Figures 2.1 and 2.2
show enough macroeconomic similarity to make the analysis comparable. The same steps are
followed as in the analysis of the WVS data.
Table 2.14: Mean job satisfaction by region and gender, 1997 and 2005
Eastern Western Dierence
Europe Europe (1)-(2)
(1) (2) (3)
All
1997 4.92 5.31 -0.39**
(0.03) (0.02) (0.03)
2005 5.06 5.31 -0.25**
(0.03) (0.02) (0.03)
Dierence 0.13** 0.00 0.13**
(0.04) (0.03) (0.05)
Women
1997 4.94 5.30 -0.35**
(0.04) (0.03) (0.05)
2005 5.02 5.28 -0.25**
(0.04) (0.03) (0.05)
Dierence 0.08 -0.02 0.10
(0.05) (0.04) (0.07)
Men
1997 4.91 5.32 -0.42**
(0.03) (0.03) (0.05)
2005 5.09 5.34 -0.26**
(0.03) (0.03) (0.04)
Dierence 0.18** 0.02 0.16**
(0.05) (0.04) (0.06)
Signicance levels: ** p<0.01, * p<0.05, + p<0.10.
Table 2.14 shows the mean job satisfaction in Eastern and Western Europe, in 1997 and 2005,
for the overall population, as well as for women and men separately. Column (1) nds a signicant
increase in job satisfaction for the transition countries by 2005, from 4.92 to 5.06. Among the two
39
genders though, the increase is statistically signicant only for men. In Western Europe, as shown
in column (2), there is no signicant change in job satisfaction, neither for the overall population,
nor for any of the two genders separately, the lack of trend therefore continuing between 1997
and 2005. This means that, unlike during the rst decade of transition, there is some convergence
between the two regions by 2005. Although transition countries continue to have lower levels of
job satisfaction than non-transition ones, the dierence in column (3) signicantly decreases from
0.38 in 1997 to 0.25 in 2005. The gap between East and West only narrows signicantly for men,
not for women.
Looking at the changes in job satisfaction at country level (table 2.15), I nd that in Eastern
Europe the increase in job satisfaction is mainly driven by Bulgaria and Hungary, who by 2005
become the transition countries with the highest levels of job satisfaction although they start o
fairly low in 1997. Less changes occur in the other three countries. In Western Europe, three of
the ve countries show the same stability, while two countries, Denmark and Spain, actually show
signicant decreases in job satisfaction. These two countries start o with the highest levels of
satisfaction in 1997, and despite the decrease, Denmark still tops the ranking in 2005.
Table 2.15: Changes in job satisfaction by country, 1997-2005
(1) (2) (3) (4)
Eastern Europe Western Europe
Hungary (+)** (+)** Great Britain 0 0
Czech R. (-)+ 0 Sweden 0 0
Slovenia 0 0 Spain (-) * (-) *
Bulgaria (+)** (+)* Portugal 0 0
Russia 0 0 Denmark (-) ** (-) **
The columns are based on the signs of OLS coecients. Columns (1)
and (3) use no controls. Columns (3) and (4) control for gender, age,
and education. Signicance based on robust standard errors.
Signicance levels: ** p<0.01, * p<0.05, + p<0.10.
The general determinants of job satisfaction in 1997 and 2005 are similar to what I found in the
WVS. For the additional variable of sector, people employed in the public sector on average have
higher levels of job satisfaction than those working in the private sector. The various subjective
40
Table 2.16: Ordinary least squares regressions of job satisfaction with interactions between the
transition country dummy and specied independent variables, 1997-2005
Objective conditions Subjective perceptions
(1) (2) (3) (4) (5)
Coe. Coe. Coe. Coe. Coe.
Variable (t-stat) (t-stat) (t-stat) (t-stat) (t-stat)
Transition*high school 0.125
(1.19)
Transition*university 0.294*
(2.60)
Transition*relative income 0.064+
(1.88)
Transition*interesting job -0.091+
(-2.00)
Transition*independent work -0.071*
(-2.89)
Transition*dangerous conditions -0.041*
(-2.32)
Observations 12,002 8,627 8,108 8,108 8,108
R-squared 0.038 0.066 0.308 0.307 0.307
The regression controls include year, ln GDP per capita, unemployment, in
ation, gender,
age group, marital status, employment level, union membership. In columns (2)-(5) sector,
occupation, and income quintile are also controlled for. In columns (4) and (5), job security,
pay, promotion, help others, useful to society, exhausted after work, physical work, stressful
work are also controlled for. Signicance levels: ** p<0.01, * p<0.05, + p<0.10.
variables in the ISSP tend to have the positive eect that one expects. The impact of having a
high pay and an interesting job are particularly strong. Also very important for satisfaction is to
have a job that is not very stressful. Table 2.16 shows the objective and subjective variables whose
impact signicantly diers between transition and non-transition countries. The importance of
income in determining job satisfaction continues to be bigger in transition countries { column
(2), with the education premium also being higher in Eastern Europe { column (1). Subjective
perceptions { how interesting one's job is, how much independent work it allows for, and how
dangerous it is { tend to matter less in transition than in non-transition countries. In general,
it appears that people in transition countries focus more on objective conditions, while in non-
transition countries subjective perceptions have a stronger impact on satisfaction with work.
Overall though, compared with the 1990s, the results are indicative of a convergence between the
41
two regions not just in terms of the level of job satisfaction, but in terms of its determinants as
well, especially the objective ones.
Table 2.17: Ordinary least squares regressions of job satisfaction on region, 1997
Subjective
Objective conditions perceptions
(1) (2) (3) (4) (5)
Coe. Coe. Coe. Coe. Coe.
Variable (t-stat) (t-stat) (t-stat) (t-stat) (t-stat)
Transition country -0.388** -0.169 -0.193 -0.322 -0.172
dummy (-3.53) (-1.49) (-1.69) (-1.49) (-1.79)
Observations 6,116 6,116 5,284 3,645 5,631
R-squared 0.025 0.029 0.047 0.079 0.287
Country level controls no yes yes yes no
Individual level controls no no yes yes no
Sector, occupation, and income no no no yes no
Job characteristics no no no no yes
Number of countries 10 10 10 10 10
Reference group: Western Europe. Country level controls include ln GDP per capita,
unemployment rate, in
ation. Individual level controls include gender, age group,
age completed education, marital status, employment level, union membership. Job
characteristics include job security, pay, promotion, interesting job, independent
work, help others, useful to society, exhausted after work, physical work, stressful
work, dangerous conditions. T-statistics are adjusted for clustering at country level.
Signicance levels: ** p<0.01, * p<0.05, + p<0.10.
The ability of these various circumstances to explain the dierence in job satisfaction between
Eastern and Western Europe is tested next. For 1997, this is done in table 2.17. The coecient
on the transition country dummy is no longer signicantly negative after controlling for objective
macroeconomic conditions in column (2). This means that these variables account for the gap in
job satisfaction observed in 1997. In table 2.18, however, the coecient is still as high as without
any controls, so the macroeconomic conditions are no longer accountable for the gap observed in
2005. This is in line with the diminishing role of macroeconomic conditions in explaining satis-
faction with work dierences that I found in the WVS between 1990 (table 2.10) and 1999 (table
2.11). The objective variables at the individual level that I consider, including sector, occupation,
and income, have even less explanatory power, as columns (3) and (4) show. The variables that
seem to be the reason behind the persisting gap are those re
ecting the respondents' subjective
42
appraisal of their job characteristics, with the coecient in column (5) no longer signicantly dif-
ferent from zero, neither in 1997 nor in 2005. As was the case for freedom of decision at work in the
WVS, the cumulative power of the subjective variables in the ISSP towards explaining variations
in individual job satisfaction is quite high, as re
ected by relatively high R-square values.
Table 2.18: Ordinary least squares regressions of job satisfaction on region, 2005
Subjective
Objective conditions perceptions
(1) (2) (3) (4) (5)
Coe. Coe. Coe. Coe. Coe.
Variable (t-stat) (t-stat) (t-stat) (t-stat) (t-stat)
Transition country -0.255* -0.308* -0.298** -0.389** -0.071
dummy (-3.09) (-3.24) (-5.80) (-5.59) (-0.93)
Observations 6,970 6,970 6,718 4,982 6,537
R-squared 0.011 0.015 0.032 0.061 0.292
Country level controls no yes yes yes no
Individual level controls no no yes yes no
Sector, occupation, and income no no no yes no
Job characteristics no no no no yes
Number of countries 10 10 10 10 10
Reference group: Western Europe. Country level controls include ln GDP per capita,
unemployment rate, in
ation. Individual level controls include gender, age group,
age completed education, marital status, employment level, union membership. Job
characteristics include job security, pay, promotion, interesting job, independent
work, help others, useful to society, exhausted after work, physical work, stressful
work, dangerous conditions. T-statistics are adjusted for clustering at country level.
Signicance levels: ** p<0.01, * p<0.05, + p<0.10.
Table 2.19 identies the determinants of the upward trend in job satisfaction in Eastern Europe.
In panel A, controlling for macroeconomic variables in column (2) makes the coecient no longer
signicant. This shows that by 2005 the eect of the economic recovery in Eastern Europe are
also felt in terms of job satisfaction. The role of objective individual level variables in driving this
upward trend is quite low, because their addition among the controls in columns (3), (4) does not
make a big dierence. The impact of subjective variables is also fairly low { column (5). Adding
controls for year of birth in panel B does not make a big dierence during this time interval.
Indeed, in Eastern Europe there is no longer a big dierence in the job satisfaction of the \old"
and the \new" cohort, in 1997 and 2005 (table 2.20). This is probably the result of the relative
43
improvement in the satisfaction of younger people compared with older ones that was already
underway before 1999. It is the job satisfaction of cohorts included at both dates that drives the
increase in transition countries. In Western Europe, the cohort eect persists but it is canceled
out by the slight increase observed for the four middle cohorts.
Table 2.19: Country xed-eects regressions of job satisfaction on time, 1997-2005, transition
countries
Subjective
Objective conditions perceptions
(1) (2) (3) (4) (5)
Coe. Coe. Coe. Coe. Coe.
Variable (t-stat) (t-stat) (t-stat) (t-stat) (t-stat)
Panel A: Without year of birth controls
Year 2005 0.124** -0.324 -0.423 -0.530 0.094**
(3.45) (-0.85) (-1.10) (-1.09) (2.84)
Observations 5,921 5,921 5,766 4,233 5,390
R-squared 0.007 0.008 0.029 0.052 0.257
Panel B: With year of birth controls
Year 2005 0.160** -0.341 -0.407 -0.525 0.143**
(4.26) (-0.90) (-1.06) (-1.08) (4.15)
Observations 5,921 5,921 5,766 4,233 5,390
R-squared 0.009 0.010 0.029 0.052 0.261
Country level controls no yes yes yes no
Individual level controls no no yes yes no
Sector, occupation, and income no no no yes no
Job characteristics no no no no yes
Countries East East East East East
Number of countries 5 5 5 5 5
Reference year: 1997. Country level controls include ln GDP per capita, unemployment
rate, in
ation. Individual level controls include gender, age group, age completed
education, marital status, employment level, union membership. Job characteristics
include job security, pay, promotion, interesting job, independent work, help others,
useful to society, exhausted after work, physical work, stressful work, dangerous
conditions. T-statistics are adjusted for clustering at country level.
Signicance levels: ** p<0.01, * p<0.05, + p<0.10.
Because the data come from two dierent datasets, it is dicult to assess how job satisfaction in
2005 compares with its level in 1990, at the onset of the transition. However, if I only consider the
seven countries that were included in both the WVS and the ISSP { Hungary, Russia, Denmark,
Great Britain, Sweden, Spain, and Portugal { it appears that the recovery in job satisfaction
has been achieved in Eastern Europe because mean satisfaction in 2005 is higher than in 1990.
44
Table 2.20: Job satisfaction means and standard errors by year of birth cohort, 1997 and 2005
Eastern Europe Western Europe
1997 2005 1997 2005
Mean Mean Mean Mean
Birth year (st. error) (st. error) (st. error) (st. error)
1932-1939 5.070 5.449
(0.15) (0.07)
1940-1949 4.995 5.317 5.387 5.498
(0.05) (0.07) (0.04) (0.04)
1950-1959 4.973 5.027 5.299 5.361
(0.04) (0.05) (0.04) (0.04)
1960-1969 4.895 5.066 5.249 5.282
(0.04) (0.04) (0.04) (0.03)
1970-1979 4.771 4.973 5.260 5.249
(0.06) (0.04) (0.06) (0.04)
1980-1987 5.089 5.118
(0.07) (0.07)
However, with data coming from two dierent datasets, with two dierent answer scales, making
any inferences about statistical signicance is hazardous.
2.4 Summary and Conclusions
In 1990, soon after the beginning of the transition process in Eastern Europe, job satisfaction was
signicantly lower than in the West, but a quick convergence was commonly expected as people
in the East were adopting the \Western" lifestyle. However, as the WVS data show, transition
countries still have lower levels of job satisfaction than Western European countries in 1999. In
fact, the dierence between the two regions becomes even bigger due to a signicant decrease in
satisfaction with work in Eastern Europe between 1990 and 1999 combined with a lack of change
in the West. The same patterns are found for the overall population, as well as for men and women
separately. The ISSP \Work Orientations" surveys of 1997 and 2005 allow me to extend the job
satisfaction analysis into the new millennium. I thus nd a signicant increase in satisfaction
with work in transition countries between these two dates, while in Western Europe the lack of
45
a signicant trend persists. Therefore, there is some convergence between the two regions during
this time interval, but Eastern Europe is still below the West at both dates.
The determinants of job satisfaction are quite similar in Eastern and Western Europe, but the
impact of subjective variables on job satisfaction is stronger in the West, while objective condi-
tions, such as income, matter more in the East. The fact that satisfaction with work in transition
countries is lower is mainly the result of dierences in macroeconomic conditions between the
two regions. Indeed, the former communist countries start o at a lower economic level than the
West and go through a dramatic collapse in the rst few years of the transition. Despite the
swift recovery, they are not quite back to the pre-transition level in 1999 and still considerably
behind Western Europe. Among the individual level variables, occupation and perceived relative
income also account for some of the dierence in job satisfaction between the two regions, but the
impact of dierences in perceived freedom of decision at work is actually slightly stronger. In fact,
the inability to close the gap with the West by 2005 is also due to the way people in transition
countries subjectively perceive their job characteristics.
In Western Europe, job satisfaction does not change signicantly between 1990 and 2005. In
Eastern Europe, there is a decline, followed by recovery. The decline is mostly driven by cohort
eects { younger cohorts are intrinsically less satised than older ones, all things equal, so as they
enter the labor market, this has a negative eect on satisfaction with work, even when objective
conditions improve. The transition, however, appears to benet young people more than older
ones. The young are better adapted to the new market conditions, which is consistent with the
ndings of Alesina and Fuchs-Sch undeln (2007), and this allows them to improve their relative
position compared with older people more than in the West. The continuous improvement in
macroeconomic conditions counteracts this cohort eect in Eastern Europe between 1997 and
2005, one of the reasons why they show an increase in job satisfaction in this time interval. At the
same time, I nd a deterioration in transition countries in the relative job satisfaction of people
between 40 and 49 years old compared with the other age groups. It is beyond the scope of this
46
paper to nd the reasons behind this deterioration, but one can speculate that people in this
age group were already well-embarked on a life course set under the conditions of socialism, they
had trained to function in this system, and were left in turmoil when the political and economic
system changed. This is made worse by the fact that many of these people are at a peak in
parental obligations generating a considerable nancial burden which could explain why they do
not consider their jobs satisfying.
Using two dierent datasets, with two dierent scales, makes it impossible to assess with
certainty whether job satisfaction in Eastern Europe is back to pre-transition levels. The results,
however, indicate that the path to recovery has started and that convergence with the West is
possible.
47
Chapter 3
Life Satisfaction and the Economic Transition in Poland
3.1 Introduction
This chapter studies subjective well-being and its determinants in Poland during the process of
transition. By subjective well-being I mean people's self-reported evaluation of their level of
happiness through survey answers. By transition I mean the process through which the countries
of Central and Eastern Europe switched from a planned economy to a Western free market model.
By the end of the 1980s, the centrally planned economies had reached their limits and were losing
their ability to expand. Serious nancial, economic, social, and political imbalances were becoming
apparent in the form of vast shortages, in
ationary pressure, social needs that could no longer be
satised. Not surprisingly, as soon as the political situation permitted it at the end of the 1980s
and the beginning of the 1990s, the socialist countries started getting rid of the old system, no
longer adaptable to the challenges of the world economy, and chose to install a market economy
(Ko lodko, 2000).
The expectation among the citizens of Central and Eastern Europe was that the adoption
of a democratic political system and the transition to a market economy would lead to material
prosperity, and that the new system would become a reality sooner rather than later (Ko lodko,
48
2000; Zuzowski, 1998). This initial euphoria was soon replaced by the realization that the transi-
tion process would not happen overnight. Establishing the economic foundations required by the
new system would bring considerable economic hardship through declining industrial production,
increasing unemployment, high in
ation, and decreasing real wages and salaries (EBRD, 1998;
Hayo and Seifert, 2003).
Economists have extensively studied the transition process, assessing its success or failure
based on various economic indicators, GDP per capita in particular. However, such indicators
do not tell the whole story. Of great importance are the perceptions of people themselves on
how satised they are with their lives and the way these perceptions change over time as reforms
are being implemented more or less successfully. Furthermore, especially when reliable objective
data are hard to nd, the subjective measures of well-being can provide a useful complement to
conventional economic data.
In this study I look precisely at the impact of the transition on people's self-reported levels of
well-being, as a measure of its success. My focus is on Poland, a country which was very quickly
identied as the \leading economic performer in the region" (EBRD, 1997, p.7). In Poland,
attempts to reform the socialist system had already been made in the early 1980s. While these
reforms made a positive contribution to the transition, from a historical perspective they were
a disappointment. As a result, a fundamental policy switch occurred in September 1989, from
a somewhat reformed socialist planned economy to a genuine market economy, Poland being
the rst Central and East European country to start the transition (Ko lodko, 2000, 1997). The
contribution of this study is to see to what extent a successful transition by economic standards
also translates into a successful transition in terms of people's happiness. Economists assume the
two go hand in hand. My results conrm that such a relationship exists, but other factors { such
as changes in people's marital and employment status, or the replacement of cohorts { also play
a role in the recovery in terms of subjective well-being.
49
3.1.1 The transition in Poland
Even before the demise of communism in 1989, Poland was quite far from the typical idea of
a centrally planned economy, with reform eorts more serious than elsewhere. Following social
unrest, the early 1970s were characterized by very fast growth. But this growth was unbalanced
and unsustainable, and as a consequence it quickly ceased and was actually negative between
1979 and 1982. Furthermore, the formation of large enterprises engaged in wasteful investment
projects nanced by foreign borrowing led to a debt burden that was strangling the economy.
Public opinion often sent contradictory signals. While regularly claiming a preference towards
price levels and structures based on market principles, people also protested loudly when the
authorities raised prices signicantly. In fact, it was another eort to raise prices in 1980 that led
to the formation of the Solidarity trade union, led by Lech Wa l esa. The potential for political
opposition against the communist regime increased as an alliance emerged between the workers'
movement, the Catholic Church, and the intellectual counter elite. The union was outlawed at
the end of 1981 and martial law followed in 1982-1983. Nonetheless, the reform eorts continued,
although the nancial resources for a profound reform did not exist, especially given the extreme
foreign debt burden. Indeed, signicant progress was made before 1989: central planning was
replace by \government contracts"; a single Ministry of Industry was formed; wage payments
were liberalized; the private sector was expanded to manufacturing, in addition to agriculture,
which had been de-collectivized since the late 1950s; a two-tier banking system, with a fairly
independent central bank was established (Ko lodko, 1997). However, these reforms were only
an attempt to adapt the socialist economic system to new circumstances, not to overturn it.
As a result, the usual deciencies of a communist regime also persisted: gigantomania in the
heavy industrial sector protected from international competition; underdeveloped light industrial,
and service and distribution sectors; shortages and poor-quality consumer goods; an inecient
50
nancial sector; a backward agricultural sector, despite being largely capitalist; very large foreign
debt (Rosser and Rosser, 2004; Sachs, 1992).
The gradual erosion of communism nally came to fruition in August 1989 when a coalition
government was established. Although communists still participated, non-communists were in
charge, and Leszech Balcerowicz was appointed nance minister. Poland became the rst East
European country to shrug o communism, and at the end of 1990 Lech Wa l esa was elected
president (Rosser and Rosser, 2004). Once the common opponent disappeared, the three pillars
of the Solidarity movement { workers, intellectuals, and the clergy { developed quite dierent
interests, and the Polish intelligentsia became increasingly fragmented along party lines (Anheier
and Seibel, 1998). In fact, the fall of communism marked the \decapitation through success" of the
Polish civil society, a civil society once strong enough to precipitate the collapse of a communist
regime (Bernhard, 1996).
Poland's market transition was implemented in January 1990 through the Balcerowicz plan,
and it took the form of a shock therapy or big-bang reform. Given the accelerating hyperin
a-
tion of 1989, macroeconomic stabilization was the highest priority. Economic liberalization and
privatization were the other types of policies to be implemented. The key provisions aimed at
achieving these objectives were: immediate decontrol of most prices; sharp devaluation of the
Polish currency to a xed exchange with the U.S. dollar, and removal of all foreign exchange
controls; a tight monetary policy signaled by a tripling of the discount rate of the Polish National
Bank; legalization of all forms of private enterprise and privatization of state owned enterprises.
Despite being a priority of the reform, privatization was not fully implemented because of political
backlash (Rosser and Rosser, 2004).
Poland's shock therapy approach had both positive and negative consequences. In
ation was
in the triple digits in 1989 and 1990, leading to a sharp increase in prices. However, it rapidly
declined to a much lower level, and since 2001 it has constantly been in the single digits. Output
declined sharply as a result of reduced demand following falling real wages, rising costs due to
51
higher taxes, and higher cost of credit and a reduction in its availability. The initial output decline
was also overstated in ocial data because activities in the private sector were not yet recorded
(Rosser and Rosser, 2004). Real GDP declined 11.6 percent in 1990 and 7.3 percent in 1991, but
economic growth resumed in 1992 and has continued ever since (Human and Johnson, 2002).
This made Poland the rst transition country to bottom out and begin growing again, and by 1996
the only country to surpass its 1989 pre-reform level of real GDP per capita. A consequence of the
decrease in output was a sharp increase in unemployment. From no open unemployment during
communism, it quickly got into the double digits by 1991. It remains higher than in neighboring
countries and continues to be a problem, staying mostly in the double digits through 2009. In
general, Poland's shock therapy had more severe side eects than the more gradual approach of
neighboring transition economies, but the recovery also started much faster.
Once Balcerowicz left oce in 1991, a more gradual approach to the transition followed. In
September 1993, a leftist coalition took power and adopted the \Strategy for Poland". Privatiza-
tion and strict monetary policies continued to be upheld, but it was emphasized that the reforms
in the state sector had to be balanced by reforms in the private sector. Fairly generous social
safety nets were maintained. The new regime claimed that this more gradual approach reduced
the transition costs, without jeopardizing its success.
The big bang versus gradualism debate is far from settled. They both have the same nal
purpose and concern to minimize disruption and avoid political opposition to reforms. However,
while the proponents of a shock therapy advocate sizing the moment of political opportunity and
implementing major changes as quickly as possible (Balcerowicz, 1993; Klaus, 1995; Sachs, 1994),
others caution that a more gradual approach will minimize disruption, output and job losses, and
prevent a consequent reversal of political will (Aghion and Blanchard, 1994; Dewatripont and
Roland, 1992). In general, theoretical analysis tends to conclude that gradualism is better, while
empirical analysis nds that early and quick reforms result in an earlier and stronger recovery,
while actually minimizing costs (Havrylyshyn, 2006).
52
Probably the biggest challenge of the Polish transition has been the rampant unemployment.
The socialist emphasis on employment as a universal entitlement led to a very high proportion of
economically active people in the working age population by international standards. This dier-
ence came especially as a result of a very high participation rate for women. At the same time,
the state-owned enterprises that dominated the economy employed more people than actually
needed. This labor hoarding made the activity rate to be more or less equivalent to the employ-
ment rate, so that there would be no open unemployment. The communist labor market was also
characterized by low wage rates, with low dierentials across skill levels, and salary expenditures
at enterprise level not constrained by revenues, but rather re
ecting policy priorities and the bar-
gaining strength of the respective enterprise (Mickiewicz and Bell, 2000). It is not surprising then
that the transition to a market economy, which removed administrative controls, also resulted in
the emergence of unemployment. The state of being unemployed was fully recognized through
the adoption of the Employment Law in December 1989, which stipulated that unemployment
compensation was no longer a discretionary benet (Maret and Schwartz, 1994). Among the
transition countries, Poland was one of the most aected by unemployment. As mentioned above,
the unemployment rate was already in the double digits in 1991 and stayed so for most of the
past two decades. In the rst few years of the transition, however, the numbers for registered un-
employment may have been in
ated by the fact that very loose eligibility criteria enabled people
who intended to withdraw from the labor market to remain registered as unemployed and claim
benets. This was driven not just by the cash value of the benets but by the desire to retain
social insurance benets, particularly free health care, and to continue to accumulate pension
rights. This was only a short-run phenomenon, and activity rates started to fall by 1992-1993,
once the unemployment benet system was reformed and stricter rules were put in place.
The increase in unemployment was not the only source of increase in social expenditures. A
steep increase in pension expenditures occurred during the early 1990s. Various regulations that
allowed early retirement to those made redundant in their jobs, or later on to those aected by
53
long-term unemployment, led to a massive increase in benet numbers. In 1991, the number of
pensioners was ve times the rate in the previous year (World Bank, 2003). Early retirement
is detrimental to the pension system's nances not only because with more beneciaries, the
expenditures increase, but because this also means that there are fewer people paying contributions
to the system. The rules governing disability pensions were also very lax, so many disability
pensioners continued gainful employment, while receiving disability pension as extra income.
However, the most important factor behind high pension expenditures during the rst few post-
communist years was the automatic indexation of benets based on increases in wage. Over time,
stricter rules were implemented in order to reduce expenditures, such as stricter eligibility criteria
and the indexation based on prices for pensioners' consumer basket (Czepulis-Rutkowska, 1999;
Rapacki, 2001).
Given these numerous challenges that people were faced with, \the sudden switch to a market
economy [. . . ] aroused profound anxieties, as most households [. . . ] wondered whether they
would be able to stay a
oat - much less prosper - in the new system" (Sachs, 1990). It is not
surprising then that the transition took its toll on people's health. Looking specically at mass
privatization programs during the transition, Stuckler et al. (2009) found that they were associated
with a short-term increase in mortality rates in working-age men, with unemployment rates as a
mediating factor. Indeed, during the rst couple of years of the transition, male life expectancy
at birth in Poland declined by almost one year, a decrease driven by men aged 20-59 years and
due mainly to external causes and circulatory disease. By 1991, the dierence in female and
male life expectancy at birth reached 9.2 years. Since then, male mortality has been showing
a steady recovery, largely attributable to falling mortality among men aged 40-64 (Nolte et al.,
2000). A phenomenon closely related to the mortality changes during the transition is alcohol
consumption. A jump in alcohol consumption occurred in the rst few years of transition, linked to
the dramatic transformations towards the market economy. On the supply side, de-monopolization
of production and distribution, rapid liberalization of alcohol control, and practical suspension of
54
its enforcement led to a higher physical availability of alcohol. In about a year, the number of
alcohol outlets in Poland increased from approximately 30,000 in the late 1980s to 150,000 at the
beginning of the 1990s. Relative prices of spirits also decreased in the 1990s to almost half of the
1980s level (Moskalewicz, 2000; Moskalewicz et al., 2000; Wojtyniak et al., 2005). On the demand
side, the challenges of the transition were a signicant source of stress, so it is not surprising that
people would use alcohol as a coping mechanism, especially given its increased availability. Even
as overall mortality started to decline and life expectancy started to increase in Poland after 1992,
mortality due to liver cirrhosis and to causes directly attributable to alcohol leveled o or even
increased.
Changes also occurred in the realm of family life. The fertility rate measured as the total births
per woman considerably decreased between 1989 and 1999 from 2.08 to 1.40. Marriage rates have
been declining over time in most Western countries as well, but the rate of decline accelerated in
most transition economies. In Poland, the marriage rate per 1,000 inhabitants fell from 6.8 in 1989
to 3.6 in 1999. However, while transition appears to have had a strong negative eect on marriage
formation and fertility, the divorce rate per 1,000 inhabitants remained relatively constant at 1.2.
This seems to indicate that transition has not destroyed existing marriages (Svejnar, 2002). I
believe, however, that this is also the result of the decline in marriage rates: as a percentage of
married people, and not of all inhabitants, divorce rates have in fact increased.
A very important aspect of the transition is the dierent impact it had on dierent population
groups, in other words, the existence of both winners and losers. The economic reforms very
quickly aected the hundreds of thousands of farmers working on small and inecient farms, who
lost their production subsidies. At the same time, the prices of goods that they sold increased at
a much slower pace than the price of products bought by peasants, which resulted in a dramatic
decline in their incomes (Gorlach and Mooney, 1998). In general, during the transition people
in rural areas tend to be worse o than people in urban areas and particularly compared with
those in the capital city (Ellman, 1997). For example, at the end of 1998, the Polish wojewodztwo
55
of Warsaw had an ocial unemployment rate of 2.6%, while in the bordering wojewodztwo of
Ciechanow unemployment was six times higher at 15.6% (Barjak, 2001).
New market conditions proved to be dicult to cope with for hundreds of thousands of people
working in industrial enterprises, where a sharp loss of jobs occurred. In general, less skilled
and less educated people are the most likely to be unemployed in post-socialist Poland. The risk
of unemployment is highest for Poles with basic vocational education, followed by those with a
general secondary education. White-collar workers are not only more likely to be employed, but
their relative earnings compared with blue-collar workers also increased after the fall of commu-
nism. This is in sharp contrast with the situation under socialism, characterized by a bias toward
production, blue-collar workers (Rutkowski, 1996). The earning dierential is particularly high in
the private sector compared with the public sector. The rate of joblessness is also higher among
young people. In 1993, about 64 per cent of the registered unemployed were younger than 35
(Kotowska, 1995). When it comes to earnings, however, young people benet from a \vintage
eect" { relatively rapid increases in earning opportunities for the newest labor force entrants.
This is the result of increased returns to schooling, while experience gained under the previous
economic conditions now has a lesser value. Again, the education premium tends to be higher
in the private compared with the public sector (Rutkowski, 1996). Gender wise, women bear a
higher burden of unemployment, have a greater risk of long-term unemployment, and the wage
gap with men is signicant. During communism, the women's participation in the labor force was
driven by economic and ideological reasons. In capitalism, the reasons are purely economic. A
decline in the access to child care facilities and the gap between male and female wages make it
more dicult for women to stay in the labor market (Fodor, 2005; Kotowska, 1995).
Given the substantial costs that the transition generated for important segments of the pop-
ulation, do the Poles still consider the change in regime to have been a positive development?
In a study carried out in 1999 by the Public Opinion Research Center in the Czech Republic,
Hungary, and Poland, the largest majority of individuals feeling that it was worthwhile to change
56
the political and economic system was found in Poland. Interestingly, at the same time, in each
country many more people believe that the losses from transition exceeded the gains than the
reverse, and that their \material conditions of living are now a little worse" (Svejnar, 2002).
There is a growing, yet far from exhaustive, literature dealing with subjective well-being in
transition countries. A number of papers compare life satisfaction in transition economies with
that recorded in non-transition countries (Frey and Stutzer, 2002; Helliwell, 2003; Hayo and Seifert,
2003; Sanfey and Teksoz, 2007). They typically nd that individuals from the former Soviet Union
report the lowest levels of life satisfaction, while Central and Eastern European countries score
higher but still below OECD countries and even below most of the countries in Asia or Central and
South America. In general, if one looks at survey evidence on individual self-reported happiness in
Central and Eastern Europe compared with Western Europe, it seems that the political isolation
of socialism was replaced by an \iron curtain" of unhappiness (Lelkes, 2006a). This is consistent
with my ndings on job satisfaction in the previous chapter. Looking at a number of transition
countries Easterlin (2009) nds that life satisfaction plummets and then recovers following the
course of the economy as indexed by real GDP, but that this recovery falls short of that in GDP.
A substantial methodological literature has established the reliability, validity, and comparability
of the answers to questions regarding life satisfaction (Frey and Stutzer, 2002; Kahneman, Daniel,
Ed Diener, and Norbert Schwarz, eds., 1999; Veenhoven, Ruut, 1993).
For Poland, I nd that the recovery of life satisfaction to its 1989 level took longer than the
recovery in terms of GDP per capita. Improvements at the macroeconomic level are an important
determinant of improvements in subjective well-being. Cohort replacement, however, also plays
an important part. It is the replacement of old generations, whose value system was formed under
communist conditions, by new generations, raised under new market conditions, that allows life
satisfaction in Poland to increase even when the economic situation levels o.
57
I will next describe the data and methodology, followed by my ndings on life satisfaction in
Poland during its transition to capitalism, between 1989 and 2009. The last section concludes the
analysis.
3.2 Data and methods
The data used in the analysis come from two main sources { the World Values Survey (WVS,
2009) and the Eurobarometer (EB)
1
. These are nationally representative surveys conducted at
multiple dates, but neither one is a panel dataset. The WVS is a multi-country survey that
covers people's attitudes toward a broad range of issues, such as economics and politics, family
and religious values, and environmental awareness. So far ve waves of the WVS have been
implemented: 1981-1984, 1989-1993, 1994-1999, 1999-2004, and 2005. For Poland, surveys were
carried out starting in the second wave { wave 2 consists of data gathered in 1989 and 1990, wave
3 was collected in 1997, wave 4 in 1999, and wave 5 in 2005. Because wave 3 does not include
any question on the employment status of the respondents, it will be omitted from most of the
analysis. In Poland, the EB surveys including life satisfaction among the questions were rst
part of the Candidate Countries Eurobarometer, between 2001 and 2003, and then part of the
Standard Eurobarometer, once Poland joint the European Union in 2004.
The WVS spans a long time interval (1989-2005), and it also has the advantage of providing
a glimpse into the Polish society right around the time when communism fell in 1989. I use the
1989-1990 surveys as a rough re
ection of pre-transition conditions and therefore a benchmark
against which later developments are judged. Is mean satisfaction in these early surveys really a
good approximation of life satisfaction under socialism? Easterlin (2009) argues that it is probably
lower than life satisfaction in the 1980s. In fact, social anomy is considered one of the fundamental
features of the Polish society throughout the 1980s. While many of the elements favoring anomy
1
The Eurobarometer data was downloaded at http://zacat.gesis.org/.
58
in the 1980s vanished with the fall of communism, other, new factors emerged { as early as 1988-
1989, before unemployment was even ocially recognized, people started realizing that it was very
likely that they would lose their jobs. The rapid deterioration of the material situation favors
anomy as well, not only through a decline in living standards, but also because a lot of eort and
stress come with the attempt to protect a falling living standard (Kolarska-Bobi nska, 1990). If
the life satisfaction in Poland corresponding to wave 2 of the WVS is indeed under the peak value
reached under socialism, then comparisons with the pre-transition situation as approximated by
the wave 2 survey will probably favor the post-1989 developments. An important downside of the
WVS data is that only a few surveys were carried out. The 1989 and 1990 surveys were both part
of wave 2 and will therefore be pooled together in the analysis. The wave 3 survey conducted in
1997 has no information on the employment status of the respondent, one of main variables of
interest, so it will be excluded from the analysis. This leaves three surveys to be analyzed for the
WVS: wave 2 (1989-1990), wave 4 (1999), and wave 5 (2005). Starting with 2001, the EB provides
data on life satisfaction in Poland on at least an annual basis until as recently as 2009. I will use
these data for a more detailed analysis of life satisfaction in Poland in the new millennium.
In the WVS life satisfaction is measured through the question \All things considered, how
satised are you with your life as a whole these days?", with answers on a scale from 1 (most
dissatised) to 10 (most satised). In the EB life satisfaction is assessed on a one to four scale.
The question is \On the whole, how satised are you with your life in general?" and the answer
categories, recoded so that higher numbers indicate a greater satisfaction, are not at all satised/
not very satised/ fairly satised/ very satised. In both surveys, each individual also provides
information on gender, age, education, marital and employment status, and occupation, which will
be used in the analysis. In the WVS there is also a question on income brackets, the respondents
being asked to rank their income on a step scale from 1 to 10, with 10 being the highest. Therefore,
this is more of a perceived income classication than a true income scale. A detailed description
of these variables and the way they are coded is provided in Appendix B.1.
59
Table 3.1: Mean life satisfaction by dataset and survey date, on original scale and on 1-10 scale,
1989-2009
Survey date Life satisfaction Year for
Year Month Mean Standard Mean macroeconomic
error (1-10) data
WVS: original scale 1-10
1989 11 6.639 (0.08) 6.639 1989
1990 05 6.531 (0.07) 6.531 1989.5
1997 10 6.421 (0.08) 6.421 1997
1999 02 6.374 (0.07) 6.374 1998.5
2005 12 7.023 (0.00) 7.023 2005
EB: original scale 1-4
2001 10 2.638 (0.03) 5.914 2001
2002 09 2.703 (0.03) 6.109 2002
2003 05 2.743 (0.03) 6.230 2002.5
2003 10 2.679 (0.03) 6.036 2003
2004 10 2.805 (0.02) 6.414 2004
2005 05 2.770 (0.02) 6.310 2004.5
2005 10 2.762 (0.02) 6.285 2005
2006 04 2.798 (0.02) 6.395 2005.5
2006 09 2.832 (0.02) 6.497 2006
2007 04 2.882 (0.02) 6.647 2006.5
2007 10 2.890 (0.02) 6.671 2007
2008 04 2.836 (0.02) 6.507 2007.5
2008 10 2.834 (0.02) 6.503 2008
2009 02 2.781 (0.02) 6.344 2008.5
2009 06 2.929 (0.02) 6.786 2008.5
2009 07 2.906 (0.02) 6.717 2009
CBOS: original scale 1-5
1992 2.651 4.714 1992
1993 2.677 4.773 1993
1994 2.761 4.962 1994
1995 2.841 5.143 1995
1996 2.939 5.362 1996
1997 2.997 5.493 1997
1998 2.973 5.440 1998
1999 2.906 5.288 1999
2000 2.868 5.202 2000
2001 2.855 5.173 2001
2002 2.874 5.217 2002
2003 2.915 5.309 2003
2004 2.958 5.405 2004
2005 2.978 5.451 2005
60
In order to get a clearer picture of the life satisfaction trend over time, I also look at data from
the Public Opinion Research Center in Poland (CBOS). A question on private satisfaction was
included { \How are your life and your family's life?" { and the answer options were very bad/
bad/ neither good nor bad/ good/ very good. Grosfeld and Senik (2008) provide the mean values
for private satisfaction for six surveys per year for the interval 1992-2005 in Table A1 of their
study. Based on these values I compute annual means for life satisfaction. Unlike for the WVS
and the EB, I do not have access to and therefore do not analyze the individual CBOS data.
The values for life satisfaction by survey are listed in table 3.1. Mean life satisfaction is shown
both on the original scale used in each survey and on the 1-10 scale of the WVS. For the EB,
LS
110
= LS
14
3 2. For the CBOS, LS
110
= LS
15
9=4 5=4. The life satisfaction
survey means will be matched with macroeconomic indicators based on the date at which each
survey was conducted. If the survey is carried out in the rst half of the year, then it is likely to
re
ect not only the economic situation of the respective year but also that of the previous year.
Consequently, the corresponding macroeconomic indicators will be an average of the values for
the previous and current year. If the survey is carried out in the second half of the year, the
corresponding indicators will be those for the current year only. The years matched with each life
satisfaction survey date are also shown in table 3.1.
The macroeconomic indicators that I use in order to follow the economic trends in Poland are
GDP per capita, the unemployment rate, the in
ation rate, and the Gini coecient for income as
an indicator of inequality. The values for these indicators for the 1989 to 2009 time interval are
shown in table 3.2.
In order to assess the success of the Polish transition, I look at the trend of life satisfaction in
Poland during the rst two decades of transition. My assumption is that this trend is the result of
the various changes that accompanied the process both at the macroeconomic and at the individual
level. I will rst establish the impact that changes in the macroeconomic indicators listed in table
3.2 have on life satisfaction. To this end, I pool the mean subjective well-being values for each
61
Table 3.2: Selected macroeconomic indicators for Poland, 1989-2009
GDP per capita Registered In
ation Gini
Index Absolute unemployment rate coecient
Year (1989=100) (%) (%) (income)
1989 100.0 9,235 0.3 244.6 0.275
1990 88.4 8,164 6.5 555.4 0.268
1991 81.9 7,568 12.2 76.7 0.265
1992 83.8 7,740 14.3 45.3 0.274
1993 86.8 8,015 16.4 36.9 0.317
1994 91.1 8,413 16.0 33.3 0.323
1995 97.4 8,991 14.9 28.1 0.321
1996 103.4 9,545 13.2 19.8 0.328
1997 110.6 10,213 10.3 15.1 0.334
1998 116.0 10,717 10.4 11.7 0.326
1999 121.3 11,204 13.1 7.3 0.334
2000 127.2 11,743 15.1 10.1 0.345
2001 129.5 11,959 17.5 5.5 0.341
2002 131.4 12,137 18.0 1.9 0.353
2003 136.6 12,615 18.0 0.8 0.356
2004 144.0 13,297 19.0 3.6 0.366
2005 149.3 13,784 17.6 2.1 0.366
2006 158.6 14,648 14.8 1.1 0.343
2007 169.3 15,638 11.2 2.4 0.345
2008 177.4 16,388 9.5 4.3
2009 10.8
Source: see Appendix B.1
survey in the three datasets { WVS, EB, CBOS { and I regress the dependent variable (mean life
satisfaction at each survey date) on the matching values for GDP, unemployment, in
ation, and
the Gini coecient:
LS
st
= +
0
Y
t
+
0
D
s
+
st
; (3.1)
where LS is life satisfaction, Y is the vector of macroeconomic variables, and D is a vector of
dummy variables identifying each of the three dierent surveys. The subscript s denotes the
survey (WVS, EB, or CBOS) and t indicates time.
The analysis will then focus on each of the two main datasets { WVS and EB { separately in
order to better understand the nature and impact on subjective well-being of life circumstances
at the individual level. For example, studies nd that marriage has a strong positive eect on
subjective well-being (Blanch
ower and Oswald, 2004; Frey and Stutzer, 2002; Layard, 2005;
62
Waite, 1995; Waite and Lehrer, 2003; Zimmermann and Easterlin, 2006) or that unemployment
has a negative impact (Blanch
ower and Oswald, 2004; Frey and Stutzer, 2002). To the extent
that such life circumstances change during the transition process, they can play a role in explaining
changes in life satisfaction.
Table 3.3 shows that the characteristics of the World Values Survey sample change quite a
lot over time, particularly between 1989 and 1999. The percentage of married people decreases
by almost ten percentage points, as a result of increases in the percentage of single and widowed
people. The increase in widowhood is consistent with the fact that the life expectancy gap
between women and men increased in the post-communist years. The proportion of unemployed
respondents also increases, re
ecting the increases in the unemployment rate shown in table 3.2.
There is also an increase, although much smaller, in the percentage of the population who are not
in the labor force. This increase is mainly driven by the retired group. The percentage retired
in the WVS sample increases from 22 in wave 2 to 30 in the next two waves, the increase being
particularly high for women. This increase is not surprising given the lax eligibility rules for
both early retirement and disability pensions (Czepulis-Rutkowska, 1999). Also not surprising
is the increase in the level of education, given the signicant jump in the returns to education
brought about by the transition (Rutkowski, 1996). Considering the impact that these various
life circumstances have been shown to have on subjective well-being, I expect them to they play
a role in explaining changes in life satisfaction during the transition process in Poland. In order
to test this hypothesis, I run ordinary least square (OLS) regressions of life satisfaction on time,
with various individual controls. I am interested in the impact that the addition of these controls
has on the coecients for time, as a re
ection of their ability to explain changes in happiness.
Because the answers to life satisfaction questions take on discrete, ordered values, I also ran
ordered logit regressions with quite similar results, suggesting that the ndings are robust with
regard to methodology. The fact that the results were quantitatively similar to those from the
OLS specications is not surprising given that the WVS answer scale is fairly wide { 1 to 10.
63
Table 3.3: WVS sample description by wave
Wave 2 Wave 4 Wave 5
1989-1990 1999 2005
Mean life satisfaction 6.58 6.37 7.02
Sample size 1,897 1,085 989
% male 47.9 47.2 47.8
Mean age 44.2 44.8 45.2
Mean ACE 16.9 18.0 18.3
% married 75.4 66.4 60.7
% single 14.8 19.7 25.7
% divorced/ separated 2.9 3.4 4.7
% widowed 6.9 10.5 8.9
% employed 61.7 49.7 45.8
% unemployed 0.8 9.0 11.5
% not in labor force 37.5 41.3 42.7
Unemployment rate 1.3 15.3 20.0
% white collar 34.3 31.2 40.6
% blue collar 65.5 62.8 53.3
% never had job 0.2 6.0 6.1
Mean income bracket (1-10) 4.66 3.95 4.03
Because the WVS is not a panel dataset, I cannot follow the same individuals over time. One
way to partially correct for this shortcoming is to use year of birth xed-eects. I can thus follow
changes over time in the life satisfaction of each cohort, even though the responses do not come
from the same members of that cohort. By including year of birth xed-eects in the regressions of
life satisfaction over time, I can see to what extent changes in life satisfaction are driven by changes
in the surveyed cohorts as opposed to changes in the objective life circumstances within each birth
cohort. Table 3.4 provides supporting evidence for the hypothesis that cohort replacement may
play a role in explaining the life satisfaction trend. It shows mean life satisfaction over time for
ve birth cohorts spanning roughly fteen years each. While the trend in life satisfaction for all
cohorts is quite similar, the levels dier signicantly, with \new" cohorts happier than \older"
ones. As people in the older generations exit the sample and are replaced by younger people, this
is expected to in
uence the overall life satisfaction of the sample. This methodology allows me
64
to see to what extent changes in life satisfaction in Poland are driven by changes in people's life
circumstances and to what extent they are the result of a replacement of generations, irrespective
of the objective conditions of each generation.
Table 3.4: Life satisfaction means and standard errors by birth cohort and wave
Birth year
1929 1930-1944 1945-1959 1960-1973 1974
Wave 2 mean 6.538 6.388 6.574 6.936
(1989-1990) st. error (0.13) (0.10) (0.09) (0.11)
Wave 4 mean 6.316 6.021 6.052 6.749 6.836
(1999) st. error (0.22) (0.20) (0.13) (0.14) (0.22)
Wave 5 mean 6.795 6.565 6.849 7.070 7.448
(2005) st. error (0.35) (0.18) (0.12) (0.14) (0.11)
With only three survey waves carried out over a decade and a half of Polish transition (1989-
1990, 1999, 2005), the WVS makes it dicult to integrate indicators at the macroeconomic level
in the analysis. In order to do this and to get a more up to date view of life satisfaction in Poland,
I extend the analysis using the Eurobarometer data collected between 2001 and 2009. Table 3.6
present the characteristics of the EB sample and how they change over time. As table 3.1 showed,
two or more surveys were conducted in some of the years. The data for each year in table 3.5 are
obtained by pooling such surveys together. The available variables are quite similar to those in
the WVS. The one missing variable is income bracket. Instead I include in my analysis data on
GDP per capita. For each observation in the EB sample, the corresponding GDP value is that
for the matching year shown in table 3.1. Some of the negative changes that occurred before 2001
start being reversed { the percentage of married people is increasing, while the unemployment
rate is decreasing. GDP per capita is also increasing fairly steadily.
The EB analysis is carried out in a similar manner to that of the WVS data. I rst run
OLS regressions of life satisfaction on time with various controls in order to establish the trend
in life satisfaction and the role that changes in life circumstances play in explaining this trend.
I then add year of birth xed-eects to the analysis as a way to see to what degree changes in
65
Table 3.5: EB sample description by year
2001 2002 2003 2004 2005 2006 2007 2008
Mean life satisfaction 2.64 2.70 2.71 2.80 2.77 2.82 2.89 2.83
Sample size 994 985 1,981 995 1,975 1,975 1,978 1,973
% male 47.0 47.3 47.5 47.7 47.9 48.0 47.7 47.8
Mean age 44.1 42.8 42.9 43.3 43.3 43.3 43.7 43.8
% primary education 28.2 26.9 21.8 19.1 17.0 16.7 17.4 16.3
% secondary education 50.8 51.1 55.6 56.8 57.1 56.5 57.4 57.4
% university 9.0 9.2 10.4 12.1 11.7 12.8 10.8 13.6
% in school 11.9 12.8 12.2 12.0 14.2 14.0 14.5 12.7
% married 54.4 57.5 64.7 64.1 63.0 63.6 61.8 62.2
% single 34.7 31.8 23.1 20.1 20.2 19.5 21.4 20.4
% divorced/separated 2.7 3.2 3.5 5.0 5.7 5.7 5.7 6.0
% widowed 8.3 7.4 8.8 10.8 11.1 11.2 11.1 11.5
% employed 39.1 38.2 36.6 37.7 39.9 42.5 43.2 45.4
% unemployed 13.0 14.4 13.6 14.8 13.2 10.5 8.3 8.7
% not in LF 47.9 47.4 49.8 47.5 46.9 47.0 48.5 45.9
Unemployment rate 24.9 27.4 27.1 28.2 24.8 19.9 16.1 16.1
% white collar 41.4 39.7 42.2 40.6 39.5 39.0 40.2 42.6
% blue collar 43.7 45.3 40.9 43.2 43.8 44.6 44.5 43.0
% never had job 14.9 15.0 15.7 15.3 16.5 16.2 15.2 14.2
GDP per capita 11,959 12,137 12,496 13,297 13,662 14,432 15,391 16,200
(PPP constant international $2005)
happiness at a country level are the result of birth cohort replacement. Even though the answer
scale in the EB is much narrower than in the WVS { 1 to 4 instead of 1 to 10 { running ordered
logit regressions that take into account the discrete nature of the answer options does not lead
to substantively dierent results than the OLS specication. The result are therefore robust to
methodology for the EB as well.
66
3.3 Findings
The life satisfaction trend in Poland after 1989 is pieced together using the three data sources:
WVS, EB, and CBOS (gure 3.1). The WVS data show that life satisfaction at the end of the
1990s is lower than it had been at the onset of the transition in 1989, and this dierence is
statistically signicant. By 2005 subjective well-being in Poland had increased quite a lot so that
it is signicantly higher than in 1989. How the 1999 and 2005 values compare with happiness
under communism is harder to infer, considering that the 1989 value might be an underestimation
of the peak pre-transition happiness. The upward trend in happiness during the 2000s is conrmed
by the EB data. Although there are some
uctuations, life satisfaction in Poland in every year
starting with 2003 is signicantly higher than it had been in 2001. The CBOS data produce
somewhat con
icting evidence with the WVS for the 1990s. They show an upward trend in
happiness between 1992 and 1997, followed by a subsequent decrease. Is it possible to reconcile
the increase in happiness after 1992 with the fact that, according to the WVS, life satisfaction in
1997 is signicantly lower than in 1989? Figure 1 (p. 134) in Easterlin (2009) appears to indicate
that it is possible. In a number of transition countries for which life satisfaction observations
are available not only soon after the fall of communism and at the end of the 1990s, but also
in the mid-1990s, happiness shows a clear U-shape, with a collapse followed by recovery. The
hypothesis that the initial collapse in life satisfaction can be very dramatic is supported by the
evidence for the former GDP. Table A.2 in Easterlin's study shows that its decline in subjective
well-being between 1990 and 1991 is of such magnitude that the almost constant increase that
follows through 1997 is not enough to make up for the initial collapse.
Subjective well-being in Poland therefore rst collapses in the rst few years of the transition,
and then recovers in a slow, but fairly steady manner. How does this compare with the changes
in objective economic indicators brought about by the transition? Figure 3.2 shows annual and
67
Figure 3.1: Life satisfaction in Poland between 1989 and 2009
WVS interpolated series for GDP per capita, Gini coecient, unemployment and in
ation rates,
for the 1989 to 2009 time interval.
When it comes to GDP per capita, the same collapse as in terms of life satisfaction is encoun-
tered in the rst few post-communist years. However, the recovery is quicker than in the case
of subjective well-being and by 1996, GDP is already higher than it was in 1989. The increase
in GDP per capita is accompanied by an increase in inequality measured by the Gini coecient.
After virtually no unemployment during communism, the unemployment rate quickly gets into
the double digits. There is some improvement in the late 1990s, but it takes until 2008 for it
to nally decrease below 10 per cent. In
ation is rampant in the rst two years of the Polish
transition, but it then improves fairly steadily and has been under 5 per cent starting in 2002.
Overall, it seems the life satisfaction trend most closely follows the trend in GDP per capita,
but with a lag because the recovery in subjective well-being takes longer. This lag might be related
to the fact that the unemployment rate remains fairly high. Indeed, when I look at the eect of
the various macroeconomic indicators on mean life satisfaction (table 3.6), GDP per capita is the
68
Figure 3.2: GDP per capita, Gini coecient, unemployment and in
ation rates in Poland between
1989 and 2009
Source: See Appendix B.1.
Notes: The scales for the unemployment and in
ation rates are inverted. The values for the in
ation rate for
1989 and 1990 are over 100% and therefore are o-scale and set at 100% in the graph. For each macroeconomic
indicator, the solid line shows the annual series, while the dashed line shows the interpolated series obtained when
using only the years when the WVS was carried out in Poland - 1989, 1990, 1997, 1999, 2005.
strongest predictor of happiness. Its impact is stronger after 2001, while the predictive power of
the unemployment rate is stronger in the rst decade of transition. During the 1989-2001 time
interval, it takes a 4.8 per cent increase in GDP to make up for an increase in unemployment
of one percentage point, all other things equal. For the 2001-2009 interval a slim 0.3 per cent
in GDP is sucient. Neither the in
ation rate, nor the Gini coecient have a signicant eect
on mean happiness. In general, the selected macroeconomic indicators can explain most of the
variation in life satisfaction from one survey date to the next, as re
ected by the R
2
values which
are very close to 1 in column (5).
69
Table 3.6: Ordinary least square regressions of life satisfaction on selected macroeconomic indi-
cators
(1) (2) (3) (4) (5)
Coe. Coe. Coe. Coe. Coe.
Variable (t-stat) (t-stat) (t-stat) (t-stat) (t-stat)
1989 - 2009
Ln GDP per capita 1.126** 1.236**
(5.92) (6.61)
Unemployment rate -0.018 -0.036**
(-1.62) (-3.36)
In
ation rate -0.0004 -0.00001
(-0.65) (-0.02)
Gini coecient 0.769 1.260
(0.79) (1.71)
Constant -5.227** 5.479** 5.216** 4.951** -6.109**
(-2.96) (30.63) (79.37) (14.94) (-3.57)
Observations 32 35 32 31 31
R
2
0.944 0.879 0.877 0.876 0.966
1989 - 2001
Ln GDP per capita 0.949* 1.161**
(2.40) (8.93)
Unemployment rate -0.041* -0.054**
(-2.44) (-8.41)
In
ation rate 0.0002 0.001*
(0.28) (2.31)
Gini coecient 1.073 5.607**
(0.35) (5.49)
Constant -3.561 5.733** 5.151** 4.823** -6.500**
(-0.98) (23.37) (64.50) (5.01) (-4.97)
Observations 15 15 15 15 15
R
2
0.925 0.926 0.886 0.887 0.996
2001 - 2009
Ln GDP per capita 1.893** 2.173**
(7.59) (4.91)
Unemployment rate -0.048** -0.006
(-4.04) (-0.32)
In
ation rate -0.022 0.002
(-0.63) (0.09)
Gini coecient -0.519 0.995
(-0.47) (1.42)
Constant -12.586** 6.178** 5.373** 5.508** -15.498**
(-5.33) (27.19) (39.31) (12.87) (-3.53)
Observations 19 22 19 18 18
R
2
0.974 0.918 0.877 0.876 0.984
Dummy variables for each dataset were included among the controls.
Signicance levels: ** p<0.01, * p<0.05, + p<0.10.
70
So far, I have only used mean life satisfaction for each survey carried out as part of the
WVS, EB, or CBOS. I will next look at the individual data in the WVS and EB. This allows
me to incorporate dierences in the respondents' life circumstances as explanatory variables for
dierences in their subjective well-being. Columns (1)-(4) in table 3.7 show the results of ordinary
least squares regressions of life satisfaction on time and various independent variables, using the
WVS data. Column (1) conrms that life satisfaction at the end of the 1990s is signicantly
lower than it had been in 1989-1990, while in 2005 it is signicantly higher. The same trend
emerges when controls for gender, age, and education are added in column (2). Therefore, the
changes in life satisfaction are not simply the result of changes in the demographic composition
of the sample across waves. In general, it appears that young and more educated people are the
happiest, with middle aged people are the least happy. Table 3.3 showed important changes in
people's marital and employment status during this decade and a half of transition. Column (3)
conrms that these changes had a negative impact on subjective well-being because married and
employed individuals, whose proportion in the overall population decreases, are the most satised
with their lives. The results for marital status are consistent with the ndings of Bernhardt and
Fratczak (2005). In column (3) the coecient on wave 4 is no longer signicant, which means that
changes in marital and employment status can explain why life satisfaction in 1999 is lower than
in 1989. The impact of the increase in unemployment is stronger than that of the decrease in the
percentage of married individuals. Adding occupation and income to the regression controls in
column (4) makes the coecient for wave 4 positive, although not signicant. The coecient on
wave 5 remains signicant even after the various life circumstances are controlled for in columns
(3) and (4). This means that, while changes in life circumstances can explain the decrease in life
satisfaction between 1989 and 1999, they cannot explain why average happiness in 2005 is higher
than in 1989. This is not surprising given that table 3.3 showed no improvement in the marital
or employment situation of the respondents between 1999 and 2005. The next step, therefore, is
to look for the driving forces behind the substantial recovery of life satisfaction after 1999.
71
Table 3.7: Ordinary least squares and year of birth xed-eects regressions of life satisfaction {
WVS
Ordinary least squares Fixed-eects
(1) (2) (3) (4) (5) (6) (7) (8)
Coe. Coe. Coe. Coe. Coe. Coe. Coe. Coe.
Variable (t-stat) (t-stat) (t-stat) (t-stat) (t-stat) (t-stat) (t-stat) (t-stat)
Wave 4 -0.211* -0.298** -0.155 0.034 -0.305** -0.438** -0.361** -0.221
(1999) (-2.16) (-2.93) (-1.54) (0.31) (-3.05) (-3.15) (-2.64) (-1.45)
Wave 5 0.439** 0.322** 0.515** 0.663** 0.196* 0.030 0.120 0.137
(2005) (5.11) (3.54) (5.56) (5.43) (2.02) (0.17) (0.68) (0.59)
Male -0.036 -0.090 -0.170+ -0.069 -0.124 -0.182+
(-0.46) (-1.13) (-1.79) (-0.89) (-1.55) (-1.89)
Age 30-44 -0.402** -0.676** -0.416** 0.068 -0.114 -0.032
(-3.70) (-5.17) (-2.82) (0.33) (-0.55) (-0.14)
Age 45-59 -0.553** -0.798** -0.425** 0.306 0.261 0.568
(-4.87) (-5.92) (-2.77) (0.92) (0.79) (1.52)
Age 60+ -0.331* -0.451** 0.128 0.545 0.655 1.229*
(-2.36) (-2.59) (0.61) (1.12) (1.37) (2.26)
Age completed educ. 0.070** 0.059** 0.035+ 0.073** 0.062** 0.034+
(4.69) (3.96) (1.70) (4.85) (4.12) (1.65)
Single -0.362** -0.155 -0.546** -0.302+
(-2.67) (-0.99) (-3.92) (-1.92)
Divorced/Separated -0.783** -0.624* -0.775** -0.616*
(-3.50) (-2.29) (-3.38) (-2.17)
Widowed -0.458** -0.617** -0.439* -0.533**
(-2.66) (-3.05) (-2.56) (-2.61)
Unemployed -1.115** -1.202** -1.172** -1.220**
(-6.15) (-4.11) (-6.48) (-4.14)
Not in labor force -0.225* -0.306* -0.278** -0.306*
(-2.35) (-2.12) (-2.86) (-2.02)
Blue collar 0.196+ 0.141
(1.80) (1.30)
Never had job 0.461 0.384
(1.61) (1.39)
Income bracket 0.272** 0.272**
(9.80) (9.75)
Constant 6.584** 5.756** 6.349** 5.103** 6.673** 5.276** 5.842** 4.725**
(126.52) (19.21) (20.64) (11.52) (122.50) (15.79) (17.45) (10.20)
Obs. 3,971 3,768 3,761 2,742 3,964 3,768 3,761 2,742
R
2
0.011 0.030 0.050 0.094 0.054 0.064 0.087 0.138
Omitted categories: wave 2 (1989-1990), age 29 or less, married, employed, white collar.
Signicance levels: ** p<0.01, * p<0.05, + p<0.10.
72
Table 3.6 already showed that the impact of GDP on happiness is much stronger after 2000 than
before. As GDP continues to increase through 2005 to levels much higher than at the beginning
of the transition, this can explain why life satisfaction in 2005 is higher than in 1989. Another
possible explanation { birth cohort replacement { is tested in columns (5)-(8) of table 3.7. The
same regressions as in the rst four columns are run, but with year of birth xed-eects instead
of OLS regressions. Each individual year of birth identies a cohort. A coecient of around
0.4 for wave 5 quickly becomes virtually zero in column (6) when demographic characteristics
are controlled for and birth cohort eects are eliminated. The coecients for the various life
circumstances do not change very much. Therefore, the replacement of generations does play a
role in the recovery in happiness.
Table 3.8: Ordinary least square regressions of life satisfaction on cohort { WVS
(1) (2) (3)
Coe. Coe. Coe.
Variable (t-stat) (t-stat) (t-stat)
Birth year 1929 0.031 0.286* 0.391*
(0.24) (2.00) (2.43)
Birth year 1930-1944 -0.161 -0.056 -0.063
(-1.52) (-0.50) (-0.55)
Birth year 1960-1973 0.425** 0.383** 0.453**
(4.27) (3.72) (4.27)
Birth year 1974 0.680** 0.556** 0.861**
(4.76) (3.82) (5.12)
Controls wave wave, gender, wave, gender,
education education,
marital status,
empl. status
Observations 3,964 3,768 3,761
R
2
0.024 0.031 0.052
Omitted cohort: 1945-1959.
Signicance levels: ** p<0.01, * p<0.05, + p<0.10.
In order to see more clearly how life satisfaction diers by cohort, I divide the sample in
four dierent cohorts in table 3.8. Younger generations are more satised with their lives than
older generations. Controlling for life circumstances in column (3) further increases the happiness
advantage of people born in 1974 or later compared with older cohorts. This means that their
73
higher life satisfaction is mainly the result of them being \[r]aised [. . . ] in the wild" (Easterlin,
2009, p.138), so that even if they are confronted with the same amount or even more problems
than older individuals, they are simply better adjusted to this new society and better equipped
to deal with the negative side eects of transition. These ndings are consistent with the higher
degree of adaptation among younger generations that Alesina and Fuchs-Sch undeln (2007) nd
in the former GDR.
Further evidence of a higher degree of adaptation among young people to the transition chal-
lenges is found by looking specically at the eect of employment status on life satisfaction for
various birth cohorts (table 3.9). Although the percentage of unemployed individuals is much
higher for the youngest cohort compared with older ones, their relative life satisfaction compared
with employed individuals is not as low as it is in the case of the older cohorts. This does not mean
that they prefer to be unemployed. However, unlike older generations, who grew up without ever
expecting to be faced with unemployment, younger people are much less likely to be blindsided
by its occurrence, which can explain why it has a smaller eect on their happiness levels.
The ordinary least squares regressions of life satisfaction on time using the Eurobarometer
data gathered starting in 2001 conrm the mostly upward trend of life satisfaction during the
second decade of transition (table 3.10). Changes in demographics and in life circumstances do
not appear to be the driving force behind this trend because adding these variables among the
regression controls in columns (2)-(4) has little eect on the coecients for time. This is not
surprising given that these socio-economic indicators do not change as much after 2001, at least
compared with the dramatic changes that marked the 1990s. However, controlling for GDP in
columns (5) and (6) makes all the time coecients but one no longer signicantly dierent from
zero. This conrms that the increase in life satisfaction is in line with the steady increase in
GDP per capita. In matching the values for GDP per capita to each surveyed individual I used
the corresponding years from table 3.1 { for example, an individual surveyed as part of the May
2005 Eurobarometer will be matched with the average GDP per capita in 2004 and 2005, while
74
Table 3.9: The impact of employment status on life satisfaction by birth cohort and wave { WVS
Birth year
1929 1930-1944 1945-1959 1960-1973 1974
Wave 2 (1989-1990)
% unemployed 0.0 0.2 1.1 1.8
Coecient 2.759** -1.404+ -0.728
% not in labor force 79.9 35.3 13.6 40.6
Coecient -0.260 -0.209 -0.309 -0.395
Wave 4 (1999)
% unemployed 0.0 2.2 7.8 11.4 20.0
Coecient -1.732 -1.416** -1.551** -0.588
% not in labor force 97.2 86.5 21.2 14.1 40.0
Coecient -2.134* -0.774 -0.601 -0.009 -0.097
Wave 5 (2005)
% unemployed 2.1 0.0 11.5 12.7 18.7
Coecient -0.246 -2.079** -0.627 -0.114
% not in labor force 96.8 97.3 41.9 9.6 31.2
Coecient -1.135 -0.524 -0.618* 0.168 0.324
The coecients are from regressions of life satisfaction for each cohort. Omitted
category: employed. Other controls: gender, age, age squared, education, marital
status. Signicance levels: ** p<0.01, * p<0.05, + p<0.10.
an individual included in the October 2005 Eurobarometer will be matched with the GDP per
capita for 2005. Young and more educated people continue to show the highest levels of subjective
well-being. Being married and having a job are also associated with a higher life satisfaction, and
so is having a white collar job.
The inclusion of year of birth xed eects in the regressions (table 3.11) makes much less of a
dierence than it did for the WVS. Because the time span of the EB is fairly short, the proportion
of the various cohorts in the sample does not change very much. Therefore it is not surprising that
the changes in life satisfaction between 2001 and 2009 do not appear to be the result of cohort
replacement.
75
Table 3.10: Ordinary least squares regressions of life satisfaction
(1) (2) (3) (4) (5) (6)
Coe. Coe. Coe. Coe. Coe. Coe.
Variable (t-stat) (t-stat) (t-stat) (t-stat) (t-stat) (t-stat)
Year 2002 0.065 0.056 0.066+ 0.071+ 0.053 0.066
(1.49) (1.37) (1.65) (1.76) (1.27) (1.62)
Year 2003 0.073* 0.061+ 0.064+ 0.065+ 0.052 0.052
(1.96) (1.73) (1.85) (1.88) (1.23) (1.22)
Year 2004 0.166** 0.145** 0.161** 0.166** 0.126+ 0.135+
(4.25) (3.92) (4.38) (4.53) (1.77) (1.91)
Year 2005 0.128** 0.096** 0.110** 0.118** 0.072 0.079
(3.66) (2.93) (3.39) (3.65) (0.86) (0.95)
Year 2006 0.177** 0.142** 0.153** 0.163** 0.108 0.107
(5.16) (4.39) (4.76) (5.07) (0.96) (0.96)
Year 2007 0.248** 0.224** 0.225** 0.234** 0.179 0.160
(7.26) (6.94) (7.01) (7.31) (1.21) (1.09)
Year 2008 0.197** 0.163** 0.167** 0.174** 0.109 0.086
(5.76) (5.03) (5.19) (5.44) (0.61) (0.49)
Year 2009 0.234** 0.206** 0.207** 0.215** 0.073 0.061
(7.12) (6.63) (6.52) (6.78) (0.40) (0.33)
Male 0.017 -0.009 0.010 0.019 0.008
(1.37) (-0.70) (0.78) (1.50) (0.62)
Age 30-44 -0.186** -0.222** -0.204** -0.190** -0.210**
(-9.68) (-10.89) (-9.96) (-9.14) (-9.74)
Age 45-59 -0.329** -0.335** -0.309** -0.332** -0.313**
(-16.74) (-15.68) (-14.29) (-15.52) (-13.70)
Age 60+ -0.245** -0.224** -0.200** -0.254** -0.207**
(-11.34) (-7.93) (-6.92) (-10.73) (-6.71)
Primary education -0.194** -0.161** -0.117** -0.195** -0.124**
(-9.83) (-8.04) (-5.67) (-9.14) (-5.74)
University 0.276** 0.241** 0.186** 0.275** 0.183**
(15.99) (13.64) (9.89) (14.69) (9.15)
Still in school 0.245** 0.270** 0.204** 0.253** 0.201**
(11.28) (9.00) (5.54) (10.83) (5.24)
Single -0.080** -0.082** -0.085**
(-4.07) (-4.21) (-4.27)
Divorced/ separated -0.278** -0.282** -0.289**
(-8.76) (-8.86) (-8.98)
Widowed -0.186** -0.182** -0.175**
(-7.37) (-7.23) (-6.80)
Unemployed -0.294** -0.292** -0.297**
(-12.81) (-12.51) (-12.34)
Not in labor force -0.088** -0.098** -0.095**
(-4.48) (-4.88) (-4.49)
Blue collar -0.133** -0.130**
(-8.86) (-8.22)
Never had job 0.040 0.050
(1.20) (1.43)
Ln GDP per capita 0.178 0.288
(0.31) (0.51)
Constant 2.638** 2.811** 2.944** 2.976** 1.144 0.275
(87.24) (86.49) (86.71) (86.96) (0.21) (0.05)
Observations 15,786 15,376 14,529 14,529 13,520 13,373
R
2
0.010 0.111 0.134 0.140 0.112 0.143
76
Table 3.11: Year of birth xed-eects regressions of life satisfaction
(1) (2) (3) (4) (5) (6)
Coe. Coe. Coe. Coe. Coe. Coe.
Variable (t-stat) (t-stat) (t-stat) (t-stat) (t-stat) (t-stat)
Year 2002 0.045 0.049 0.058 0.062 0.046 0.059
(1.10) (1.22) (1.45) (1.56) (1.11) (1.46)
Year 2003 0.047 0.044 0.044 0.046 0.038 0.038
(1.35) (1.28) (1.28) (1.33) (0.90) (0.90)
Year 2004 0.133** 0.120** 0.130** 0.137** 0.110 0.120+
(3.58) (3.23) (3.50) (3.70) (1.53) (1.69)
Year 2005 0.086** 0.071* 0.078* 0.088** 0.058 0.067
(2.64) (2.16) (2.36) (2.68) (0.69) (0.81)
Year 2006 0.121** 0.104** 0.107** 0.120** 0.088 0.092
(3.74) (3.17) (3.25) (3.64) (0.78) (0.82)
Year 2007 0.184** 0.180** 0.173** 0.185** 0.159 0.148
(5.71) (5.44) (5.19) (5.58) (1.07) (1.01)
Year 2008 0.124** 0.114** 0.108** 0.120** 0.090 0.075
(3.85) (3.39) (3.19) (3.55) (0.51) (0.43)
Year 2009 0.151** 0.150** 0.142** 0.155** 0.051 0.046
(4.87) (4.54) (4.17) (4.56) (0.28) (0.25)
Male 0.016 -0.008 0.010 0.020 0.008
(1.32) (-0.62) (0.80) (1.50) (0.62)
Age 30-44 -0.024 -0.030 -0.032 -0.015 -0.049
(-0.63) (-0.79) (-0.86) (-0.37) (-1.21)
Age 45-59 -0.016 -0.003 -0.002 -0.002 -0.024
(-0.27) (-0.05) (-0.03) (-0.03) (-0.38)
Age 60+ 0.099 0.106 0.101 0.098 0.068
(1.25) (1.31) (1.26) (1.13) (0.78)
Primary education -0.192** -0.164** -0.120** -0.195** -0.127**
(-9.55) (-8.10) (-5.74) (-8.96) (-5.80)
University 0.272** 0.239** 0.185** 0.273** 0.183**
(15.60) (13.46) (9.81) (14.37) (9.10)
Still in school 0.206** 0.199** 0.156** 0.196** 0.143**
(7.41) (5.72) (3.98) (6.53) (3.55)
Single -0.101** -0.100** -0.104**
(-5.11) (-5.10) (-5.19)
Divorced/ separated -0.273** -0.279** -0.284**
(-8.50) (-8.63) (-8.71)
Widowed -0.191** -0.187** -0.180**
(-7.34) (-7.18) (-6.75)
Unemployed -0.294** -0.289** -0.294**
(-12.81) (-12.41) (-12.24)
Not in labor force -0.088** -0.094** -0.093**
(-4.29) (-4.48) (-4.17)
Blue collar -0.131** -0.129**
(-8.69) (-8.11)
Never had job 0.017 0.022
(0.50) (0.63)
Ln GDP per capita 0.061 0.143
(0.11) (0.25)
Constant 2.690** 2.657** 2.793** 2.837** 2.085 1.515
(95.29) (62.50) (63.67) (64.42) (0.39) (0.28)
Observations 15,757 15,376 14,529 14,529 13,520 13,373
R
2
0.090 0.121 0.144 0.150 0.122 0.153
77
3.4 Summary and Conclusions
The present analysis supports the idea that changes in economic circumstances are an impor-
tant driving force behind changes in life satisfaction. Other factors, however, also play a role. In
Poland, the rst couple of years after the fall of communism were marked by an economic collapse,
accompanied by all appearances by a collapse in subjective well-being. The economic recovery
was fairly swift, especially in terms of GDP per capita which was back to its pre-transition level as
early as 1996. The recovery in terms of life satisfaction was much slower. In addition to economic
improvements, it was made possible, to an important extent, by birth cohort replacement. As
the generations raised under capitalism replace the generations raised under communism in the
population, subjective well-being tends to increase even if objective circumstances do not nec-
essarily improve. Young people are simply better equipped to deal with the specic problems
of capitalism compared with the older generations \well-embarked on a life course set under the
conditions of the socialist greenhouse" (Easterlin, 2009, p.138).
Was the Polish transition a success when considering the level of self-reported happiness?
Given the fragmentary nature of the data, it is hard to say with certainty how current life satis-
faction in Poland compares with the pre-transition level. Piecing the various data sources together,
however, it appears that happiness today is higher than in 1989 even when cohort eects are con-
trolled for. As older generations are replaced by younger ones, average life satisfaction is even
more likely to increase. It is also clear that the transition paid o for some people { the young,
more educated, or better skilled { but not for others { the middle-aged, less educated, or un-
skilled. If more had been done to help those most vulnerable to the transformations involved by
the transition, the overall outcome could have been better. In other words, for a transition to
be successful, eort has to be put not only in maximizing the benets for the winners, but also
in minimizing the costs for the less fortunate ones and in making sure that people are not made
78
redundant in the new society. Ultimately, however, what will allow for a high proportion of the
population to embrace the new regime is the replacement of generations.
The results for life satisfaction in Poland are quite similar to what I found in the previous
chapter when looking at job satisfaction in a larger number of transition and non-transition
countries. The same fall followed by recovery is found, and the same population segments {
young and more educated people { can be counted among the winners of the transition process.
An interesting research topic would be to look at life satisfaction in a country in Central and
Eastern Europe, such as Hungary, that adopted a more gradual approach to the transition. This
could shed some light on whether the impact of a gradual approach on people's subjective well-
being is dierent from the impact that the shock therapy had in Poland, especially when looking
at a long time horizon.
79
Chapter 4
Summary
The purpose of my dissertation is to assess the process of transition using not only objective
economic indicators, but measures of subjective well-being as well. Indeed, previous research
has suggested that objective measures of well-being such as GDP should be complemented by
subjective measures (Veenhoven, 2002; Shah and Marks, 2004; Donovan and Halpern, 2002).
Furthermore, a recent report commissioned by French President Nicolas Sarkozy from a panel
of highly distinguished economists, including Nobel winners Joseph Stiglitz and Amartya Sen,
called for new measures of growth including subjective well-being. Life satisfaction research
therefore has the potential to signicantly in
uence public policy. My focus is on the trend in
subjective well-being as a measure of the success of the transformation in the years following the
fall of communism in Central and Eastern Europe. Whether looking at overall life satisfaction
or at satisfaction with the particular domain of work, I nd a decrease in the rst few years of
transition, followed by a subsequent recovery. This trend in subjective well-being follows fairly
closely the trend in GDP per capita, but the recovery in terms of GDP typically is achieved faster
than the recovery of life satisfaction.
Chapter 2 is a comparative analysis of job satisfaction in Eastern and Western Europe between
1990 and 2005. I am interested in the determinants of job satisfaction in the two regions and
particularly in how the level and the trend of job satisfaction may dier between Eastern and
80
Western Europe. Labor relations are quite dierent under communism compared with a free
market economy. Under communism, there is no open unemployment, but real wages are low,
and skill and sectoral pay dierences are narrow. Capitalism comes with its own set of problems
- jobs become scarcer as unskilled work moves towards poorer countries, and this has stressful
consequences on people's lives. What is then the eect on people's levels of job satisfaction of the
Eastern European transition from communism to capitalism? Do they value the newfound freedom
enough to compensate for the job insecurity that comes with the transition to a free market
economy? Do Eastern and Western Europe become more similar in terms of job satisfaction
and its determinants as the transition progresses? After all, Western Europe is the example of
capitalist model that Eastern Europe is aiming towards.
In answering these questions I use data from the World Values Survey and the International
Social Survey Programme \Work Orientations" module. The WVS provides data for 6 Eastern
European and 13 Western European countries around 1990 and 1999. In order to get more
updated information, I carry out a separate analysis using the ISSP surveys of 1997 and 2005,
which included 5 Eastern European and 5 Western European countries. I focus on employed
individuals between 18 and 65 years old, roughly the age group for which it is typical to be
involved in the job market. The analysis is carried out using ordinary least square regressions
with standard errors adjusted to allow for clusters in the error term within countries. As far as
the grouping of the countries is regarded, the hypothesis here is that the communist experience
is suciently similar to leave an identiable common legacy aecting outcomes and views of the
labor market in these Eastern European countries. Indeed a series of F-tests comparing each
country with the overall region to which it belongs conrms the grouping of countries in Eastern
and Western Europe. In order to account for any heterogeneity within regions, country xed-
eects are used where appropriate. This approach also removes the estimation bias caused by
individual-invariant country characteristics.
81
Job satisfaction in Eastern Europe is lower than in the West throughout the 1990 to 2005
time interval. In fact, the gap between the two regions increases between 1990 and 1999, before
decreasing by 2005. These changes in the job satisfaction gap are the result of a decrease followed
by recovery in transition countries, while in Western Europe there is no signicant change in
satisfaction with work during this decade and a half. What are some of the factors that could
explain the trend and dierences in the level of job satisfaction in Eastern and Western Europe?
Macroeconomic indicators such as GDP, unemployment and in
ation, have a signicant in
uence
on satisfaction with work, the rst indicator having a positive impact, while increases in the last
two have a negative impact. At the individual level, a higher level of education, having a full time
job, a white collar occupation, a higher income, or more freedom of decision at work all translate,
on average, into higher levels of job satisfaction. When it comes to age, people between 50 and 65
years old have the highest work satisfaction. The impact of these determinants on job satisfaction
is quite similar in transition and non-transition countries. However, objective factors such as edu-
cation or income tend to have a stronger impact in Eastern Europe, while subjective perceptions
of one's job characteristics matter more in the West. Over time, young people close some of the
job satisfaction gap compared with older workers, particularly in transition countries. In these
countries, the premium for education when it comes to satisfaction with work also increases. This
points to young and more educated people as winners of the transition process relative to older,
less educated workers.
What role do these factors play in explaining the job satisfaction dierences between Eastern
and Western Europe? The lower level of satisfaction with work in transition compared with non-
transition countries is mainly the result of dierences in macroeconomic conditions between the
two groups of countries. Indeed, GDP per capita in Eastern Europe is systematically about half
that in Western Europe. While the transition countries start o with virtually zero unemployment,
by the late 1990s their unemployment rate is higher than in the West. The rst few years of
82
transition also bring about rampant in
ation. Among these economic indicators, the role of GDP
per capita is strongest in explaining job satisfaction dierences between East and West.
While job satisfaction in Western Europe stays virtually unchanged, in the East there is a
decrease during the 1990s followed by a recovery by 2005. Can these same determinants explain
the changes over time in job satisfaction in transition countries? By 1999, economic conditions in
Eastern Europe had generally improved after the collapse of the rst few years of transition, and in
most countries GDP had recovered to its 1990 level. Job satisfaction, however, was still lower than
in 1990. The failure of satisfaction with work in transition countries to recover commensurately
with the economic recovery by 1999 is the result of a birth year cohort eect. As other authors
previously found, new generations entering the labor market are less satised with work than
the older cohorts exiting the labor market, when everything else is equal. Because such a cohort
replacement occurs between 1990 and 1999, although economic conditions in Eastern Europe at
the two dates are quite similar, job satisfaction is lower at the latter date. In Western Europe,
the cohort eect is counteracted by the fact that economic conditions are better in 1999 compared
with 1990. What explains then the increase in job satisfaction in transition countries between 1997
and 2005? First of all, the economic improvements in Eastern Europe during this time interval
are quite dramatic, with GDP per capita in 2005 being about 30 percentage points higher than
in 1997. At the same time, young people in transition countries are better adapted to the new
market conditions compared with older individuals, which allows them to improve their relative
job satisfaction. The combined eect of these changes explains why the cohort eect is no longer
a factor driving job satisfaction in transition countries by 2005.
Because I use two dierent datasets, it is impossible to assess with certainty whether job
satisfaction in Eastern Europe is back to its pre-transition level. However, the results indicate
that the transition countries are on their path to recovery and convergence with the West is
possible.
83
Chapter 3 focuses on overall life satisfaction during the transition in Poland. Since 1989
Poland has been considered a leader in economic reform. It is also an example of a shock therapy
approach to the transition process, as opposed to the more gradual approach that countries such
as Hungary chose. This type of transition policies led to a very sharp economic collapse in the rst
couple of years of transition, but also to a fairly quick recovery. As a result, Poland was the rst
transition country to bottom out and begin growing again, so that by 1996 it was the only country
to have surpassed its pre-1989 level of GDP per capita. The decrease in output also led to a sharp
increase in unemployment, which in Poland tends to be higher than in neighboring countries and
stays in the double digits even in the 2000s. It is not surprising then that the transition aroused
profound anxieties about surviving in the new system. Indeed, alcohol consumption increased in
the rst few years of transition as a re
ection of an increase in stress levels. This aected men
more than women, leading to an increase in the life expectancy gap between the two genders in
favor of women. Changes also occurred in the realm of family life, with both marriage and fertility
rates decreasing.
How does life satisfaction in Poland change as a result of the transformation from a planned
economy to a free market model? How does the trend in life satisfaction compare with the changes
in economic conditions? Who are the winners and the losers of the Polish transition? In order
to answer these questions I use data from the World Values Survey and the Eurobarometer. The
WVS surveys were carried out in Poland around 1989-1990, 1999, and 2005. The Eurobarometer
brings the analysis up to date with surveys carried out at least once a year between 2001 and 2009.
The data from these surveys paint the same picture of increase in unemployment and decrease
in marriage rates that emerges from ocial statistical data, especially during the rst decade
of transition. The 1989-1990 survey is considered a re
ection of the pre-transition conditions in
Poland and I therefore use it as a benchmark against which later developments are judged. Does
this early survey really provide a good approximation of life satisfaction under socialism? Easterlin
(2009) argues that it is actually an underestimation of life satisfaction under communism, given
84
that social anomy is one of the fundamental features of the Polish society around the time of the
transition, with negative eects on life satisfaction.
Life satisfaction in Poland in 1999 is signicantly lower than it was in 1989. Despite the
fact that GDP per capita quickly recovered to pre-transition levels, increased unemployment and
involuntary early retirement as a way to avoid the even worse alternative of unemployment have
taken a toll on the happiness of the Polish people. The decrease in the percentage of married
people also plays a role in this decline, especially because it is typically dictated by the harsh
economic conditions and not by a change in people's attitudes towards marriage.
By 2005 life satisfaction in Poland reaches levels signicantly higher than in 1989. In fact, the
upward trend continues through 2009, as the Eurobarometer data show. This happens despite
the fact that both the unemployment rate and the percentage of unmarried people are higher in
2005 compared with 1989. What signicantly improves in this time interval is GDP { by 2005
it is around 50 percent higher than it was at the onset of transition. Indeed, improvements in
GDP per capita are an important driving force behind the recovery in life satisfaction. Another
important factor in this recovery of life satisfaction is birth cohort replacement { new generations
appear to be better adjusted to the new society and better equipped to deal with the negative side
eects of transition. As they replace older people in the population, life satisfaction is more likely
to improve, all other things being equal. This is consistent with the higher degree of adaptation
among younger generations that Alesina and Fuchs-Sch undeln (2007) nd in the former GDR.
Unlike my ndings for job satisfaction, when it comes to overall life satisfaction, people younger
than 30 are the happiest group, while those between 45 and 59 years old are the least happy.
This happens despite the fact that young people are more vulnerable to unemployment and more
likely to postpone marriage as a result of a precarious economic situation. Because they have
been raised under capitalism, however, they are more apt at dealing with these issues than the
older generations raised in the socialist \greenhouse".
85
In general, it is quite clear that the transition is characterized by the existence of both winners
and losers { the young, more educated and skilled people are among the winners, while the
less educated and middle-aged are generally worse o than during communism. It is important
therefore to be aware of the costs that the transition brings for certain population categories and
to try and make the process more manageable for these social groups.
My analysis is carried out for the overall population, without dierentiation by gender. Of
course, the situation of men and women during the transition is not identical, but it is similar
enough to warrant this unitary analysis. Indeed, the general trends in life and job satisfaction
are quite similar for the two genders, and the coecients on gender in my regressions are not
signicant. In order to further conrm this, I carried out separate regressions of life and job
satisfaction for men and women and I nd that, while the magnitude of the coecients sometimes
diers, their signs are fairly consistently the same.
For both life and job satisfaction, the initial decrease and the following increase during the
transition are statistically signicant. Furthermore, changes in subjective well-being of the mag-
nitude of those I nd over the rst couple of decades of transition are clearly a manifestation of
the dramatic transformations that these societies are undergoing and are much rarer under more
stable circumstances. Is a 0.2 average decrease (on a scale from one to ten) in job satisfaction in
Eastern Europe over a ten year time period really large enough to re
ect more than a spurious
change? The virtually zero change that occurs in Western Europe is one indication that the
changes in transition countries are indeed of a signicant magnitude. In a number of developed
countries in both Europe and North America, the WVS included a job satisfaction question in
the early 1980s as well. When looking at job satisfaction in these countries over the almost two
decades between the early 1980s and the late 1990s, there is an overall decrease of only about 0.1,
only half the change occurring in Eastern Europe over a time interval half as long. How about the
magnitude of the changes in life satisfaction in Poland? Is the 0.2 decrease over ten years, followed
by a 0.6 increase over a mere six years (on a one to ten scale), a noteworthy change? Such life
86
satisfaction changes over about a decade and a half are almost never encountered in developed
countries. In addition to transition countries, only the countries of Latin America, themselves
going through important transformations, show similarly large changes.
Even after controlling for unemployment, in
ation, or relative income, my analysis nds a
strong impact of GDP per capita on both life and job satisfaction, which some might nd sur-
prising. However, both the collapse and the following recovery of GDP in transition countries
have been of a magnitude rarely encountered over such a short time span. Because my datasets
do not include information on the respondents' absolute income, GDP per capita can be seen as
a measure of how the change from a planned economy to a free market and the implementation
of this new economic system aect people's incomes. The changes in GDP are also a re
ection
of the chaotic situation in transition countries { the labor markets were restructured, large state
enterprises were closed, even those workers who were able to keep their jobs were not certain if
their salaries would be paid on time. All these changes had stressful consequences on people's
lives, aecting their physical and mental health, leading to alcohol abuse, consequences presented
in mode detail for Poland in chapter 3. GDP can be seen as a summary measure of all this turmoil
and, therefore, its signicant eect on subjective well-being is not surprising.
My research consistently points to birth cohort replacement as an important explanation for
the eventual recovery in subjective well-being in transition countries. As my analysis of life
satisfaction in Poland has shown (chapter 3), young people do not actually enjoy better life
circumstances { they are more likely to be unemployed, they have to postpone marriage and stay
single longer than previous generations used to. One might argue that this latter development is
just a re
ection of changing times, and a sign of development as people choose to pursue a higher
education, to focus on their careers and only later on to start their family lives. In my opinion,
the higher likelihood to stay single after the onset of the transition is more complex than this.
In the WVS, there is a question on the importance of family. In Eastern Europe this stays fairly
constant throughout the transition, despite a considerable decrease in the percentage of married
87
individuals. However, signicant correlations exist across countries between the change in this
percentage and the change in employment and in a number of macroeconomic indicators. This
points toward changing economic circumstances as the source of changing marriage patterns,
rather than a fundamental shift in attitudes toward family formation and dissolution. People
need to adjust to the economic shocks brought about by transition, and one way to do this is
by postponing family formation. What makes new generations report higher levels of subjective
well-being despite these challenges is the fact that, having been raised in the wild, they are better
equipped to deal with the negative side eects of transition.
An important constraint on my research has been data availability. The available surveys
also have various limitations. The WVS has the important advantage of having been carried
out very soon after the fall of communist, which makes it a fairly good re
ection of conditions
during communism. A signicant downside, however, is the low frequency with which the surveys
have been carried out, which makes it dicult to pinpoint the eect on subjective well-being
of specic transition reforms carried out at particular times. This is also a reason why I use
additional datasets, with dierent scales, in order for my analysis to span a longer time interval.
Furthermore, it is clear that the transition has dierent eects in dierent life domains { for
example, child care might have been better during communism, but goods availability is much
better under capitalism. The WVS includes some questions on domain satisfaction, such as
job satisfaction, but they are not consistently asked in each survey, which makes it dicult to
get an overall picture of changes in these domains and their impact on overall life satisfaction.
The fact that the R-square in my regressions is usually less than 0.1 shows that there are other
life circumstances that have an important impact on subjective well-being but which were not
captured in these surveys.
My research so far already points to some possible extensions to this analysis of subjective
well-being in transition countries. Given that a number of studies have found that marriage has
a strong positive eect on life satisfaction (Waite, 1995; Waite and Lehrer, 2003; Zimmermann
88
and Easterlin, 2006), it would be interesting to look at the impact of the transition in the realm
of family life. For two or three decades before the start of the transition from a planned to
a market economy, the countries of Central and Eastern Europe displayed a pattern of early
and nearly universal entry into marriage. This pattern was generally stable during this long
time span, with some moderate changes due primarily to the eect of pro-natal policies which
provoked a slightly earlier entry into rst marriages (Philipov and Dorbritz, 2007). With the
transition came dramatic changes in the family and demographic behavior. From 1989 to 1999,
the crude marriage and birth rates generally fell by between one-quarter and one-half (UNICEF,
2001). Substantial delays in the timing of marriage occurred, as well as signicant increases in
nonmarital cohabitation and childbearing outside of marriage (Thornton and Philipov, 2003).
Are the drop in marriage rates and the corresponding increase in cohabitation and divorce part
of a wide adoption of \individualistic Western values", including a change in people's attitudes
toward marriage? Or are such phenomena mainly the result of the need to adjust to the economic
shocks brought about by transition in the form of lower household incomes, higher unemployment
and resulting job insecurity, or high in
ation? As stated earlier, my preliminary analysis points
to changing economic circumstances as the source of changing marriage patterns, rather than a
fundamental shift in attitudes toward family formation and dissolution.
It has been shown that overall satisfaction depends on satisfaction with various domains of
life, such as family, health and work (Campbell et al., 1976; Easterlin, 2006). Another important
line of research, therefore, is to look at satisfaction with various life domains and their impact on
overall life satisfaction during the transition. This can be done for East and West Germany, using
data from the German Socio Economic Panel, carried out in West Germany starting in 1984 and
in East Germany starting in 1990. This dataset contains information not only on respondents'
overall life satisfaction, but also on their satisfaction with a variety of life domains, such as living
level, health, household income, work, child care, the supply of goods and services in the area,
and so on. Before unication, one would expect substantial dierences in these domains between
89
East and West Germany. For instance, child care was more readily available in East Germany,
while goods availability was much better in the West. Have such dierences become narrower
since the unication? Is the impact of these life domains on overall subjective well-being dierent
in East and West Germany? These are some of the questions worth investigating.
90
Bibliography
Philippe Aghion and Olivier J. Blanchard. On the speed of transition in Central Europe. In Stanley
Fischer and Julio J. Rotemberg, editors, National Bureau of Economic Research Macroeconomic
Annual, pages 283{320. MIT Press, Cambridge, 1994.
George A. Akerlof, Andrew K. Rose, Janet L. Yellen, Laurence Ball, and Robert E. Hall. Job
switching and job satisfaction in the U.S. labor market. Brookings Papers on Economic Activity,
1988:495{594, 1988.
Alberto Alesina and Nicola Fuchs-Sch undeln. Goodbye Lenin (or not)? The eect of communism
on people. The American Economic Review, 97(4):1507{1528, 2007.
Helmut K. Anheier and Wolfgang Seibel. The nonprot sector and the transformation of soci-
eties: A comparative analysis of East Germany, Poland, and Hungary. In Walter W. Powell
and Elisabeth S. Clemens, editors, Private Action and the Public Good, pages 177{189. Yale
University Press, New Haven and London, 1998.
Michael Argyle. The Psychology of Happiness. Routledge, London, 1989.
Leszech Balcerowicz. Economic transition in Central and Eastern Europe: Comparisons and
lessons. Second Annual IFC Lecture, Washington, DC, 1993.
Leszek Balcerowicz. Understanding postcommunist transitions. Journal of Democracy, 5(4):75{
89, 1994.
Franz Barjak. Regional disparities in transition economies: A typology for East Germany and
Poland. Post-Communist Economies, 13(3):289{311, 2001.
Ulrich Beck. Risk Society. Towards a New Modernity. Sage, London, 1992.
Ulrich Beck. The New Brave World of Work. Polity Press, Cambridge, 2000.
Michael Bernhard. Civil society after the rst transition: Dilemmas of post-communist democ-
ratization in Poland and beyond. Communist and Post-Communist Studies, 29(3):309{330,
1996.
Eva Bernhardt and Ewa Fratczak. Family status and subjective well-being: Comparing Poland
and Sweden. Presented at the International Union for the Scientic Study of Population Con-
ference, Tours, France, 2005.
David G. Blanch
ower. Unemployment, well-being, and wage curves in Eastern and Central
Europe. Journal of the Japanese and International Economies, 15(4):364{402, 2001.
David G. Blanch
ower and Richard B. Freeman. The attitudinal legacy of communist labor
relations. Industrial and Labor Relations Review, 50:438{459, 1997.
91
David G. Blanch
ower and Andrew Oswald. Well-being over time in Britain and the USA. Journal
of Public Economics, 88:1359{1386, 2004.
Robert Blauner. Alienation and Freedom: The Factory Worker and His Industry. University of
Chicago Press, Chicago, 1964.
Friedel Bolle and Simon Kemp. Can we compare life satisfaction between nationalities? Evaluating
actual and imagined situations. Social Indicators Research, 90:397{408, 2009.
Michael Bruno. Stabilization and reform in Eastern Europe: A preliminary evaluation. Sta
Papers - International Monetary Fund, 39(4):741{777, 1992.
Donald H. Brush, Michael K. Moch, and Abdullah Pooyan. Individual demographic dierences
and job satisfaction. Journal of Organizational Behavior, 8(2):139{155, 1987.
Angus Campbell, Philip E. Converse, and Willard L. Rodgers. The Quality of American Life:
Perceptions, Evaluations, and Satisfactions. Russell Sage, New York, 1976.
Nauro F. Campos and Fabrizio Coricelli. Growth in transition: What we know, what we don't,
and what we should. Journal of Economic Literature, 40:793{836, 2002.
Nauro F. Campos and Roman Horv ath. Reform redux: Measurement, determinants and reversals.
Institute for the Study of Labor Discussion Paper No. 2093, 2006.
Andrew E. Clark. Job satisfaction and gender: Why are women so happy at work? Labour
Economics, 4:341{372, 1997.
Andrew E. Clark. Work, jobs, and well-being across the millennium. In Ed Diener, John Helliwell,
and Daniel Kahneman, editors, International Dierences in Well-Being. Oxford University
Press, Oxford, forthcoming.
Andrew E. Clark, Andrew J. Oswald, and Peter B. Warr. Is job satisfaction U-shaped in age?
Journal of Occupational and Organizational Psychology, 69:57{81, 1996.
Zoa Czepulis-Rutkowska. The Polish pension system and its problems. In Katharina M uller,
Andreas Ryll, and Hans-J urgen Wagener, editors, Transformation of Social Security: Pensions
in Central-Eastern Europe, pages 143{158. Physica-Verlag, Heidelberg, 1999.
Ertu grul Delikta s and Mehmet Balcilar. A comparative analysis of productivity growth, catch-up,
and convergence in transition economies. Emerging Markets Finance and Trade, 41(1):6{28,
2005.
Mathias Dewatripont and G erard Roland. Economic reform and dynamic political constraints.
Review of Economic Studies, 59:703{730, 1992.
Christyn E. Dolbier, Judith A. Webster, Katherine T. McCalister, Mark W. Mallon, and Mary A.
Steinhardt. Reliability and validity of a single-item measure of job satisfaction. American
Journal of Health Promotion, 19:194{198, 2005.
Nick Donovan and David Halpern. Life Satisfaction: The State of Knowledge and Implications
for Government. Cabinet Oce analytical paper; REVIEW, London, 2002.
Richard A. Easterlin. Life Cycle Happiness and Its Sources: Intersections of Psychology, Eco-
nomics, and Demography. Journal of Economic Psychology, 27(4):463{482, 2006.
Richard A. Easterlin. Lost in transition: Life satisfaction on the road to capitalism. Journal of
Economic Behavior & Organization, 71:130{145, 2009.
92
Richard A. Easterlin and Anke C. Plagnol. Life satisfaction and economic conditions in East and
West Germany pre- and post-unication. Journal of Economic Behavior & Organization, 68
(3-4):433{444, 2008.
EBRD. Transition Report Update 1997. European Bank for Reconstruction and Development,
London, 1997.
EBRD. Transition Report Update 1998. European Bank for Reconstruction and Development,
London, 1998.
Economic Commission for Europe. Economic Survey of Europe, No. 1. United Nations, New York
and Geneva, 2003.
Michael Ellman. The political economy of transformation. Oxford Review of Economic Policy, 13
(2):23{32, 1997.
Stanley Fischer and Alan Gelb. The process of socialist economic transformation. The Journal
of Economic Perspectives, 5(4):91{105, 1991.
Stanley Fischer, Ratna Sahay, and Carlos A. V egh. Stabilization and growth in transition
economies: The early experience. The Journal of Economic Perspectives, 10(2):45{66, 1996.

Eva Fodor. Women at Work: The Status of Women in the Labour Markets of the Czech Republic,
Hungary and Poland. United Nations Institute for Social Development Occasional Paper 3,
Geneva, 2005.
Richard B. Freeman. Job satisfaction as an economic variable. American Economic Review, 68:
135{141, 1978.
Bruno S. Frey and Alois Stutzer. Happiness and Economics: How the Economy and Institutions
Aect Human Well-Being. Princeton University Press, Princeton, N.J., 2002.
Paul Frijters, John P. Haisken-DeNew, and Michael A. Shields. Money does matter! Evidence from
increasing real income and life satisfaction in East Germany following reunication. American
Economic Review, 94(3):730{740, 2004a.
Paul Frijters, Michael A. Shields, and John P. Haisken-DeNew. Investigating the patterns and de-
terminants of life satisfaction in Germany following reunication. Journal of Human Resources,
39(3):649{674, 2004b.
Paul Frijters, Ingo Geishecker, John P. Haisken-DeNew, and Michael A. Shields. Can the large
swings in Russian life satisfaction be explained by ups and downs in real incomes? Scandinavian
Journal of Economics, 108(3):433{458, 2006.
Krzysztof Gorlach and Patrick H. Mooney. Defending class interests: Polish peasants in the rst
years of transfomation. In John Pickles and Adrian Smith, editors, Theorising Transition: The
Political Economy of Post-Communist Transformations, pages 262{283. Routledge, London and
New York, 1998.
Jaroslaw G orniak. Poverty in transition: Lessons from Eastern Europe and Central Asia. In
Alejandro Grinspun, editor, Choices for the Poor: Lessons from National Poverty Strategies,
pages 145{172. United Nations, New York, 2000.
Carol Graham and Stefano Pettinato. Frustrated achievers: Winners, losers and subjective well-
being in new market economies. The Journal of Development Studies, 38(4):100{140, 2002.
93
Carol Graham, Andrew Eggers, and Sandip Sukhtankar. Does happiness pay? An exploration
based on panel data from Russia. Journal of Economic Behavior & Organization, 55(3):319{342,
2004.
Francis Green and Nicholas Tsitsianis. An investigation of national trends in job satisfaction in
Britain and Germany. Bristish Journal of Industrial Relations, 43(3):401{429, 2005.
Irena Grosfeld and Claudia Senik. The emerging aversion to inequality: Evidence from Poland
1992-2005. Institute for the Study of Labor Discussion Paper No. 3484, 2008.
Daniel S. Hamermesh. Economic aspects of job satisfaction. In Orley C. Ashenfelter and Wallace E.
Oates, editors, Essays in Labor Market Analysis, pages 53{72. John Wiley, New York, 1977.
Daniel S. Hamermesh. The changing distribution of job satisfaction. Journal of Human Resources,
36:1{30, 2001.
James K. Harter, Frank L. Schmidt, and Theodore L. Hayes. Business-unit-level relationship
between employee satisfaction, employee engagement, and business outcomes: A meta-analysis.
Journal of Applied Psychology, 87:268{279, 2002.
Oleh Havrylyshyn. Divergent Paths in Post-Communist transformation: Capitalism for All or
Capitalism for the Few. Palgrave Macmillan, New York, 2006.
Bernd Hayo. Happiness in transition: An empirical study of Eastern Europe. Economic Systems,
31:204{221, 2007.
Bernd Hayo and Wolfgang Seifert. Subjective economic well-being in Eastern Europe. Journal of
Economic Psychology, 24(3):329{348, 2003.
John F. Helliwell. How's life? Combining individual and national variables to explain subjective
well-being. Economic Modelling, 20(2):331{360, 2003.
Frederick Herzberg, Bernard Mausner, and Barbara Snyderman. The Motivation to Work. John
Wiley, New York, 1957.
Sonya Human and Stanley R. Johnson. Re-evaluation of welfare changes during the transition
in Poland. Post-Communist Economies, 14(1):31{46, 2002.
International Labor Organization. ILO/KILM Key Indicators of the Labour Market Programme,
5th edition. http://www.ilo.org/public/english/employment/strat/kilm/download.htm (re-
trieved on April 10, 2009), 2009.
Timothy A. Judge and Shinichiro Watanabe. Another look at the job satisfaction - life satisfaction
relationship. Journal of Applied Psychology, 6:939{948, 1993.
Timothy A. Judge, Carl J. Thoresen, Joyce E. Bono, and Gregory K. Patton. The job satisfaction
- job performance relationship: A quantitative and qualitative review. Psychological Bulletin,
127:376{407, 2001.
Hendrik J urges. Age, cohort, and the slump in job satisfaction among West German workers.
Labour, 17(4):489{518, 2003.
Kahneman, Daniel, Ed Diener, and Norbert Schwarz, eds. Well-being: Foundations of Hedonic
Psychology. Russell Sage Foundation Press, New York, 1999.
V aclav Klaus. The ten commandments of a systematic reform revisited. IFC Fourth Annual
Lecture, Washington, DC, 1995.
94
Lena Kolarska-Bobi nska. Civil society and social anomy in Poland. Acta Sociologica, 33(4):
277{288, 1990.
Grzegorz W. Ko lodko. The Polish Alternative: Old Myths, Hard Facts and New Strategies in the
Successful Polish Transformation. UNU/WIDER, Helsinki, 1997.
Grzegorz W. Ko lodko. From Shock to Therapy: The Political Economy of Postsocialism Trans-
formation, chapter Transition from Socialism: Where to?, pages 9{47. Oxford University Press,
Oxford, 2000.
J anos Kornai. The Socialist System: The Political Economy of Communism. Princeton University
Press, Princeton, NJ, 1992.
J anos Kornai. The great transformation of Central and Easter Europe. Economics of Transition,
14(2):207{244, 2006.
Irena E. Kotowska. Discrimination against women in the labor market in Poland during the
transition to a market economy. Social Politics, 2:76{90, 1995.
Thomas Lange and Yannis Georgellis. Revisiting the union - job satisfaction paradox: Empirical
evidence from Eastern and Western European labor markets. Warsaw International Economic
Meeting Presentation, 2007.
Richard Layard. Happiness, Lessons from a New Science. The Penguin Press, New York, 2005.
Orsolya Lelkes. Tasting freedom: Happiness, religion and economic transition. Journal of Eco-
nomic Behavior & Organization, 59(2):173{194, 2006a.
Orsolya Lelkes. Knowing what is good for you: Empirical analysis of personal preferences and
the \objective good". Journal of Socio-Economics, 35(2):285{307, 2006b.
Louis L evy-Garboua and Claude Montmarquette. Reported job satisfaction: What does it mean?
Journal of Socio-Economics, 33:135{151, 2004.
Edwin A. Locke. The nature and causes of job satisfaction. In Marvin Dunnette, editor, Handbook
of Industrial and Organizational Psychology, pages 1297{1350. Rand McNally College Pub. Co.,
Chicago, 1976.
Xavier Maret and Gerd Schwartz. Poland: Social protection and the pension system during the
transition. International Social Security Review, 47(2):51{69, 1994.
Tomasz Mickiewicz and Janice Bell. Unemployment in Transition: Restructuring and Labour
Markets in Central Europe. Hardwood Academic Publishers, Amsterdam, 2000.
Jacek Moskalewicz. Alcohol in the countries in transition: The Polish experience and the wider
context. Contemporary Drug Problems, 27(3):561{592, 2000.
Jacek Moskalewicz, Bogdan Wojtyniak, and Daniel Rabczenko. Alcohol as a cause of mortality
in societies undergoing rapid transition to market economy. In Giovanni A. Cornia and Re-
nato Panicci a, editors, The Mortality Crisis in Transitional Economies, pages 83{104. Oxford
University Press, Oxford, 2000.
Ceema Namazie and Peter Sanfey. Happiness and transition: The case of Kyrgyzstan. Review of
Development Economics, 5(3):392{405, 2001.
Ellen Nolte, Vladimir Shkolnikov, and Martin McKee. Changing mortality patterns in East and
West Germany and Poland.II: Short-term trends during transition and in the 1990s. Journal
of Epidemiology and Community Health, 54:899{906, 2000.
95
Claus Oe. Varieties of Transition: The East European and East German Experience. MIT Press,
Cambridge, 1997.
Dimiter Philipov and J urgen Dorbritz. Developmental Idealism and Family and Demographic
Change in Central and Eastern Europe. European Demographic Research Papers, No. 3, Vienna
Institute of Demography of the Austrian Academy of Sciences, 2007.
Ryszard Rapacki. Public expenditures in Poland: Major trends, challenges and policy concerns.
Presented at the UN DESA Workshop on Financial Management and Accountability in the Face
of Globalization, Rome, Italy, 2001.
Jerry D. Rogers, Kenneth E. Clow, and Toby J. Kash. Increasing job satisfaction of service
personnel. Journal of Service Management, 8:14{26, 1994.
J. Barkley Rosser and Marina V. Rosser. Comparative Economics in a Transforming World
Economy, 2nd ed., chapter Poland: The Peril and Promise of Shock Therapy, pages 337{362.
MIT Press, London, 2004.
Jacques Rupnik. Eastern Europe: The international context. Journal of Democracy, 11(2):115{
129, 2000.
Jan Rutkowski. High skills pay o: The changing wage structure during economic transition in
Poland. Economics of Transition, 4(1):89{112, 1996.
Jerey Sachs. Electoral Tremors in Poland. Washington Post, page A23, November 28 1990.
Jerey Sachs. The economic transformation of Eastern Europe: The case of Poland. Economics
of Planning, 25:5{19, 1992.
Jerey Sachs. Poland's Jump to a Market Economy. MIT Press, Cambridge, 1994.
Peter Sanfey and Utku Teksoz. Does transition make you happy? Economics of Transition, 15
(4):707{731, 2007.
Willem E. Saris. What in
uences subjective well-being in Russia? Journal of Happiness Studies,
2(2):137{146, 2001.
Claudia Senik. When information dominates comparison: A panel data anlysis using Russian
subjective data. Journal of Public Economics, 88:2099{2123, 2004.
Hetan Shah and Nic Marks. A Well-Being Manifesto for a Flourishing Society. New Economics
Foundation, London, 2004.
Alfonso Sousa-Poza and Andr es A. Sousa-Poza. Taking another look at the gender/ job-
satisfaction paradox. KYKLOS, 53:135{152, 2000a.
Alfonso Sousa-Poza and Andr es A. Sousa-Poza. Well-being at work: A cross-national analysis of
the levels and determinants of job satisfaction. Journal of Socio-Economics, 29:517{538, 2000b.
Paul E. Spector. Job Satisfaction: Application, Assessment, Causes, and Consequences. Sage,
London, 1997.
David Stuckler, Lawrence King, and Martin McKee. Mass privatization and the post-communist
mortality crisis: A cross-national analysis. The Lancet, 373:399{407, 2009.
Jan Svejnar. Transition economies: Performance and challenges. Journal of Public Economics,
16(1):3{28, 2002.
96
Arland Thornton and Dimiter Philipov. Demographic Consequences of Economic Transition in
Countries of Central and Eastern Europe. Population Studies, No. 39, Council of Europe, 2003.
Ming-Chang Tsai. Market openness, transition economies and subjective wellbeing. Journal of
Happiness Studies, 10:523{539, 2009.
UNICEF. A Decade of Transition. The MONEE Project: CEE/ CIS/ Baltics. Regional Monitoring
Report No. 8, Florence, Italy: UNICEF, 2001.
UNICEF. TransMONEE 2009 Database. UNICEF Regional Oce for CEE/CIS, Geneva, 2009.
Jir Vecern k. Skating on thin ice: A comparison of work values and job satisfaction in CEE and
EU countries. International Journal of Comparative Sociology, 44:444{471, 2003.
Ruut Veenhoven. Are the Russians as unhappy as they say they are? Journal of Happiness
Studies, 2(2):111{136, 2001.
Ruut Veenhoven. Why social policy needs subjective indicators. Social Indicators Research, 58
(1):33{45, 2002.
Veenhoven, Ruut. Happiness in Nations, Subjective Appreciation of Life in 56 Nations 1946-1992.
Erasmus University, Rotterdam, 1993.
Victor H. Vroom. Work and Motivation. Wiley, New York, 1964.
Linda J. Waite. Does marriage matter? Demography, 32(4):483{507, 1995.
Linda J. Waite and Evelyn L. Lehrer. The benets from marriage and religion in the United
States: A comparative analysis. Population and Development Review, 29(2):255{275, 2003.
John P. Wanous, Arnon E. Reichers, and Michael J. Hudy. Overall job satisfaction: How good
are single-item measures? Journal of Applied Psychology, 82:247{252, 1997.
Thomas White and Paul E. Spector. An investigation of age-related factors in the age-job-
satisfaction relationship. Psychology and Aging, 2(3):261{265, 1987.
Bogdan Wojtyniak, Jacek Moskalewicz, Jakub Stokwiszewski, and Daniel Rabczenko. Gender-
specic mortality associated with alcohol consumption in Poland in transition. Addiction, 100:
1779{1789, 2005.
World Bank. Poland: Toward a Fiscal Framework for Growth. Report No. 25033-POL, Washing-
ton, DC, 2003.
World Bank. World Development Indicators. http://go.worldbank.org/IW6ZUUHUZ0, 2009.
James D. Wright and Richard F. Hamilton. Work satisfaction and age: Some evidence for the
\job change" hypothesis. Social Forces, 56:1140{1158, 1978.
WVS. World Values Survey 1981-2008 Ocial Aggregate v.20090901. World Val-
ues Survey Association. Aggregate File Producer: ASEP/JDS, Madrid, 2009. URL
http://www.worldvaluessurvey.org.
Gerald Zeitz. Age and work satisfaction in a government agency: A situational perspective.
Human Relations, 43(5):419{438, 1990.
Anke C. Zimmermann and Richard A. Easterlin. Happily ever after? Cohabitation, marriage,
divorce, and happiness in Germany. Population and Development Review, 32(3):511{528, 2006.
Robert Zuzowski. Political Change in Eastern Europe Since 1989: Prospects for Liberal Democracy
and a Market Economy. Praeger, Westport, London, 1998.
97
Appendix A
Chapter 2
A.1 World Values Survey (WVS) and International Social
SurveyProgramme(ISSP)\WorkOrientations"module
Job satisfaction:
WVS: Overall, how satised or dissatised are you with your job? \1" means dissatised,
\10" means satised.
ISSP: How satised are you in your (main) job?
{ [1] Completely satised
{ [2] Very satised
{ [3] Fairly satised
{ [4] Neither satised nor dissatised
{ [5] Fairly dissatised
{ [6] Very dissatised
{ [7] Completely dissatised
The answer options were recoded so that [1] Completely dissatised and [7] Completely
satised.
Education:
WVS: At what age did you (or will you) complete your full time education, either at school
or at an institution of higher education? Please exclude apprenticeships. (ACE)
The variable was truncated to values between 7 and 23, with a higher value taken to imply a
higher level of education.
ISSP: Respondent's education (categories)
{ [1] None
{ [2] Incomplete primary
{ [3] Primary completed
{ [4] Incomplete secondary
98
{ [5] Secondary completed
{ [6] Incomplete university
{ [7] University completed
The dummy variables for education were created as follows:
\Less than high school" is equal to one if education level is [1] None [2] Incomplete primary
[3] Primary completed or [4] Incomplete secondary
\High school" is equal to one if education level is [5] Secondary completed or [6] Incomplete
university
\University" is equal to one if education level is [7] University completed.
Marital status:
The dummy variables were created as follows:
\married" equal to one if married or living together as married
\single" equal to one if single/ never married
\divorced/ separated widowed" equal to one if divorced, separated, or widowed.
Employment level:
WVS: Are you employed now or not?
[1] Full time
[2] Part time
[3] Self employed
[4] Retired
[5] Housewife
[6] Student
[7] Unemployed
[8] Other
Only those in categories [1] through [3] were kept in the sample, and the respective employment
level dummy variables were created.
ISSP:
Respondent: current employment status { current economic position, main source of living
[1] Full-time employed, main job
[2] Part-time employed, main job
[3] Less than part-time
99
[4] Helping family member
[5] Unemployed
[6] Student, school, education, vocational training
[7] Retired
[8] Housewife/man
[9] Permanently disabled
[10] Other, not in labour force
In your (main) job are you an employee or self-employed?
Self-employed
Work for someone else
The dummy variables for employment level are: \Self employed" if the respondent is self-employed;
\Full time" if the respondent is an employee and [1] Full-time employed, main job; \Part time" if
the respondent is an employee and [2] Part-time employed, main job or [3] Less than part-time.
All others were excluded from the sample.
Occupation:
WVS: In which profession/occupation do you or did you work? If more than one job, the main
job? What is/was your job there?
[11] Employer/manager of establishment with 500 or more employed
[12] Employer/manager of establishment with 100 or more employed
[13] Employer/manager of establishment with 10 or more employed
[14] Employer/manager of establishment w. less than 500 employed
[15] Employer/manager of establishment w. less than 100 employed
[16] Employer/manager of establishment with less than 10 employed
[21] Professional worker
[22] Middle level non-manual oce worker
[23] Supervisory Non manual -oce worker
[24] Junior level non manual
[25] Non manual-oce worker
[31] Foreman and supervisor
[32] Skilled manual
[33] Semi-skilled manual worker
100
[34] Unskilled manual
[41] Farmer: has own farm
[42] Agricultural worker
[51] Member of armed forces
[61] Never had a job
[81] Other
The dummy variables for occupation are: \white collar" equal to one if [11] Employer/manager of
establishment with 500 or more employed through [25] Non manual-oce worker, or [51] Member
of armed forces; \blue collar" equal to one if [31] Foreman and supervisor through [42] Agricul-
tural worker, or [81] Other; \never had job" equal to one if [61] Never had a job.
ISSP: The occupational categories are based on the ISCO - International Code 1988
The dummy variables for occupation are: \white collar" equal to one if the ISCO - International
Code 1988 is between 1000 and 6000; \blue collar" equal to one if the ISCO - International Code
1988 is less than 1000 or greater than 6000.
Scale of incomes:
WVS: Income categories range from [1] Lower step to [10] Tenth step.
ISSP: Income quintiles.
Freedom of decision at work:
WVS: How free are you to make decisions in your job? \1" means None at all, \10" means A
great deal
Job characteristics:
ISSP:
For each of these statements about your (main) job, please tick one box to show how much you
agree or disagree that it applies to your job:
My job is secure
My income is high
My opportunities for advancement are high
My job is interesting
I can work independently
In my job I can help other people
My job is useful to society
Answer options:
[1] Strongly agree
101
[2] Agree
[3] Neither agree nor disagree
[4] Disagree
[5] Strongly disagree
The answer options were recoded so that [1] Strongly disagree and [5] Strongly agree.
Now some more questions about your working conditions. Please tick one box for each item below
to show how often it applies to your work. How often...
do you come home from work exhausted?
do you have to do hard physical work?
do you nd your work stressful?
do you work in dangerous conditions?
Answer options:
[1] Always
[2] Often
[3] Sometimes
[4] Hardly ever
[5] Never
Country coverage:
WVS: Estonia, Hungary, Latvia, Lithuania, Poland, the Russian Federation, Austria, Belgium,
Denmark, Finland, France, Great Britain, Iceland, Ireland, Italy, Netherlands, Portugal, Spain,
and Sweden.
ISSP: Bulgaria, Czech Republic, Hungary, the Russian Federation, Slovenia, Great Britain, Den-
mark, Portugal, Spain, and Sweden.
102
Appendix B
Chapter 3
B.1 WorldValuesSurvey(WVS)andEurobarometer(EB)
data
Life satisfaction:
WVS: All things considered, how satised are you with your life as a whole these days? \1"
means dissatised, \10" means satised.
EB: On the whole, are you very satised, fairly satised, not very satised or not at all
satised with the life you lead? Answers were recoded so that \1" means not at all satised
and \4" means very satised.
CBOS: How are your life and your family's life? The answer options were very bad, bad,
neither good nor bad, good, very good, coded so that \1" means very bad and \5" means very
good.
Education:
WVS: At what age did you (or will you) complete your full time education, either at school
or at an institution of higher education? Please exclude apprenticeships. (ACE)
The variable was truncated to values between 7 and 23, with a higher value taken to imply a
higher level of education.
EB: The education variable was built using two dierent survey questions, depending on
their availability:
{ 2001 - 2004 surveys: What is your level of education?
[1] uncompleted primary school
[2] primary school
[3] basic vocational
[4] general and technical secondary school
[5] university degree or more
{ 2001 - 2009 surveys: How old were you when you stopped full-time education? (ACE)
[0] still studying
[1] illiterate
separate categories for 7 or older
103
The dummy variables for education were created as follows:
\Primary education or less" is equal to one if education level is [1] uncompleted primary
school or [2] primary school; if education level is not available, I use ACE of 16 or less as a
criterion
\Secondary school" is equal to one if education level is [3] basic vocational or [4] general and
technical secondary school; if education level is not available, I use ACE between 17 and 22
as a criterion
\University" is equal to one if education level is [5] university degree or more; if education
level is not available, I use ACE of 23 or older as a criterion
\Still in school" is equal to one if ACE is [0] still studying
Marital status:
The dummy variables were created as follows:
\married" equal to one if married, remarried, or living together as married
\single" equal to one if single/ never married
\divorced/ separated" equal to one if divorced or separated
\widowed" equal to one if widowed
Employment status:
WVS: Are you employed now or not?
[1] Full time
[2] Part time
[3] Self employed
[4] Retired
[5] Housewife
[6] Student
[7] Unemployed
[8] Other
The dummy variables for employment status are: \employed" equal to one if [1] Full time, [2]
Part time, or [3] Self employed; \unemployed" equal to one if [7] Unemployed; \not in the labor
force (LF)" equal to one if [4] Retired, [5] Housewife, [6] Student, or [8] Other.
104
EB: What is your current occupation?
NON-ACTIVE
[1] Responsible for ordinary shopping and looking after the home, or without any current
occupation, not working
[2] Student
[3] Unemployed or temporarily not working
[4] Retired or unable to work through illness
SELF EMPLOYED
[5] Farmer
[6] Fisherman
[7] Professional (lawyer, medical practitioner, accountant, architect, etc.)
[8] Owner of a shop, craftsmen, other self-employed person
[9] Business proprietors, owner (full or partner) of a company
EMPLOYED
[10] Employed professional (employed doctor, lawyer, accountant, architect)
[11] General management, director or top management (managing directors, director gen-
eral, other director)
[12] Middle management, other management (department head, junior manager, teacher,
technician)
[13] Employed position, working mainly at a desk
[14] Employed position, not at a desk but travelling (salesmen, driver, etc.)
[15] Employed position, not at a desk, but in a service job (hospital, restaurant, police,
reman, etc.)
[16] Supervisor
[17] Skilled manual worker
[18] Other (unskilled) manual worker, servant
The dummy variables for employment status are: \employed" equal to one if [5] Farmer through
[18] Other (unskilled) manual worker, servant; \unemployed" equal to one if [3] Unemployed or
temporarily not working; \not in the labor force (LF)" equal to one if [1] Responsible for ordinary
shopping and looking after the home, or without any current occupation, not working, [2] Student,
or [4] Retired or unable to work through illness.
105
Occupation:
WVS: In which profession/occupation do you or did you work? If more than one job, the main
job? What is/was your job there?
[11] Employer/manager of establishment with 500 or more employed
[12] Employer/manager of establishment with 100 or more employed
[13] Employer/manager of establishment with 10 or more employed
[14] Employer/manager of establishment w. less than 500 employed
[15] Employer/manager of establishment w. less than 100 employed
[16] Employer/manager of establishment with less than 10 employed
[21] Professional worker
[22] Middle level non-manual oce worker
[23] Supervisory Non manual -oce worker
[24] Junior level non manual
[25] Non manual-oce worker
[31] Foreman and supervisor
[32] Skilled manual
[33] Semi-skilled manual worker
[34] Unskilled manual
[41] Farmer: has own farm
[42] Agricultural worker
[51] Member of armed forces
[61] Never had a job
[81] Other
The dummy variables for occupation are: \white collar" equal to one if [11] Employer/manager of
establishment with 500 or more employed through [25] Non manual-oce worker, or [51] Member
of armed forces; \blue collar" equal to one if [31] Foreman and supervisor through [42] Agricul-
tural worker, or [81] Other; \never had job" equal to one if [61] Never had a job.
EB: Same question as for employment status, plus a similar question asked of those not doing any
paid work currently: Did you do any paid work in the past? What was your last occupation? It
includes answer option Never did any paid work, in addition to the answer options above.
The dummy variables for occupation are: \white collar" equal to one if [7] Professional (lawyer,
medical practitioner, accountant, architect, etc.) through [15] Employed position, not at a desk,
but in a service job (hospital, restaurant, police, reman, etc.); \blue collar" equal to one if [5]
Farmer, [6] Fisherman, or [16] Supervisor through [18] Other (unskilled) manual worker, servant;
\never had job" equal to one if Never did any paid work in the additional question asked of those
not currently doing any paid work.
106
Scale of incomes:
Only available in the WVS. It ranges from [1] Lower step to [10] Tenth step.
GDP per capita:
Data from the World Bank (2009) World Development Indicators in PPP constant 2005 in-
ternational dollars (retrieved from http://data.worldbank.org/data-catalog/world-development-
indicators), except 1989 obtained by extrapolation from Economic Commission for Europe (2003),
Table B.1.
Registered unemployment:
Data from the Central Satistical Oce of Poland
(http://www.stat.gov.pl/gus/index ENG HTML.htm). 1989 value from January 1990; 1990-2008
values from December of the respective year; 2009 value from July.
In
ation:
Data from the World Bank's World Development Indicators (retrieved from
http://data.worldbank.org/data-catalog/world-development-indicators), based on consumer prices
(annual %).
Gini coecient:
Data from UNICEF (2009) TransMONEE Database, based on the distribution of income.
107

Abstract (if available)

Abstract The two essays in this dissertation share in common an attempt to study the transition process in terms of its effect on people's subjective well-being, in addition to the objective conditions traditionally emphasized in economics. The first essay focuses on the particular area of job satisfaction, while the second analyzes overall life satisfaction.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Adaptation, assets, and aspiration. Three essays on the economics of subjective well-being

PDF

Essays on health and well-being

PDF

Four essays on how policy, the labor market, and age relate to subjective well-being

PDF

Explaining well-being: essays on the socio-economic factors accompanying life satisfaction

PDF

Three essays on the credit growth and banking structure of central and eastern European countries

PDF

Essays in empirical health economics

PDF

Three essays on social policy: institutional development, and subjective well-being as a cause and consequence of labor market outcomes

PDF

Measurement error in income and consumption data and economic mobility

PDF

Essays on labor and development economics

PDF

What leads to a happy life? Subjective well-being in Alaska, China, and Australia

PDF

Armed conflict, education and the marriage market: evidence from Tajikistan

PDF

The relationship of gratitude and subjective well-being to self-efficacy and control of learning beliefs among college students

PDF

Beyond revealed preferences: how gender, relative socioeconomic status, and social norms drive happiness and behavior

PDF

The effect of guidance on learning in a web-based, media-enriched learning environment

PDF

Essays in behavioral and financial economics

PDF

Empirical studies of monetary economics

PDF

Essays in international economics

PDF

Essays in political economics

PDF

The link between the expression of negative emotion and future well-being: the role of disclosure modality in an East Asian American population

PDF

Essays on the econometrics of program evaluation

Asset Metadata

Creator Angelescu, Laura (author)

Core Title Essays on the economics of subjective well-being in transition countries

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Economics

Publication Date 08/05/2010

Defense Date 05/07/2010

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag eastern Europe,job satisfaction,OAI-PMH Harvest,Poland,subjective well-being,transition

Place Name eastern Europe (region), Poland (countries), western Europe (region)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Easterlin, Richard A. (committee chair), McArdle, John J. (committee member), Strauss, John (committee member)

Creator Email angelesc@usc.edu,lauraangelescu@yahoo.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m3280

Unique identifier UC1262142

Identifier etd-Angelescu-3907 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-367091 (legacy record id),usctheses-m3280 (legacy record id)

Legacy Identifier etd-Angelescu-3907.pdf

Dmrecord 367091

Document Type Dissertation

Rights Angelescu, Laura

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu