You are on page 1of 26

Int Rev Econ (2014) 61:127–152

DOI 10.1007/s12232-014-0201-0

RESEARCH ARTICLE

Satisfaction and comparison income in transition


and developed economies

Devrim Dumludag

Received: 26 August 2013 / Accepted: 28 March 2014 / Published online: 19 April 2014
Ó Springer-Verlag Berlin Heidelberg 2014

Abstract The main purpose of this paper is to evaluate the impact of different
types of income comparison on subjective well-being in transition countries and
developed European countries. The paper relies on the Life in Transition Survey
(European Bank for Reconstruction and Development 2011), which was conducted
in late 2010 jointly by the European Bank for Reconstruction and Development and
the World Bank. The emphasis of the paper is on income comparisons, specifically;
local comparisons and self-ranking. The main findings reveal that comparisons have
a significant impact on life satisfaction in transition countries, whereas the rela-
tionship between comparison and life satisfaction is ambiguous in developed
European countries. In transition countries, the impact of comparisons is asym-
metric: in most cases, under-performing one’s benchmark has a greater effect than
out-performing it. In transition countries, both downward and upward evaluations
have an impact on life satisfaction, while it is worthy of note that all upward
evaluations have no effect on life satisfaction in developed European countries.

Keywords Life satisfaction  Transition countries  Economic development 


Income comparisons

JEL Classification O1  D00  D6  Z00


1 Introduction

Since Easterlin’s (1974) work on relative income and subjective well-being, there
has been a substantial increase in the number of researches on the influence of

D. Dumludag (&)
Department of Economics, Marmara University, Istanbul, Turkey
e-mail: dumludag@marmara.edu.tr; d.dumludag@maastrichtuniversity.nl

D. Dumludag
Department of Economics (AE1), Maastricht University, Maastricht, The Netherlands

123
128 D. Dumludag

income comparisons on happiness. The majority of researches into income


comparisons and life satisfaction have been carried out for developed countries
due to the greater availability of data. However, research is also on the rise in
developing countries (thanks to growing number of surveys that include questions
on subjective well-being and the rising interest of scholars).
Do income comparisons have a similar impact on life satisfaction at different
levels of development? Do they perform in the same way? This paper investigates
the differences between transition countries and developed countries in Europe in
terms of the self-reported life satisfaction levels and income comparisons. Since
transition countries differ in income per capita, socioeconomic factors and political
environment, the findings of income comparisons would be useful for the
measurement of relative income literature, providing evidence on different levels
of development.
The objective of the present paper is to examine the relationship between income
comparisons and life satisfaction using a comparative approach. For this purpose,
the paper exploits Life in Transition Survey II, applied in 2010 jointly by the
European Bank for Reconstruction and Development (EBRD) and the World Bank
(WB). The first survey was applied in 2006 and included 28 transition countries,
whereas the second survey, carried out in 2010, included 29 transition countries and
five western European countries.
The study includes two different specifications to test the importance of income
comparisons on individual well-being: local comparisons (evaluation of living
conditions with parents’ situation at the same age, concern for whether household
lives were better than 4 years ago) and the impact of self-ranking (for the past,
present and future) on an economic ladder.
The econometric results reveal that for the analysis of relative income concerns,
downward evaluations of individuals have a negative effect on life satisfaction,
whereas upward evaluations have a positive impact in transition countries, and this
relationship is asymmetric. Under-performing one’s benchmark tends to have a
greater impact than out-performing it; however, this relationship is ambiguous in
developed European countries, where upward evaluations have no positive impact
on life satisfaction, while downward evaluations have a negative impact.
The paper begins with an interpretation of income comparisons, after which, the
methodology of the study, data and empirical strategy are introduced. The
correlation between income comparisons, interaction effects and life satisfaction is
then examined, and to check for robustness, regressions for men and women are run
in the sample. In the final section, the main results are discussed and conclusions are
drawn.

2 Income comparisons

The attitudes that arise from comparisons have been known since Veblen (1899),
who used the term ‘‘conspicuous consumption’’ to refer to expenditure on goods that
signal the consumer’s position in society, and Duesenberry (1949), who emphasized
the importance of income in relation to others in determining consumption and

123
Transition and developed economies 129

savings patterns over time.1 Duesenberry’s relative income hypothesis is known as a


good example of a theory that rests on a relative utility concept. Duesenberry’s
arguments were later developed by Pollak (1976), who set out a model of
interdependent preferences that depend on other people’s consumption and
examined its implications on demand behavior. Van Praag (1968), and later
Kapteyn (1977), formalized the notion of relative utility into a theory of preference
formation.2 Kapteyn et al. (1978) developed a model of interdependent welfare
functions and suggested that satisfaction with one’s possessions is determined
through a process of preference formation defined over ‘‘reference spaces’’.
According to Kapteyn, individuals evaluate their consumption according to the
ranking of their consumption in the perceived distribution of consumption. Frank
(1985a), following Hirsch (1976), analyzed consumption behavior using the term
‘‘positional goods’’, which depend strongly on comparisons. Frank (1985b), Clark
and Oswald (1996) developed a model of behavior in which individuals care about,
and respond to, the actions of others and in explaining human behavior, the method
considered interdependent utility functions rather than interpersonally independent
ones, challenging the established welfare propositions (Layard 1980).

2.1 Types of comparisons

The question of whether happiness is relative has been a hot topic in the happiness
literature since Easterlin’s (1974) work on relative income and subjective well-being.
In his seminal paper, published in 1974, Easterlin pointed out that the average self-
reported happiness in the United States stayed practically unchanged in the postwar
period although real incomes increased, and nearly doubled, during this period. The
issue gained in importance as in most nations, at a given point in time, a higher income
is positively associated with people’s happiness, yet over the life cycle, happiness
remains more or less unchanged (McBride 2001). Easterlin’s (1974, 1995) explanation
for this paradox was the relative income effect on subjective well-being.
The relative income argument is based on the idea that the impact of income or
consumption depends on changeable standards, such as those derived from
expectancies, habituation levels and comparisons. Empirical studies have turned
up evidence in favor of the relative income approach, primarily through the use of
data that supports the argument that an individual’s happiness may depend (at least
in part) on their evaluation of income, comparison and expectation.
There are several variations to the relative income approach: comparison with
others, comparison with earlier living conditions (adaptation or habituation) and
expectations. Social comparison theory (linked with Duesenberry’s relative income
hypothesis) suggests that comparison of one’s status with others play an important

1
Earlier classical economists such as Adam Smith and Karl Marx also mention relative comparisons.
Dealing with the growth of wealth Smith (1937) recognized them as a result of ‘‘passions’’, while Marx
(1849) claimed that they explained the social aspect of utility.
2
In economics, the oldest and most developed neo-cardinal methodology leading to a cardinal
representation of individual utility is the Leyden approach based on a complicated procedure that involves
an infinite number of consumption groups with a cumulative density function of a lognormal distribution.
For details see: Van Praag (1968).

123
130 D. Dumludag

role in evaluating and constructing social reality (Festinger 1954), proposing that
people compare their income with those around them (reference group). The better-
off (the gap between present status and the status of the reference group) people
perceive themselves to be relatively, the happier they feel. For instance, Campbell
et al. (1976) tested the relative income hypothesis directly using reference groups
and satisfaction within particular domains, and found a positive association between
reported satisfaction within the domain, and the gap between the respondents’
present status and the status of the reference groups, as perceived by the
respondents. Income comparisons also include comparisons with one’s status in the
past (adaptation or habituation). The memory of past observations could also affect
one’s evaluation. Adaptation theory states that happiness increases temporarily with
increases in income, but that over time, people adjust to their higher income, and
their happiness reverts back toward its original level. Aspiration level theory states
that it is the gap between one’s aspirations and achievements that determines life
satisfaction (Stutzer 2004). According to the theory, if an increase in income leads
to an equal increase in income aspirations, the magnitude of this gap will remain
constant, and so life satisfaction does not increase. Expectations are highly
dependent on the information and opinions provided by others, and Michalos’ (1985)
integrated these variations into his Multiple Discrepancy Theory, which holds that
people use several standards (such as what others have, the best experience in the past
and expectations for the first years) when evaluating their lives. He demonstrates that
the evaluation of these facts predicts happiness better when combined rather than when
assessed separately. Finally, satisfaction judgments depend on the gap between the
actual situation and the comparison benchmarks.
One way of measuring relative standings is to ask the individuals hypothetical
questions regarding their choice among alternative states or outcomes for relative
positions. People compare themselves to other people, but it is important to know with
whom such a comparison is being made. A number of earlier surveys have included
questions about reference groups such as neighbors, colleagues, school friends, former
schoolmates and relatives. The reference group comparison section generally
comprises two questions: ‘‘How important is it to compare yourself with the reference
groups below?’’ or ‘‘Whose income would you be most likely to compare your own
with?’’ and ‘‘How do you evaluate your income in regard to these reference groups?’’
Using these questions, scholars can easily analyze the impact and direction of income
comparisons and the interaction effects on life satisfaction (Mayraz et al. 2009; Clark
and Senik 2010; Goerke and Pannenberg 2013). Comparison benchmarks also contain
questions that encourage individuals to rank themselves in terms of their welfare
position in society for the past, present and future.
There are a substantial number of studies (especially for developed countries)
assessing the influence of income comparisons on happiness, and the findings reported in
these studies are consistent with the hypothesis that relative standing is as important as
the absolute level of consumption in determining individual well-being.
The datasets from developed countries provides evidence demonstrating the
impact of income comparisons on life satisfaction. For instance, using the German
Socioeconomic Panel (GSOEP), Di Tella et al. (2010) found that there is a
significant adaptation to income, and that the level of adaptation is sufficiently large

123
Transition and developed economies 131

that no significant income effect on happiness remains after the fourth year. Ferrer-i-
Carbonell (2005), using the same panel, found that the income of the reference
group was about as important as the own income for individual happiness, that
individuals are happier the larger their income is in comparison with the income of
the reference group, and that for West Germany this comparison effect is
asymmetric. Using the British Household Panel Survey (BHPS), Clark and Oswald
(1996) provided a test of the theory that happiness depends upon a comparison level
of income by using data garnered from a random sample of workers who were asked
how content they felt with their jobs. Stutzer (2004), using data and a survey based
on more than 6,000 interviews with residents of Switzerland, provided empirical
evidence for the effect of income aspirations on individual well-being and found
that a higher income aspiration reduce one’s satisfaction with life.
Examining information from the General Social Surveys of the United States
(which have for many years been interviewing people to measure their levels of
happiness) and Eurobarometers, Blanchflower and Oswald (2004) studied happiness
in the United States and Great Britain. They stated that relative income effects play
a significant role in explaining why the reported levels of well-being had declined
over the last quarter of a century in the United States; and that life satisfaction had
remained at a similar level through time in Britain. Using the General Social
Survey, McBride (2001) found a negative correlation between subjective well-being
and the individual’s reference income and the financial situation of the parents.
Luttmer (2005), using the National Survey of Families and Households in the
United States provided evidence that suggested that utility depends in part on one’s
relative position. In summary, it can be argued that income comparisons play a
significant role in developed countries.
As mentioned previously, there is a growing interest among scholars in the
subject of life satisfaction in transition countries with such questions as ‘‘What has
happened to subjective well-being as the former communist nations of Europe have
transitioned from central planning to market economies?’’, ‘‘Are people more or less
satisfied with their lives?’’ attracting the interest of scholars studying life
satisfaction in transition countries (Easterlin 2009).3
By using such datasets as the Life in Transition Survey, the Russian Longitudinal
Monitoring Survey (RLMS), the European Values Study, the European Social
Survey and the World Values Survey, scholars have analyzed the determinants of
life satisfaction and the role of relative income in transition countries. For instance,
using the World Values Survey, Sanfey and Teksoz (2007) showed that individuals
in transition economies on average record lower values of self-reported satisfaction
with life compared with those in non-transition countries. In addition, they assert
that for a smaller sample of countries, life satisfaction levels have returned close to
pre-transition levels in most cases after a dip in the mid-1990s. Veenhoven (2001)
stated that the unhappiness of Russians in the 1990s was due not to the Russian
national character, but was more to do with the troublesome transitions taking place
in Russian society. Frijters et al. (2004) reported that life satisfaction in East

3
For a detailed survey about life satisfaction and transition countries see: Selezneva (2011) and Clark
and Senik (2010).

123
132 D. Dumludag

Germany was significantly below that of West Germany throughout the first decade
after reunification; however, thanks to higher real household incomes there has been
a clear convergence that has resulted from a continual increase in life satisfaction in
East Germany up until 1999.
In regard to the relative income effect on life satisfactions in transition countries,
one of the most notable contributions was made by Senik (2004, 2008, 2009). Using
rounds 5–9 of the RLMS (1994–2000), Senik made some unusual findings:
Although basic socio-demographic variables are correlated in a quite standard way
with individual satisfaction, social comparison indicators are not. The reference
group income exerts a positive influence on individual satisfaction, which is in
contrast with other studies on the subject (Senik 2004). Using individual-level data
from a large number of countries (European Community Household Panel (ECHP)
Russian Longitudinal Monitoring Survey (RLMS), INSEE database, TARKI
Hungarian Household Panel NORBALT II survey of Estonia national representative
household survey in Poland, General Social Survey (GSS) and European Social
Survey 2003), Senik examined how self-reported subjective well-being depends on
own income and reference income, where reference income is defined as the income
of one’s professional peers.
In 2009, Senik studied the relative income effect on life satisfaction in transition
countries by using the Life in Transition Survey (EBRD 2006) conducted by the
EBRD (2006). In her study of 28 post-transition countries, Senik showed that
internal comparisons with one’s own past living standards outweighed any other
comparison benchmark. One major finding was that local comparisons (to one’s
parents, former colleagues or high school friends) are more powerful than self-
ranking in the social ladder.
Schyns (2000) analyzed the relationship between income, changes in income and
life satisfaction. Using data from a German and Russian panel study, Schyns tested
whether there was an absolute effect, a relative effect or a combination of both
effects of income on life satisfaction. Namazie and Sanfey (2001) using 1993
household survey data, analyzed self-reported measures of life satisfaction in a
transition country, Kyrgyzstan testing whether higher levels of satisfaction were
associated with greater economic well-being. There is some evidence that income
relativities, as measured by one’s perceived position on the wealth ladder, also have
a strong effect on life satisfaction.

3 Data

This paper uses the data garnered from the Life in Transition Survey conducted by
the EBRD and the World Bank in 29 transition countries and in five developed
European countries in 2010.4 The main aim of the survey was to gain a better
understanding of how people’s lives had been shaped and affected during the
4
Although the term ‘‘transition economies’’ usually covers the countries of Central and Eastern Europe
and the Former Soviet Union, this term may have a wider context. The categorization of European Bank
for Reconstruction and Development (EBRD) include central Eastern Europe and the Baltic States, south-
eastern Europe, Eastern Europe and the Caucasus and Central Asia.

123
Transition and developed economies 133

countries’ transition processes and the survey contains direct measures of both the
intensity and the direction of income comparisons. Although some comparison
questions are removed from the survey, coverage has been expanded to include five
western European ‘‘comparator’’ countries—France, Germany, Italy, Sweden and
the United Kingdom. This allows scholars to benchmark the transition region
against some advanced market economies, thereby giving a clearer perspective on
the remaining challenges facing transition countries. Using the information from the
Life in Transition Survey II, the main objective of the paper is to provide evidence
of income comparisons among transition and developed European countries, and the
correlation between consumption and life satisfaction will also be presented.
Respondents were drawn randomly using a two-stage sampling method, and with
census enumeration areas as primary sampling units and households as secondary
sampling units. The survey asked several series of general questions about the
household and expenses to a first respondent (Sects. 1, 2, 8) including the three
subjective ranking of ten scale questions and asked all the other questions (Sects. 3–7)
to the person selected at the bottom of Sect. 1 in the household. Since Section 7
included questions on life satisfaction and other control variables, it is deemed
important to follow the respondents who answered all sections. Following Senik’s
approach (2009), the observations were retained only in cases where the respondents
were the same person, which resulted in a loss of 40 % (using a sample of 23,671
rather than 38,865) of the total observations. Using a total of 22,671 observations,
split 4,573 for developed European countries and 19,098 for transition countries
(excluding Turkey), a test was made to ascertain whether income comparisons are
associated with life satisfaction.5
With the help of questions in the Life in Transition Survey, it is possible to obtain
an indication of the evaluations of the levels of life satisfaction of the individual,
based on the individual responses to the question: ‘‘All things considered, how
satisfied are you with your life as a whole these days?’’ Answers were given on a
scale ranging from 1 (dissatisfied) to 10 (satisfied). The measure of household
expenditure was ascertained from the level of logarithmic consumption per
household calculated from the survey.6
The main attitudinal questions for the local comparison were posed in the
following way: ‘‘to what extent do you agree with the following statements’’: ‘‘I
have done better in life than my parents; and ‘‘my household lives better now than it
did around 4 years ago’’. For each separate question, the respondents had to tick one
answer out of seven proposed choices: ‘‘strongly disagree/disagree/neither disagree
nor agree/agree/strongly agree/not applicable/don’t know’’.

5
For income comparison effects on life satisfaction for the Turkish case see Dumludag (2013).
6
Although information related to income was not used in the survey, expenditure may give a much better
indication of the material resources available to individuals. According to the Life In Transition Survey
(2011), standard of living is measured using a series of questions regarding household expenditure over
the past 12 months based on a comprehensive list of several items including food, beverages and tobacco;
utilities (electricity, water, gas, heating, fixed-line phone); transportation (public transportation, fuel for
car); education (including tuition, books, kindergarten expenses); health (including medicines and health
insurance); clothing and footwear; and durable goods (e.g., furniture, household appliances: TV, car, etc).

123
134 D. Dumludag

The following questions were asked for the self-ranking: ‘‘Please imagine a
10-step ladder, on the bottom of which, on the first step, stand the poorest 10 %
people in our country, and on the highest step, the tenth, stand the richest 10 % of
people in our country. On which step of the 10 is your household today?’’ ‘‘Now,
imagine the same 10-step ladder 4 years ago. On which step was your household at
that time?’’ ‘‘Finally, where on the ladder do you believe your household will be
4 years from now?’’
The survey, did not include questions that allowed a comparison with school
classmates and colleagues, although they were included in the survey in 2006 [The
statements were as follows: I have done better in life than most of my high school
mates; I have done better in life than most of the colleagues I had in around 1989
(EBRD 2006)]. On the other hand in the 2011 survey, the respondents’ expectations
for the future in regard to living standards was introduced [The statement is as
follows: and where on the ladder do you believe your household will be 4 years
from now? (EBRD 2011)].

4 Methodology

Cross-sectional models are estimated in the present study using ordinary least
squares (OLS) regressions, which treat life satisfaction as a cardinal construct, as the
results of cardinal models are more intuitive and easier to interpret than estimates
from ordinal probit models. In addition, cardinal and ordinal analyses of life
satisfaction yield, in general, similar results (Ferrer-i-Carbonell and Frijters 2004).
The dependent variable is the reported satisfaction of the individuals. This is
regressed on a number of socio-demographic characteristics, as well as income
comparison questions.
Since the main focus of this paper is the effect of income comparisons on
satisfaction with life, in all regressions there is systematical control to take into
account socio-demographic factors and the level of household consumption. The
dataset contains information on such socio-demographic control variables as gender,
age, education level, employment, marital status and household size. The gender
dummy takes the value zero if the respondent is female and one if the respondent is
male. The marital status dummy variable takes the value one if the respondent is
currently married or living with a partner and zero otherwise. The employment
variable takes the value one if the respondent is currently employed and zero
otherwise.
For the individual’s local comparison norm, the subjective ‘‘parents’ standard of
living’’ and ‘‘household lives better’’ measures are used. The comparison question
related to the parents is, ‘‘compared to your parents when they were the age you are
now, do you think your own standard of living now is: much better, somewhat
better, about the same, somewhat worse, or much worse?’’ Dropping the not
applicable and don’t know choices, the dummy variables are ‘‘better than parents’’
(taking a value 1 for the choices ‘‘much better’’ and ‘‘somewhat better’’ and zero
otherwise) and ‘‘worse than parents’’ (taking a value of 1 for the choices of
‘‘somewhat worse’’ and ‘‘much worse’’ and zero otherwise).

123
Transition and developed economies 135

For the household comparison question ‘‘my household lives better nowadays
than 4 years ago’’, two dummy variables are used: for the answer ‘‘my household
lives better’’, the dummy variable takes a value of 1 for the choices ‘‘agree’’ and
‘‘strongly agree’’, and 0 otherwise, and for the answer ‘‘my household lives worse’’,
the dummy variable takes a value of 1 for the choices ‘‘disagree’’ and ‘‘strongly
disagree’’, and 0 otherwise.
In order to evaluate the effect of income evaluations (general ranking), the
responses to the following questions are used: ‘‘Please imagine a 10-step ladder, on
the bottom of which, on the first step, stand the poorest 10 % people in our country,
and on the highest step, the tenth, stand the richest 10 % of people in our country.
On which step of the 10 is your household today?’’ ‘‘Now, imagine the same 10-step
ladder 4 years ago. On which step was your household at that time?’’ ‘‘Finally,
where on the ladder do you believe your household will be 4 years from now?’’
Concerning the income ladder question, a number of dummy variables were
created: econ rankdown is a dummy variable that takes the value 1 for people who
position themselves under the 5th step and 0 otherwise; and econ rankup is a dummy
that takes the value 1 if the respondents declare their position to be higher than the
5th step. This categorization is made also for rankings for 4 years ago (evaluation of
the past) and for 4 years later (evaluation of the future).

4.1 Comparisons and life satisfaction

The regressions in Table 1 suggest that there are significant differences between
model 2 and model 3 in regard to the comparison variables: For developed
countries, all upward evaluations are statistically insignificant. In other words,
upward evaluations (feeling better in economic terms—for present, past and the
future feeling better than the parents and believing that living standards of
household have improved) do not have a positive impact on life satisfaction. This is
in line with the strict version of the relative income approach, which states that a
downward comparison does not result in an extra increase in individual happiness.
For developed countries, in model 2, among the general ranking variables,
positioning oneself under the 5th step for both the present and for the future are
statistically significant and such downward evaluations have a negative impact on
life satisfaction. For the ‘‘household’’ and ‘‘parents’’ questions, similar to the ladder
questions, only downward evaluations are statistically significant. The coefficient of
‘‘household lives worse’’ is the highest with a negative impact of -0.620, whereas
the lowest coefficient is -0.150 for the respondents who believe that their economic
position will be below the fifth step 4 years from now. On the other hand, it cannot
be said with any certainty that downward local evaluations are more powerful than
general rankings, since the impact of positioning oneself under the fifth step on life
satisfaction is -0.509.
Model 3 reveals that for transition countries, both upward and downward
evaluations are statistically significant. Stating a position on the upper side (above
5) of the economic ladder (for the present and past) has a positive impact on life
satisfaction and vice versa. Interestingly, there is also a clear effect of individuals’
expectations about their economic situation on life satisfaction. For respondents

123
136 D. Dumludag

Table 1 OLS estimations, dependent variable: life satisfaction


Comparisons Model 1 Model 2 Model 3

All sample Developed countries Transition countries

Coefficients SE Coefficients SE Coefficients SE

General ranking
Present
Econ Rank (Up) 0.324*** 0.045 0.053 0.093 0.510*** 0.052
Econ Rank Ref. Ref. Ref.
(Middle) group group group
Econ Rank (Down) -0.604*** 0.041 20.509*** 0.089 -0.589*** 0.046
Past
Econ Rank (Up) 0.143*** 0.041 0.006 0.080 0.190*** 0.047
Econ Rank Ref. Ref. Ref.
(Middle) group group group
Econ Rank (Down) -0.199*** 0.039 0.077 0.081 -0.231*** 0.044
Future
Econ Rank (Up) 0.119*** 0.043 0.076 0.088 0.154*** 0.049
Econ Rank Ref. Ref. Ref.
(Middle) group group group
Econ Rank (Down) -0.365*** 0.042 -0.150** 0.088 -0.460*** 0.479
Local comparisons
Household
Lives better 0.196*** 0.038 0.094 0.073 0.249*** 0.045
Lives similar Ref. Ref. Ref.
group group group
Lives worse -0.459*** 0.035 -0.620*** 0.070 -0.396*** 0.040
Parents
Better than parents 0.293*** 0.035 0.096 0.067 0.327*** 0.040
Similar Ref. Ref. Ref.
group group group
Worse than parents -0.250*** 0.038 -0.442*** 0.079 -0.166*** 0.044
R2 0.340 0.210 0.280
Observations 17,270 4,196 13,073

Controls: age, age square, gender, household size, marital status, education, country dummies, employ-
ment and log household consumption
The regressions in Table 1 suggest that there is a positive relationship between satisfaction and education,
marital status, household expenditure and employment for both developed and transition countries.
Household consumption is positively correlated with reported satisfaction with life, suggesting that
consumption does raise happiness. The impact of consumption on life satisfaction is higher in both
regressions in comparison to other control variables. See ‘‘Appendix’’ Tables 7 and 8
*** p \ 0.01; ** p \ 0.05; * p \ 0.10

who expect to be above step five on the economic ladder 4 years later, the
coefficient is positive (0.154), whereas for respondents who expect to be below step
five, the coefficient is negative and higher (-0.460). Among the general ranking

123
Transition and developed economies 137

variables and all comparison variables, the most powerful comparisons are the
variables related to economic rankings for the present, with coefficients of 0.510 and
-0.589.
Similar to the general rankings, the upward and downward evaluations in local
comparisons are statistically significant for transition countries. Feeling better than
one’s parents and stating that the living standard of the household is better have a
positive impact on life satisfaction. A negative effect is observed when a person
states that his/her standards of living are worse than their parents and that the living
standards of the household have not improved.
In Model 3, in line with Duesenberry‘s approach, an asymmetry exists among the
comparison variables. Except for comparison with parents, under-performing one’s
benchmark has a greater effect than out-performing it. Unfavorable evaluations
have a more significant (negative) impact on life satisfaction than when evaluations
are favorable [except for the ‘‘parents’’ variable in Models 1 and 3, in which the
coefficient of ‘‘better than parents’’ (0.327) is higher than the coefficient of worse
than parents (-0.166)]. In Model 3 it cannot be stated with any certainty that
downward local evaluations or downward general rankings are more powerful than
the other.
The available evidence reveals two important outcomes: Income comparisons
seem to have various effects on subjective well-being at different levels of
development; plus, the direction of comparison is not the same in developed
countries and transition countries. In developed economies, only downward
evaluations are statistically significant, whereas in transition countries comparisons
are significant for both directions. The evidence represented in Models 2 and 3
raises a question about impact of comparisons among countries at different levels of
income: Is there a threshold (in terms of economic development level) for upward
evaluations? ‘‘After a certain level of development, do upward evaluations not
matter?’’
A similar question (is there a threshold for the utility of growth?) was raised by
Layard (2005), Inglehart (1997) Frey and Stutzer (2002) related to the effect of
income/consumption on life satisfaction: ‘‘If we compare countries, there is no
evidence that richer countries are happier than poorer ones; however, once a country
has over $15,000 per head, its level of happiness appears to be independent of its
income per head’’ (Layard 2005). In the Life in Transition Survey, life satisfaction
statistics confirmed that the average satisfaction is higher for developed countries
than transition countries. In the survey, those in developed European countries
report the highest life satisfaction scores and those in transition countries the lowest,
where the life satisfaction average is 5.11. Splitting transition countries into two
groups as central, east, southeast European countries and former Soviet countries, it
can be seen that the life satisfaction average for the former is 5.24 and 4.99 for the
latter. The average life satisfaction level in developed European countries is 7.01.7
At low levels of economic development, one may expect that income/
consumption would have a greater impact (once a society reaches a certain

7
For detailed information about descriptive statistics, see ‘‘Appendix’’ Tables 4 and 5; for regression
results for SEE and CEE and former Soviet countries, see ‘‘Appendix’’ Table 6.

123
138 D. Dumludag

threshold of development, economic gains bring a diminishing return in terms of


subjective well-being). Since Life in Transition Survey II was applied at a single
point in time, it is not possible to show the impact of changes in income/
consumption on life satisfaction at different levels of development; however, it is
possible to see the impact of consumption on satisfaction levels among developed
and transition countries. The Life in Transition Survey does not include a
household income question, and so for the purpose of this study, household
consumption is used as a proxy for income. There appears to be a positive
relationship between life satisfaction and (log) household consumption. Separate
regressions in Table 8 reveals that the impact of consumption is 0.875 in
developed countries and the impact of consumption on life satisfaction is 0.564 in
transition countries. When income comparisons are introduced into the models,
the consumption coefficient of both groups diminishes to 0.637 and 0.456,
respectively. On the other hand, Table 9 reveals that the consumption coefficient
is higher for individuals who place themselves under step five on the economic
ladder than for individuals who place themselves above step five. This is valid for
every country/region and confirms the statement that at a low level of income,
consumption has a greater impact on life satisfaction. The difference here is that a
low level of income is subjectively stated by the respondents rather than being
drawn from objective indicators.

4.2 Interaction effects

In order to understand the interaction effect, first, the general rankings and local
comparisons demonstrated in Table 2 are matched after which, the variables are
matched within the general rankings and within local comparisons. These
interaction effects can be seen in Table 3. These two sets of interaction groups
are not combined into one large set of interaction groups as the interaction groups in
one set may overlap too much with the other sets and hence may give rise to
problems of multicollinearity.
Generally, in all regressions upward interactions (up and better, up and up, better
and better) have a positive sign, whereas downward interactions (down and worse,
worse and worse, down and down) have a negative sign. For opposite interactions
(up and worse, down and better, up and down, worse and better), the sign varies
among regressions.
Interestingly, in regressions showing the interaction effects between general
rankings and local comparisons for developed countries, the opposite interaction
(rankup_houseworse) is statistically significant in each regression, and the sign is
negative. For individuals who feel that they are positioned above step five on the
economic ladder and at the same time believe that their household is worse off than
4 years ago, life satisfaction is negatively affected by this. In other words,
(houseworse) outweighs the (rankup) effect in the interaction. In each regression,
there is an asymmetry for (up_better) and (down_worse) interactions. The
coefficients of (down_worse) interactions are relatively very high in comparison
to (up_better) interactions. Interestingly, the (down_better) variables in each
regression are statistically insignificant for developed countries.

123
Table 2 Interaction effects, OLS estimations, dependent variable: life satisfaction

Interactions Developed countries Transition countries

Coefficients SE t value Coefficients SE t value


Regression 1 Regression 2

Present
rankup_housebetter 0.153 0.094 1.628 0.285 0.070 4.072***
rankup_houseworse 20.407 0.077 -5.268*** 20.183 0.080 -2.274**
rankdown_housebetter 0.071 0.113 0.631 0.103 0.058 1.789*
rankdown_houseworse 20.882 0.088 -9.976*** 20.763 0.042 -18.091***
Transition and developed economies

rankup_parentsbetter 0.195 0.082 2.383** 0.715 0.034 21.033***


rankup_parentsworse -0.007 0.125 -0.059 0.405 0.091 4.466***
rankdown_parentsbetter 0.001 0.090 0.0104 20.637 0.049 -13.073***
rankdown_parentsworse 20.829 0.103 -8.039*** 20.599 0.047 -12.825***
Sample 4,351 15,902
R2 0.208 0.258

Regression 3 Regression 4

Past
rankup_housebetter 0.380 0.106 3.587*** 0.592 0.074 8.017***
rankup_houseworse 20.512 0.090 -5.673*** 20.298 0.058 -5.106***
rankdown_housebetter 0.142 0.097 1.475 0.381 0.055 6.890***
rankdown_houseworse 20.696 0.097 -7.169*** 20.930 0.048 -19.531***
rankup_parentsbetter 0.167 0.083 2.020** 0.750 0.057 13.128***
rankup_parentsworse -0.097 0.117 -0.831 0.157 0.074 2.115**
rankdown_parentsbetter 0.043 0.091 0.473 -0.044 0.048 -0.901
rankdown_parentsworse 21.014 0.094 -10.777*** 20.661 0.0512 -12.743***
Sample 4,351 15,794
139

123
R2 0.197 0.203
Table 2 continued
140

Regression 5 Regression 6

123
rankup_housebetter 0.272 0.092 2.960*** 0.524 0.060 8.749***
rankup_houseworse 20.199 0.107 -1.867* 20.132 0.066 -2.016**
rankdown_housebetter 0.175 0.121 1.450 -0.074 0.077 -0.960
rankdown_houseworse 20.895 0.091 -9.866*** 20.902 0.050 -18.183***
Future
rankup_parentsbetter 0.185 0.085 2.194** 0.646 0.056 11.506***
rankup_parentsworse 20.202 0.118 -1.703* 0.282 0.072 3.902***
rankdown_parentsbetter 0.117 0.091 1.284 -0.015 0.053 -0.275
rankdown_parentsworse 20.684 0.109 -6.263*** 20.651 0.057 -11.485***
Sample 4,210 13,143
R2 0.198 0.222

All of the interactions are included in this table. The omitted categories are ‘‘similar’’ and ‘‘middle’’. Controls and notes: same as Table 1
Coefficients in bold
*** p \ 0.01; ** p \ 0.05; * p \ 0.10
D. Dumludag
Table 3 Interaction effects, OLS estimations, dependent variable: life satisfaction
Interactions Developed countries Transition countries

Coefficients SE t value Coefficients SE t value


Regression 1 Regression 2

rankup_present_rankup_past -0.143 0.132 -1.086 0.557 0.096 5.808***


rankup_present_rankdown_past -0.209 0.231 -0.904 -0.083 0.143 -0.579
rankdown_present_rankup_past 0.021 0.175 0.123 -0.003 0.090 -0.028
rankdown_present_rankdown_past -0.132 0.135 -0.972 20.501 0.062 -8.127***
rankup_present_ rankup_future 0.293 0.118 2.481** 0.662 0.083 7.975***
Transition and developed economies

rankup_present_rankdown_future 0.399 0.280 1.425 20.318 0.172 -1.851*


rankdown_present_rankup_future 20.787 0.189 -4.161*** 20.216 0.079 -2.717***
rankdown_present_rankdown_future 20.715 0.115 -6.245*** 20.927 0.062 -14.910***
rankup_past_rankup_future 0.176 0.139 1.265 -0.065 0.072 -0.897
rankup_past_rankdown_future 20.418 0.142 -2.936*** -0.057 0.081 -0.705
Rankdown_past_rankup_future 0.4761 0.148 3.215*** 0.412 0.070 5.891***
rankdown_past_rankdown_future 0.250 0.159 1.567 0.026 0.079 0.330
Sample 4,285 13,803
R2 0.172 0.246

Regression 3 Regression 4

parentsbetter_housebetter 0.306 0.078 3.939*** 0.772 0.044 17.530***


parentsbetter_houseworse 20.420 0.083 -5.035*** 20.100 0.045 -2.225**
parentsworse_housebetter 20.224 0.126 -1.776* 0.189 0.076 2.500**
parentsworse_houseworse 21.094 0.085 -12.888*** 20.995 0.041 -24.512***
Sample 4,401 16,106
R2 0.174 0.156

All of the interactions are included in this table. The omitted categories are ‘‘similar’’ and ‘‘middle’’. Controls: same as Table 1
Coefficients in bold
141

123
*** p \ 0.01; ** p \ 0.05; * p \ 0.10
142 D. Dumludag

For the transition case, almost all interactions are statistically significant in each
regression—except in the (rankdown_past_parentsbetter) and (rankdown_future_-
housebetter) interactions. In regressions 2 and 4, both opposite interactions are
statistically significant, however, the sign of (rankup_houseworse) is negative,
whereas the sign of (rankdown_housebetter) is positive. The household variable
outweighs the general ranking in regard to the direction of interaction. Interestingly,
regression 2 reveals that the situation is somewhat different for parents. Here,
(rankup_parentsworse) has a positive sign, whereas (rankdown_parentsbetter) has a
negative sign. In other words, the general ranking outweighs the parents in regard to
the direction of interaction.
Table 3 demonstrates the interactions within the comparison groups. Regres-
sions 1 and 2 reveal an interaction among the general ranking variables.
Regressions 3 and 4 demonstrate an interaction among local comparisons.
In regression 1, all interactions between the present and past rankings are
statistically insignificant, and there is an asymmetry between up and up (present and
future) and down and down (present and future), with the coefficient being negative
and much higher in the latter. (Rankdown_present_rankup_future) has a negative
sign, which shows that positioning oneself above step 5 on the economic ladder
(present) outweighs his/her positive expectations for the future. On the other hand,
between the past and future interactions, the results are somewhat different. In this
case, expectations for the future outweigh evaluations of the past: (rankup_pas-
t_rank_down_future) has a negative sign and (rank_down_past_rank_up_future) has
a positive sign.
All interactions among local comparisons are statistically significant for
developed countries. The (better_better) interaction has a positive sign and the
(worse_worse) interaction has a negative sign. In addition, the (worse_worse)
interaction outweighs the (better_better) interaction: The coefficient of (parents-
worse_houseworse) (-1.094) is very high when compared to the coefficient of
(parentsbetter_housebetter) (0.306). In regard to interactions with opposite direc-
tions, the downward evaluation dominates the upward evaluation: (parentsbet-
ter_houseworse) and (parentsworse_housebetter) have negative signs.
For transition countries, all of the interactions between the present and the future
rankings are statistically significant. (Up and up) interactions have positive sign
whereas (down and down) interactions have negative sign. Interaction (rankdown_-
present_rank_down_future) has the highest coefficient with a negative sign, which
means that the life satisfaction of individuals who rank themselves below step five
on the economic ladder, and at the same time expect to be below step five on the
ladder in the future, is negatively affected by this interaction. Among the opposite
interactions (up and down), the (present and future) interactions and (rankup_pas-
t_rankdown_future) are statistically significant. The sign of present and future
interactions is negative. In other words, downward evaluations outweigh upward
evaluations. On the other hand, (rankdown_past_rankup_future) has a positive
sign, which means that upward evaluation outweighs downward evaluation. If
individuals (although they rank themselves below step five on the ladder for the
past) expect to be above step five on the ladder, there is a positive effect on life
satisfaction.

123
Table 4 OLS estimations (gender), dependent variable: life satisfaction
Transition countries Regression 1 Regression 2

Women Men

Coefficients SE t value Coefficients SE t value

General ranking
Present
Econ rank (up) 0.572 0.074 7.706*** 0.447 0.074 6.075***
Econ rank (middle) Ref. group Ref. group
Econ rank (down) 20.630 0.062 -10.077*** 20.533 0.067 -7.976***
Transition and developed economies

Past
Econ rank (up) 0.195 0.065 2.997*** 0.194 0.067 2.911***
Econ rank (middle) Ref. group Ref. group
Econ rank (down) 20.227 0.060 -3.795*** 20.224 0.064 -3.517***
Future
Econ rank (up) 0.061 0.068 0.899 0.266 0.072 3.712***
Econ rank (middle) Ref. group Ref. group
Econ rank (down) 20.513 0.065 -7.865*** 20.388 0.071 -5.489***
Local comparisons
Household
Lives better 0.197 0.062 3.174*** 0.311 0.064 4.818***
Lives similar Ref. group Ref. group
Lives worse 20.456 0.055 -8.333*** 20.326 0.059 -5.488***
Parents
Better than parents 0.363 0.055 6.626*** 0.282 0.059 4.811***
Similar Ref. group Ref. group
Worse than parents 20.160 0.059 -2.693*** 20.173 0.064 -2.699***
R2 0.284 0.278
Observations 7,116 5,957
143

123
Table 4 continued
144

Developed countries Regression 3 Regression 4

Women Men

123
Coefficients SE t value Coefficients SE t value

General ranking
Present
Econ rank (up) 0.072 0.130 0.554 0.038 0.133 0.286
Econ rank (middle) Ref. group Ref. group
Econ rank (down) 20.601 0.122 -4.923*** 20.395 0.132 -2.990***
Past
Econ rank (up) -0.076 0.113 -0.672 0.101 0.114 0.894
Econ rank (middle) Ref. group Ref. group
Econ rank (down) 0.172 0.115 1.501 -0.041 0.116 -0.354
Future
Econ rank (up) 0.088 0.122 0.722 0.104 0.128 0.809
Econ rank (middle) Ref. group Ref. group
Econ rank (down) 20.266 0.121 -2.200** -0.006 0.129 -0.050
Local comparisons
Household
Lives better 0.169 0.106 1.599 0.0210 0.102 0.206
Lives similar Ref. group Ref. group
Lives worse 20.510 0.098 -5.220*** 20.732 0.102 -7.206***
Parents
Better than parents 0.101 0.094 1.076 0.091 0.097 0.936
Similar Ref. group Ref. group
Worse than parents 20.465 0.108 -4.284*** 20.395 0.115 -3.435***
R2 0.125 0.213
Observations 2,186 2,010

Controls: same as Table 1. *** p \ 0.01; ** p \ 0.05; * p \ 0.10


D. Dumludag

Coefficients in bold
Transition and developed economies 145

Similar to developed countries, all local interactions are statistically significant


for transition countries. The (worse_worse) interaction outweighs the (better_better)
interaction, but the difference is not as high as it is for developed countries: The
coefficient of (parentsworse_houseworse) is (-0.995) and the coefficient of
(parentsworse_houseworse) is (0.772). Again, different to the developed sample,
no single direction dominates the interaction variables: (parentsworse_housebetter)
has a positive sign, whereas (parentsbetter_houseworse) has a negative sign.

4.3 Robustness of the findings

In this section, separate OLS regression results for women and men are presented as
a robustness check. In both estimations, for the socio-demographic control variables
in the models, coefficients of a similar size are estimated.
In Table 4, regressions 1 and 2 show the general ranking and local comparison
effects for both men and women in transition countries, in which it can be seen that
all comparison variables are statistically significant for both sexes, with one
exception for women being that the expectation to be above step five on the
economic ladder is statistically insignificant. The directions of all variables are
similar to the general OLS results for transition countries, with upward evaluations
having positive signs and downward evaluations having negative signs. For women,
to be below step five on the economic ladder (present) has the greatest impact on life
satisfaction: The coefficient of (econ rankdown) is -0.630. This is the same for the
male sample: the coefficient of (econ rankdown) is -0.533 for men in regression 2.
Comparisons are asymmetric the downward evaluations outweigh the upward
evaluations, except for comparison with parents: in the samples (better than parents)
outweighs (worse than parents).
In Table 4, regressions 3 and 4 show the impact of comparisons on life satisfaction
for women and men in developed countries. All downward evaluations (except econ
rankdown for the past) are statistically significant and have a negative sign, whereas
all upward comparisons are statistically insignificant for both women and men. (Econ
rankdown—present) has the greatest effect (-0.601) on life satisfaction in
comparison to other variables for women. On the other hand, for men, the highest
coefficient is -0.732 with (household lives worse). In regression four, among the
general ranking variables, only (econ rankdown—present) is statistically significant
for men, whereas, all local downward evaluations are statistically significant.

5 Conclusions

This paper expands upon the work of previous papers that have examined
empirically the relationship between comparisons and subjective well-being. The
results demonstrate that people’s subjective well-being is significantly affected by
their income comparison level, controlling for the effect of consumption and other
socio-economic variables. This is a significant outcome, indicating that comparisons
in the utility function seem to matter, despite the strong emphasis of the economic
theory approach on the role of absolute consumption.

123
146 D. Dumludag

Using ordered OLS estimations, this paper reveals that all comparison variables
have a significant explanatory power on life satisfaction levels in transition
countries, with a high ranking on the social ladder having a positive impact on
individual life satisfaction, and vice versa. In addition, comparisons are asymmetric.
Unfavorable evaluations have a more important (negative) impact on life
satisfaction than when evaluations are favorable, except for evaluations of parents’
lives. In particular, the coefficients of downward ranking on the social ladder are
more powerful than the coefficients of local comparisons to parents and household.
On the other hand, for developed European countries the relationship between
comparison variables and life satisfaction is ambiguous. Among the ladder questions
(evaluation for the past present and future), only downward evaluations for the present
and the future are statistically significant, while for the comparison questions related to
household and parents, only downward evaluations are statistically significant. In other
words, the upward evaluations of an individual in terms of their economic conditions are
not statistically significant for life satisfaction.
Comparisons are asymmetric: In most cases, under-performing one’s internal
benchmark is more important than out-performing it. Unfavorable comparisons, in
most cases, are more powerful than positive ones on life satisfaction. People
experience a lower level of well-being when they find themselves to be not as
successful as their parents (in terms of living standards) and when comparisons are
unfavorable; this has a more significant (negative) impact on life satisfaction than
when comparisons are favorable.
Why do downward evaluations work rather than upward ones in developed European
countries? There is evidence relating to norms that supports the notion that higher
incomes do not lead to greater happiness because material aspirations increase with a
society’s income (Easterlin 1995). Parallel to this view, at a high level of development
upward evaluations may not lead to an increase in life satisfaction since material
aspirations are already high (or increased) as well. On the other hand, Veenhoven (1991),
providing a reply to the Easterlin hypothesis, suggested that income helps people meet
innate, universal needs and, therefore, that the relationship between money and happiness
is not based on such changeable standards as social comparison, expectancies or
adaptation at a low level of development. As those with greater income are more likely to
be able to meet their inborn needs, for example for food, safety, health and comfortable
housing, they are prone to have a greater subjective well-being, regardless of social
comparison and so forth. Therefore, in this level of development, it is expected that the
impact of income comparisons on life satisfaction is insignificant. However, this study
has revealed that within a country group with (relatively) low output levels, income
comparisons also matter. In contrast to developed countries, both upward and downward
evaluations are in effect in the transition countries, and furthermore, it is interesting to
note that upward evaluations are statistically significant and have a positive impact on life
satisfaction in transition countries, which is not the case for developed countries.

Appendix

See Tables 5, 6, 7, 8 and 9.

123
Transition and developed economies 147

Table 5 Descriptive statistics—countries


Country Life satis. SD Country weight Country Life satis. SD Country weight

Albania 5.41 1,858 2.25 Kyrgyzstan 4.67 1,566 1.84


Armenia 4.85 1,853 2.22 Latvia 4.93 1,958 3.26
Azerbaijan 4.87 1,986 2.28 Lithuania 4.74 2,102 3.14
Belarus 5.43 1,796 3.41 Moldova 4.89 2,036 3.22
Bosnia 4.99 1,801 2.49 Mongolia 5.812 2,308 2.21
Bulgaria 4.30 1,950 2.91 Montenegro 5.72 2,089 2.14
Croatia 5.84 2,273 2.88 Poland 5.84 1,923 3.93
Czech Rep. 6.05 1,801 1.84 Romania 4.43 2,146 3.06
Estonia 5.66 2,034 2.52 Russia 5.20 2,062 6.51
Macedonia 4.75 2,146 1.46 Serbia 5.17 2,181 4.12
France 6.66 2,133 4.1 Slovakia 5.92 1,688 1.91
Georgia 3.73 1,867 2.36 Slovenia 6.21 1,915 2
Germany 6.76 2,035 4.16 Sweden 8.05 1,652 3.93
Hungary 4.876 2,131 4.04 Tajikistan 5.62 2,461 1.61
Italy 6.40 1,516 2.63 Gr. Britain 7.04 2,007 5.35
Kazakhstan 5.12 2,050 2.41 Ukraine 4.54 1,899 4.27
Kosovo 4.84 1,773 0.73 Uzbekistan 5.63 1,912 2.99

Table 6 Descriptive statistics—variables


Transition countries Developed European countries

Value SD Min–Max Value SD Min–Max

Weight 0.79 1.1 0–1 0.21 0.86 0–1


Household size 2.57 1.53 1–10? 2.18 1.23 1–10?
Age 50.0 17.1 14–97 51.42 16.28 18–94
Gender 0.44 0.5 0–1 0.48 0.50 0–1
Life satisfaction 5.13 2.084 0–10 7.02 2 0–10
Health 3.34 0.95 1–5 3.83 0.87 1–5
Marital status 0.53 0.5 0–1 0.49 0.5 0–1
Education 3.24 1.44 0–6 3.29 1.69 0–6
Consumption 3.53 0.85 4.20 0.43
Income evaluation (past) 4.54 1.82 0–10 5.10 1.76 0–10
Income evaluation (present) 4.16 1.68 0–10 4.90 1.70 0–10
Income evaluation (future) 4.71 2.20 0–10 5.03 1.87 0–10
Household better 0.26 0.44 0–1 0.30 0.46 0–1
Household worse 0.48 0.5 0–1 0.38 0.49 0–1
Better than parents 0.46 0.5 0–1 0.49 0.5 0–1
Worse than parents 0.30 0.46 0–1 0.4 0.43 0–1

123
Table 7 OLS estimations, dependent variable: life satisfaction
148

Model 1 Model 2

123
CEE and SEE Former soviet

Coefficients SE t value Coefficients SE t value

General ranking
Present
Econ rank (up) 0.460 0.067 6.874*** 0.569 0.084 6.741***
Econ rank (middle) Ref. group Ref. group
Econ rank (down) 20.662 0.059 -11.242*** 20.461 0.072 -6.399***
Past
Econ rank (up) 0.145 0.057 2.549** 0.184 0.081 2.271**
Econ rank (middle) Ref. group Ref. group
Econ rank (down) 20.366 0.056 -6.540*** 0.001 0.069 0.011
Future
Econ rank (up) 0.181 0.065 2.780*** 0.181 0.076 2.375**
Econ rank (middle) Ref. group Ref. group
Econ rank (down) 20.390 0.061 -6.377*** 20.538 0.077 -7.007***
Local comparisons
Household
Lives better 0.205 0.061 3.341*** 0.287 0.066 4.328***
Lives similar Ref. group Ref. group
Lives worse 20.385 0.050 -7.673*** 20.474 0.067 -7.081***
Parents
Better than parents 0.314 0.050 6.293*** 0.352 0.067 5.280***
Similar Ref. group Ref. group
Worse than parents 20.204 0.054 -3.760*** -0.088 0.072 -1.217
D. Dumludag
Table 7 continued

Model 1 Model 2

CEE and SEE Former soviet

Coefficients SE t value Coefficients SE t value

R2 0.316 0.245
Observations 7,898 5,175

Coefficients in bold
*** p \ 0.01; ** p \ 0.05; * p \ 0.10 (controls: same as Table 1)
Transition and developed economies
149

123
150 D. Dumludag

Table 8 OLS estimations (consumption effect), dependent variable: life satisfaction


Consumption Regressions without comparison Regressions with comparison
2
Categories Coefficients SE Sample R Coefficients SE Sample R2
size size

Developed 0.875*** 0.058 4,517 0.128 0.637*** 0.060 4,196 0.210


Transition 0.564*** 0.036 17,293 0.087 0.456*** 0.038 13,073 0.280
Developed (women) 0.773*** 0.085 2,366 0.122 0.563*** 0.087 2,186 0.125
Developed (men) 0.963*** 0.078 2,151 0.139 0.706*** 0.083 2,010 0.213
Transition (women) 0.605*** 0.048 9,609 0.084 0.522*** 0.051 7,116 0.284
Transition (men) 0.518*** 0.054 7,683 0.090 0.387*** 0.056 5,957 0.278
SEE and CEE 0.548*** 0.048 9,608 0.119 0.484*** 0.048 7,898 0.316
Former soviet 0.382*** 0.062 8,015 0.062 0.185*** 0.068 5,175 0.245

Controls and notes: controls are same as Table 1. Each cell of the table corresponds to a separate
regression. It displays the regression coefficients (consumption) and standard errors, as well as the sample
size and R2 of the regression
Coefficients in bold
*** p \ 0.01; ** p \ 0.05; * p \ 0.10

Table 9 OLS estimations (consumption effect), dependent variable: life satisfaction


Consumption Regressions without comparison Regressions with comparison
2
Categories Coefficients SE Sample R Coefficients SE Sample R2
size size

Developed (high 0.708*** 0.081 1,543 0.084 0.673*** 0.084 1,479 0.115
rank)
Developed (low 0.956*** 0.113 1,741 0.117 0.836*** 0.116 1,604 0.193
rank)
Transition (high 0.3656*** 0.065 3,013 0.049 0.379*** 0.070 2,524 0.101
rank)
Transition (low 0.551*** 0.050 9,528 0.057 0.505*** 0.057 7,032 0.146
rank)
SEE and CEE 0.437*** 0.076 1,885 0.059 0.425*** 0.083 1,657 0.131
(high rank)
SEE and CEE 0.554*** 0.070 5,181 0.081 0.559*** 0.074 4,183 0.164
(low rank)
Former soviet -0.068 0.137 1,128 0.036 -0.021 0.153 867 0.068
(high rank)
Former soviet 0.381*** 0.083 4,347 0.047108 0.222*** 0.101 2,849 0.149
(low rank)

Each cell of the table corresponds to a separate regression. It displays the regression coefficients and
standard errors, as well as the sample size and R2 of the regression. The table only displays the con-
sumption variable but all the controls in Table 1 are included
Coefficients in bold
*** p \ 0.01; ** p \ 0.05; * p \ 0.10

123
Transition and developed economies 151

References

Blanchflower D, Oswald A (2004) Well-being over time in Britain and the USA. J Public Econ
88:1359–1386
Campbell A, Converse PE, Rodgers WL (1976) The quality of American life: perceptions, evaluations,
and satisfactions. Russell Sage Foundation, New York
Clark AE, Oswald A (1996) Satisfaction and comparison income. J Public Econ 61:359–381
Clark AE, Senik C (2010) Who compares to whom? The anatomy of income comparisons in Europe.
Econ J 120(544):573–594
Di Tella R, Haisken Denew J, MacCulloch R (2010) Happiness adaptation to income and to status in an
individual panel. J Econ Behav Organ 76:834–852
Duesenberry JS (1949) Income, saving and the theory of consumer behaviors. Harvard University Press,
Cambridge
Dumludag D (2013) Life satisfaction and income comparison effects in Turkey. Soc Indic Res
114(3):1199–1210
Easterlin R (1974) Does economic growth improve the human lot? In: David PA, Reder MW (eds)
Nations and households in economic growth: essays in honor of Moses Abramovitz. Academic
Press, New York
Easterlin R (1995) Will raising the incomes of all increase the happiness of all? J Econ Behav Organ
27(1):35–48
Easterlin RA (2009) Lost in transition: life satisfaction on the road to capitalism. J Econ Behav Organ
71(2):130–145
EBRD (2006) Life in Transition Survey, http://www.ebrd.com/pages/research/publications/special/
lifeintransition.shtml
EBRD (2011) Life in Transition Survey II. http://www.ebrd.com/pages/research/publications/special/
transitionII.shtml
Ferrer-i-Carbonell A (2005) Income and well-being: an empirical analysis of the comparison income
effect. J Public Econ 89:997–1019
Ferrer-i-Carbonell A, Frijters P (2004) How important is methodology for the estimates of the
determinants of happiness? Econ J 114(497):641–659
Festinger L (1954) A theory of social comparison processes. Hum Relat 7:117–140
Frank RH (1985a) The demand for unobservable and other nonpositional goods. Am Econ Rev
75:101–116
Frank RH (1985b) Choosing the right pond: human behavior and the quest for status. Oxford University
Press, New York
Frey B, Stutzer A (2002) What can economists learn from happiness research? J Econ Lit XL:402–435
Frijters P, Haisken-DeNew JP, Shields MA (2004) Money does matter! Evidence from increasing real
income and life satisfaction in East Germany following reunification. Am Econ Rev 94(3):730–740
Goerke L, Pannenberg M (2013) Direct evidence on income comparisons and subjective well-being,
SOEP papers on multidisciplinary panel data research no. 549
Hirsch F (1976) Social limits to growth. Harvard University Press, Cambridge
Inglehart R (1997) Modernization and postmodernization: cultural, economic, and political change in 43
societies, vol 19. Princeton University Press, Princeton
Kapteyn A (1977) A theory of preference formation. Ph.D. thesis, Leyden University
Kapteyn A, Van Praag BMS, Van Herwaarden FG (1978) Individual welfare functions and social
reference spaces [0165-1765]. Econ Lett 1(2):173
Layard R (1980) Human satisfactions and public policy. Econ J 90:737–750
Layard R (2005) Happiness. Lessons from a new science. Penguin Books, New York
Luttmer EFP (2005) Neighbors as negatives: relative earnings and well-being. Q J Econ 120:963–1002
Marx K (1849) Wage-labour and capital. May 2002. http://eserver.org/marx/1849-wage.labor.capital/6-
labor.and.capital.txt
Mayraz G, Wagner GG, Schupp J (2009) Life satisfaction and relative income: perceptions and evidence.
In: CEP discussion paper, no. 938. Centre for Economic Performance, London School of Economics
and Political Science, London, UK
McBride M (2001) Relative-income effects on subjective well-being in the cross-section. J Econ Behav
Organ 45(3):251–278
Michalos AC (1985) Multiple discrepancies theory (MDT). Soc Indic Res 16(4):347–413

123
152 D. Dumludag

Namazie C, Sanfey P (2001) Happiness and transition: the case of Kyrgyzstan. Rev Dev Econ
5(3):392–405
Pollak RA (1976) Interdependent preferences. Am Econ Rev 66(3):309–320
Sanfey P, Teksoz U (2007) Does transition make you happy? Econ Transit 15(4):707–731
Schyns P (2000) The relationship between income, changes in income and life-satisfaction in West
Germany and the Russian Federation: relative, absolute, or a combination of both? In: Social
indicators research series: advances in quality of life theory and research. Springer, Netherlands,
pp 83–109
Selezneva E (2011) Surveying transitional experience and subjective well-being: income, work, family.
Econ Syst 35(2):139–157
Senik C (2004) When information dominates comparison: learning from Russian subjective panel data.
J Public Econ 88(9):2099–2123
Senik C (2008) Ambition and jealousy: income interactions in the ‘Old’Europe versus the ‘New’Europe
and the United States. Economica 75(299):495–513
Senik C (2009) Direct evidence on income comparisons and their welfare effects. J Econ Behav Organ
72:408–424
Smith A (1937) The wealth of nations. Modern Library, New York
Stutzer A (2004) The role of income aspirations in individual happiness. J Econ Behav Organ
54(1):89–109
Van Praag BMS (1968) Individual welfare functions and consumer behavior. North-Holland, Amsterdam
Veblen T (1899) The theory of leisure class. Modern Library, New York
Veenhoven R (1991) Is happiness relative? Soc Indic Res 24:1–34
Veenhoven R (2001) Are the Russians as unhappy as they say they are? J Happiness Stud 2(2):111–136

123

You might also like