THE EASTERLIN ILLUSION:

Economic growth does go with greater happiness

1 2
Ruut Veenhoven * and Floris Vergunst

International Journal of Happiness and Development, Vol. 1 (4) 311-343

DOI: 10.1504/IJHD.2014.066115

ABSTRACT:
The ‘Easterlin Paradox’ holds that economic growth in nations does not buy greater
happiness for the average citizen. This thesis was advanced in the 1970s on the basis of the
then available data on happiness in nations. Later, data have disproved most of the
empirical claims behind the thesis, but Easterlin still maintains that there is no long-term
correlation between economic growth and happiness. This last claim was tested using the
time trend data available in the World Database of Happiness, which involve 1,531 data
points in 67 nations that yield 199 time-series ranging from 10 to more than 40 years. The
analysis reveals a positive correlation between GDP growth and rise of in happiness in
nations. Both GDP and happiness have gone up in most nations, and average happiness
has risen more in nations where the economy has grown the most; r = +0.20 p < 05. On
average a 1% growth in income per capita per year was followed by a rise in average
happiness on scale 0-10 of 0.0034; thus, a gain in happiness of a full point would take 60
years with an annual economic growth of 5%.

Keywords: happiness; economic growth; trend; cross national; progress.

1 INTRODUCTION

The economy has grown over the last decades in most contemporary nations and economic
growth is likely to continue in the future. Though generally welcomed, there are also
reservations about this ongoing economic growth. Over the ages, there have been warnings
against the lures of material wealth and since the 1960s there has also been growing
concern about sustainability issues. In this context, the question arises as to what
increasing wealth does to human wellbeing and that question has become a subject of
empirical research.
1 Erasmus University Rotterdam, POB 1738, 3000DR, Rotterdam, Netherlands and North-West University,
Vaal Triangle Campus, P.O. Box 1174, Vanderbijlpark 1900, South Africa
E-mail: [email protected]
*Corresponding author

2 Erasmus University Rotterdam,POB 1738, 3000DR Rotterdam, Netherlands

E-mail: [email protected]
Ruut Veenhoven & Floris Vergunst 2 The Easterlin Illusion

1.1 The Easterlin Paradox

In 1974, Easterlin (1974) published the seminal paper ‘Does economic growth improve the
human lot?’ His answer was no. This conclusion was based on the then available survey
data on happiness. Comparing within nations he saw that rich individuals are typically
happier than their poor compatriots. Yet, comparing across nations, Easterlin noted that
average happiness was not higher in rich nations than in poor nations and comparing over
time he observed no rise in average happiness in the USA between 1944 and 1970, in spite
of impressive economic growth in that period.
This finding came to be known as the ‘Easterlin Paradox’ and is commonly seen as the
start of ‘happiness economics’. It prompted a stream of scientific publications.1 A search in
Google Scholar on ‘Easterlin Paradox’ yields more than 4,500 hits. Most of these
publications are about explanations for the phenomenon, but a considerable body of
literature is on whether it really exists.
Over the years, this has resulted in a watering down of the thesis. As we will see in
more detail in Section 1.3, Easterlin (1995) first dropped the claim that there is no same-
time correlation between economic development and average happiness in nations, but
maintained that there is no over-time correlation. Current textbooks typically mention that
second version and illustrate with the case of the USA, where income rose but happiness
not. Recently, Easterlin admitted that economic growth tends to boost happiness in the
short run, but still maintains that there is no long-term effect of economic growth on
happiness in nations (10 years and more). He also added that in the long run higher rates of
economic growth do result in proportionally greater happiness (Easterlin and Angelescu,
2009; Easterlin et al., 2010).

1.2 Explanations for the paradox

In his 1974 paper, Easterlin explained the non-relation between the wealth of nations and
average happiness of its citizens in terms of interpersonal comparison, in line with the
sociological theory of ‘relative deprivation’, which Stouffer (1949) uses to explain
dissatisfaction with rank among American soldiers. In the same vein, Easterlin assumes
that satisfaction with life depends on interpersonal comparison and since reference persons
are typically compatriots, the distance to these is about the same in rich and poor countries.
Easterlin also mentions the role of rising aspiration and in this context refers to the notion
of a ‘hedonic treadmill’, a term coined by Brickman and Campbell (1971) in this days.2
Though these explanations seem plausible, it has appeared difficult to prove that they
really apply. Over time, the explanations have also lost plausibility. One reason is in
comparison theory itself, because the related thesis of shifting standards (e.g., VanPraag
and Kaptein, 1973) predicts that satisfaction will tend to the neutral level and do so both
among the poor and the rich. This is clearly not the case: average happiness is above
neutral, at least in developed nations, and the rich remain happier than the poor. At a more
basic level, Veenhoven (1991, 2009a) has cast doubt on the theory that happiness is a
matter of comparison. In his view not all satisfaction is equally dependent on social
comparison and satisfaction with domains, such as rank and income, more so than
satisfaction with life as a whole. He argues that life satisfaction depends on affective
experience rather than on cognitive comparison, and that affective experience depends on
meeting innate ‘needs’ rather than learned ‘wants’.
The paradox has also been explained by the negative effects of economic growth. One
line of argumentation is that economic growth involves high costs, such as the work it
involves and the stresses of competition (e.g., Schorr, 1993, 1999). Another negative effect
is seen in the effects of economic prosperity on life style, such as increased consumption
Ruut Veenhoven & Floris Vergunst 3 The Easterlin Illusion

of high caloric food and too much television viewing. In this line, Scitovsky (1976) makes
the case that mass-consumption does not really satisfy and Lane (2000) argues that the
attendant culture of materialism tends to blind people to more solid sources of happiness.

1.3 Tests of the reality of the paradox

Rather than seeking for explanations, some scholars have inspected the evidence for the
empirical claim. Ever more data on happiness in nations has become available over the
years and this has gradually changed the picture.

1.3.1 Comparison across nations

The first piece of evidence presented by Easterlin is that average happiness does not differ
between rich and poor nations. This claim is based on analyses of average happiness in a
dozen countries against the log of income per head. Yet, consideration of absolute income,
rather than log-income radically changes the picture and yields a correlation of +0.59
(Veenhoven, 1989). Later studies in this line using an ever greater number of nations have
also shown strong correlations and a recent study among 123 nations even found a strong
correlation between average happiness and log-income
across nations (Deaton, 2008) of an equal magnitude as the within country correlation
between individual income and happiness (Stevenson and Wolfers, 2008, 2013).

1.3.2 Comparison over time

Easterlin also used time series data on happiness and income per head to show that
economic growth in the USA was not followed by a rise in average happiness. In later
publications, he presents longer time series on the USA, which also show a pattern of
stagnant happiness in spite of rising wealth (Easterlin, 1995).
One line of criticism does not doubt these facts, but questions the interpretation. A
positive effect of economic growth on happiness may have been suppressed by other
developments. In this line, Fisher (2008) has argued that in the USA happiness has
stagnated because of a demonstrable deterioration in family life, which is unrelated to
economic development. Likewise, Phelps (2011) explains the stagnation of happiness as a
response to a decline in the percentage of altruists in the population. In a similar vein,
Veenhoven (2007) suggests that the pains of cultural change have delayed a rise in
happiness in Japan and South Korea.
Another approach is to check the empirical claim made by Easterlin as such: Is there
really no correlation between economic growth and happiness in nations? One question in
this context is whether average happiness in nations tends to remain at the same level, as
Easterlin observed in the USA, or if rising happiness is more common. Another question is
whether rising happiness goes together with economic growth.

Evidence of rising happiness in nations

Using time series data from 21 nations, Hagerty and Veenhoven (2003) note that stagnant
happiness is the exception rather than the rule, since average happiness has gone up in
most nations of the world, even in the USA, since the 1970s. In reply, Easterlin (2005)
argues that the evidence for rising happiness is still ‘fragmentary’. Yet, evidence of rising
happiness is mounting. In an analysis of the World Values Surveys, Inglehart et al. (2008)
found a pattern of rising happiness in most participating nations since the 1980s and
happiness has also increased in most of the nations where the Eurobarometer started in the
early 1970s (Vergunst, 2011; Veenhoven, 2012c).
Ruut Veenhoven & Floris Vergunst 4 The Easterlin Illusion

Evidence of link with economic growth

The last question is whether this rise in happiness is due to economic growth. Hagerty and
Veenhoven (2003) argue that rising happiness typically coincides with economic growth.
Yet, Easterlin replies that there is still no correlation between the rate of growth in GDP
and happiness, at least when the analysis limits to cases of significant change (Easterlin
and Angelescu, 2009). Recently, he has presented more data to support this view, based on
an analysis of 54 countries (Easterlin et al., 2010).
In another recent paper (Easterlin and Angelesco, 2009), Easterlin states that there are
countries where economic growth and average happiness go hand in hand, but claims that
this is a short-term effect. On that basis, he maintains that economic growth does not add
to happiness in the long run.
Research findings on this matter are mixed so far. In an analysis of 15 European nations
over the years 1973 to 2002, Bjornskov et al. (2008) found no relation of happiness with
economic growth as such, but observed growing happiness in response to accelerations of
economic growth. Graham (2011) reports a negative effect of growth rate in the last 5
years, which she calls the ‘paradox of unhappy growth’. Yet, several studies have shown
positive correlations between happiness and economic growth (e.g., Di Tella et al., 2003;
Stevenson and Wolfers, 2008). All these studies consider a limited number of nations.
The best available data of the moment are reported by Diener et al. (2012), who use
five yearly waves of the Gallup World Poll over the years 2007 to 2011 in 158 nations.
Contrary to Easterlin’s thesis they found a positive effect of economic growth on
happiness. One particular strong point of this study is that it considers both change in GDP
per capita and change in household consumption. Another strong point is that they
measured both cognitive contentment and affective experience. Remarkably, this study did
not provide evidence of a reference shift. A weak point of the study is that it covers only 5
years, so Easterlin can still maintain that economic growth does not add to happiness in the
long run.
In this context, the present study was focused on the long-term effects of economic
growth on average happiness in nations and used time-series of at least ten years.

1.4 Aims of this study

This paper is about that last surviving empirical claims of the Easterlin Paradox, that is,
version 3 as mentioned in Section 1.1. We consider whether economic growth in nations
tends to go together with rising average happiness in the long run, and if so, whether more
growths yields proportionally greater happiness. In line with Easterlin, we conceive ‘long
run’ as periods of 10 years or more. We use the latest available data on the largest
available number of nations to answer the following research questions:

1 Did average happiness in nations remain at the same level over the last decade? a Is the
number of cases of rise in happiness about the same as the number of
decline in happiness? b If happiness risen, does this tendency manifest both in the short
and the long run?

2 If happiness has risen, is its rise typically paralleled by economic growth?

a Has happiness risen more often in countries where the economy has grown than in
countries where the economy has stagnated?
b Has average happiness raised more in nations where the economy has grown the
most?
Ruut Veenhoven & Floris Vergunst 5 The Easterlin Illusion

3 If rising happiness has tended to go together with economic growth, is this correlation
similar across situations, or:
a Has the correlation existed only in the short-term and not in the long-term?
b Does the correlation exist only in poor nations, or also in rich nations?
c Is the correlation stronger in nations where economic growth is constant than in nations
where economic growth has been erratic?

4 Do the correlations between happiness and economic growth differ across measures of
happiness used?

Note the difference between questions 2a and 2b. Economic growth may add to happiness
irrespective of its size; some growth may have a similar effect on happiness as much
growth has. Tests of the Easterlin Paradox deal typically with question 2b, neglecting
question 2a. Question 3 deals with contingencies. The Easterlin Paradox may apply to
specific conditions and may not apply in other contexts. From a policy point of view it is
important to know where it applies and where not. Question 4 is a test of robustness.

2 DATA

The data used in this study were obtained from the following sources.

2.1 Data on change of average happiness in nations

The data on average happiness in nations were taken from the World Database of
Happiness (Veenhoven, 2012e). This is a ‘findings archive’ on happiness in the sense of
subjective enjoyment of one’s life as a whole.

2.1.1 World database of happiness

The archive contains research findings yielded with measures that fit this concept of
happiness as life-satisfaction. All acceptable indicators are included in the collection
‘measures of happiness’ (Veenhoven, 2012b).
Most measures are single survey questions, such as the famous item ‘taking all
together, how happy would you say you are these days, are you very happy, pretty happy
or not too happy?’ This is just one of many acceptable measures of happiness. Survey
questions have used different key words, such as ‘satisfaction with life’, and different
response options, such as numerical scales. Next to these single questions there are also
multiple questions, some of which constitute a ‘balance scale’.
This diversity of measures of happiness used in the many surveys makes it difficult to
compare scores and in particular to assess change in average happiness over time. The
different measures of happiness are therefore sorted into ‘equivalent’ kinds, that is,
questions that address happiness using the same keyword and a rating scale of the same
length.
Research findings yielded using these acceptable measures of happiness are described
in standard excerpts using standard terminology. Two kinds of findings are distinguished,
‘distributional findings’ and ‘correlational findings’. Distributional findings denote how
happy people are in a particular population and are often summarised in a measure of
central tendency, typically the mean. Correlational findings are about things that go
together with more or less happiness and summarised using measures of association, such
as Pearson’s correlation coefficient.
Ruut Veenhoven & Floris Vergunst 6 The Easterlin Illusion

Distributional findings are sorted into findings among special publics, such as elderly
persons, and findings in the general population. The findings on happiness in the general
public are further subdivided by the kind of areas from which samples were drawn, such as
‘regions’, ‘cities’ and ‘nations’. These latter findings are gathered in the collection of
‘happiness in nations’ (Veenhoven, 2012c), which we used for this research.

2.1.2 Collection happiness in nations

To date (November 2012) the collection ‘happiness in nations’ contains 5204 findings on
average happiness of the general population3 in 164 nations over the years 1946 to 2011.
These findings are sorted in three levels, one by nation, two within nations by kind of
measure used and three within measures of the same kind by year.
An example of a ‘nation page’ is presented on Appendix A. This is the case of
Argentina for which 35 distributional findings in the general public are available. These
findings are sorted in blocks of equivalent survey questions. The first block consists of
seven findings yielded by a survey question on how ‘happy’ one is, the answers to which
were rated on a four step verbal response scale. The measure codes link to the precise text
of that question and detailed information about the investigation can be found behind the
‘i’ icon.
Findings are sorted by year within each block, and this first block consists of the years
1981, 1991, 1995, 1999, 2002, 2005 and 2008. Looking at the blocks in Appendix A, we
see no clear trend in the responses to the question on happiness (measure type 111C)
between 1981 and 2008, but a gradual change to the better in the responses to questions
about life-satisfaction (measure type 121C and 122F) and the Cantril ladder (measure type
31D).

2.1.3 Identical questions

Within these blocks of equivalent questions there are still small differences in the wording
of the lead question and/or response options. These variations are marked by the last
symbol in the measure code. There are also variations in the timeframe addressed in the
question, and these are marked with the third letter code, where ‘c’ stand for ‘current’, ‘g’
for in ‘general’ and ‘u’ is used for ‘unclear’. These minor variations in the wording of
questions can result in small differences in the mean scores and could as such overshadow
the small changes in actual happiness over time. For that reason we limited our data set to
time-series based on identical questions, that is, questions with the same measure code.4
In the above mentioned case of seven questions on how ‘happy’ one is in Argentina
this meant that we considered only the five findings based on the question variant ‘a’.
Since the series of answers to question variant ‘f covered only 6 years, these were left out.

2.1.4 Transformation to common 0-10 numerical scale

We decided to use the transformed means, provided in the World Database of Happiness,
for reasons of comparability. These transformed means are expressed on a common
numerical scale ranging from 0 (low) to 10 (high). Scores on numerical response scales,
shorter than this, are linearly stretched to give a range of 0-10. Scores on scales with verbal
response options are transformed using a procedure first described by Thurstone (1927), in
which experts rate the numerical value of response options. This procedure is described in
more detail in Veenhoven (1993), chapter 7 ‘How the data are homogenised’.

2.1.5 Series
On this basis, we constructed several series of responses to identical questions on
happiness in the same nation over time. Since we focus on the long-term, we limited our
Ruut Veenhoven & Floris Vergunst 7 The Easterlin Illusion

analysis to series that covered a minimum of 10 years. We also limited the analysis to data
gathered using probability samples. If the same question had been used in several surveys
in the same year in the same country, we used the average response to that question. We
did not require that a series involved more than two data points, though most series involve
more.
This resulted in 199 time-series for average happiness in 67 nations, which together
gave 1,531 data points. The data matrix is presented in Appendix B. This database is about
twice the size of the one used in the latest test of this aspect of the Easterlin Paradox by
Stevenson and Wolfers (2008).

2.2 Data on economic growth in nations

Data on the economic performance of nations were taken from the World Bank (2012)
online database. We considered both the income per capita in the year before each data
point on happiness and the growth of that nation’s economy in the same year as a
percentage of the GDP. Additional data for Taiwan were found in the library of the IMF
(2012).
Income per capita for each of the years in each of the countries was expressed in US
dollars. Economic growth in nations was measured using the yearly growth of the GDP as
a percentage measured in local currency.
These data did not cover all the periods for which we had data on happiness. One
reason is that the World Bank does not provide data for the years before 1960. Another
reason is that some nations have only emerged recently, for example, Croatia, a nation
which was born from the death of former ‘Great’ Yugoslavia in 1991.

3 METHODS

The following methods were used to find answers to the questions set out in Section 1.4.

3.1 Assessing change of average happiness over time

Question 1 dealt with whether average happiness has typically remained at the same level,
or has risen in most nations. We answered that question in two ways. We first assessed
change in each of the 199 series of responses to the same question on happiness in the
same country. Next, we computed the average change over all series in the same country.

3.1.1 Change of average happiness in series of identical questions

We regressed happiness against year in all the 199 time series. The resulting regression
coefficients were used to indicate the yearly change in happiness in the period covered by
the series. Since happiness is expressed on range 0-10, a regression coefficient of 0.01
means a rise of 0.1 point per year, which amounts to a 1 point gain in happiness over 10
years. These yearly coefficients were used in the following ways.

Ratio of rise or decline

We first counted the number of series in which happiness had gone up and the number in
which happiness had gone down. On that basis, we assessed the ratio; a ratio greater than
1 indicates that increasing happiness is more common than decline; a ratio of 1 that rising
and declining happiness are equally frequent, and a ratio smaller than one that a decline in
happiness is the most common. The Easterlin Paradox holds that happiness tends to remain
at the same level, around which is at best some random fluctuation and this predicts a ratio
of about 1.5
Ruut Veenhoven & Floris Vergunst 8 The Easterlin Illusion

Average change coefficient

The above bi-partitions provide a view on the relative frequency of rise and decline in
happiness, but do so at the cost of loss of variation. In order to use the available variance
more fully we computed the average change over all 199 series and assessed whether that
average coefficient was positive or negative.

3.1.2 Change of average happiness in countries

Using the change coefficients in the series, we computed the average change coefficients
for each of the 67 nations. Where only one series was available, we took the change
coefficient observed in that one and when more series were available we computed the
average change score.
These change scores in nations were analysed in the same way as the change scores in
the series. First, a ratio of rise or decline in happiness was obtained and then the average
change scores were computed and we assessed the statistical significance of these scores.

3.2 Assessing economic growth

We assessed the average yearly growth of GDP per capita for each of the countries, for
each of the periods for which a series on average happiness was available. These change
coefficients ranged from -0.034 in Belarus 1990 to 2000 to +0.120 in China 2000 to 2009.
With three exceptions the coefficients were positive.

3.3 Assessing correspondence between growth of GDP and happiness

Different questions are at stake here: research question 2a deals with whether economic
growth tends to go with any change in happiness. Next, question 2b deals with whether
more economic growth is followed by a greater rise in happiness. Question 3 deals with
the possible variations on these general patterns of correlation.

3.3.1 Is there any correlation?

Question 2a serves as a first test of Easterlin’s claim that economic growth does not buy
greater happiness. To check this hypothesis we counted the number of cases in which this
thesis applied and compared it with the number of cases in which it did not.
To that end we first ordered the possible combinations of growth and decline, see the 2
x 2 table presented in Table 1. The ‘paradoxical’ cases are called ‘discordant’ and coloured
red. The cases that fit common sense are called ‘concordant’ and coloured green. We next
counted the number of cases in each category and assessed the odds ratio. If Easterlin is
right, that ratio must be about 1.

3.3.2 More cases of rising happiness in strong growing economies?

Question 2b deals with whether happiness has raised more in countries where the economy
has grown the most. One answer to this question is whether rising happiness is more
common in the fastest growing economies. A related, but not identical, question is whether
the rise of happiness is proportionally greater in the fast growing economies.

Odds of rise or decline of happiness in slow and fast economic growth

We distinguished between cases of modest economic growth and strong economic growth,
using the median of the growth percentages found.

Correlation between change scores of happiness and GDP

Next, we assessed the correlation between the rates of yearly change in GDP and
happiness. We did this for all of the 199 time series for happiness and then computed the
Ruut Veenhoven & Floris Vergunst 9 The Easterlin Illusion

average correlation over all the time series. We then considered the average change
coefficients per nation and regressed the average change in happiness against the growth of
GDP in that nation in the same period, again with a one-year interval.

Estimates of effect size

In a first analysis, we considered the average effect of a 1% change in GDP on happiness
on a range 0-10. Then we computed the average effect of a $100 rise in GDP per capita on
happiness.

3.3.3 Correspondence contingent to situations?

Question 3 deals with possible variations on this general pattern. Question 3a covers
whether the effect of economic growth on happiness is short lived or rather manifests in
the long-term. Question 3b deals with whether the effect differs across poor and rich
nations and question 3 c with whether smooth economic growth works out better on
average happiness than bumpy growth.

Short and long-term

To answer question 3, we repeated the above mentioned analyses for each of the three time
intervals considered, the short-term of 10 to 20 years, the medium term of 21 to 40 years
and the long-term of more than 40 years.

Poor and rich nations

To answer question 3b, we measured economic prosperity of nations using the GDP per
capita in US$. The data used for this analysis were again taken from the website of the
World Bank (2012). We followed the distinction made by the World Bank between: ‘low’
income’ ‘lower-middle’ income, ‘upper-middle’ income and ‘high income nations. Our
data set did not include any ‘low-income’ countries, as rated by the World Bank.

Smooth versus bumpy growth

To answer question 3c, we estimated instability (i) in economic growth by taking the
variance in yearly GDP per capita that cannot be explained by the linear trend over the
years. To that end we first regressed GDP on year, with GDP dependent and year
independent. In the case of perfect stability, all GDP scores should be on the trend line,
and the regression coefficient 1. We computed the deviation from that pattern, using the
following formula: i = 1 - R2.

3.3.4 Robust across measures of happiness?

Finally, we checked whether these patterns differed across the measures of happiness used
in these time series.

3.4 Significance
So far how we assessed average change in happiness and GNP per capita and the
correlation between these changes. A next question was whether the observed coefficients
are worth considering or not.
One way to answer that question is to consider the effect size and pick a minimum, such as
over a 10 year period, a 0.1 point change in happiness and a correlation
coefficient with economic growth of at least +0.20. In that case, our conclusion limits to
the series studied here.
Another way is to generalise beyond the observations and in that context it is common
practice to infer the probability that the correlation observed in the sample is positive,
while there is actually no correlation in the population from which this sample is drawn. In
Ruut Veenhoven & Floris Vergunst 10 The Easterlin Illusion

that context, a 95% probability is usually deemed ‘significant’. This is how Easterlin
treated the time series.
Though routinely performed, this test for significance involves making strong
assumptions that do not fully apply in this case. One such assumption is that the 199 series
provide a random sample of all possible time series in the 67 nations. Another dubious
assumption is that the 67 nations provide a random sample of all nations in the world. Still
another thing to be kept in mind is that significance depends very much on the sample size,
small effects are significant in big samples, and big effects insignificant in small samples.
Significance depends also on the dispersion in the observations and on choices made by
the investigator, with respect to the null-hypothesis, one-sided or two-sided testing, and the
probability level. All this makes tests for significance precarious.
In our view, the descriptive approach is the most informative in this case. The number
of series at hand is large and covers all we can ever have over this period. The
interpretation is straight forward, we can easily see in Appendix B where the Easterlin
Paradox has applied (coefficient 0) and where not (all the positive coefficients).
Still, we realise that many readers are accustomed to significance testing and some are
willing to buy the above mentioned perils, even when acknowledged. We therefore did
some significance tests. We tested whether the observed positive change in happiness is
more common than negative change and whether concordance with economic growth is
more common than discordance.
In these analyses, we used all available coefficients, even those based on a limited
number of data-points. A reviewer argued that we should consider only significant
coefficients. We disagree. The number of data-points in most of the separate series is too
small for a meaningful significance test and restricting to the sufficiently dense series
would lead into the selectiveness that has played Easterlin false. We also avoid his fault of
taking non-significance in too small series for evidence of non-correlation.6

4 RESULTS

Our findings contradict Easterlin’s predictions: average happiness has increased in most
nations and increased more in the nations where the economy had grown. This pattern was
found over long time spans and appeared both among poor and rich nations.

4.1 Happiness rose in most nations

Question 1 dealt with whether average happiness in nations tends to remain at the same
level over the years (as Easterlin holds), or that average happiness tends to rise (as Hagerty
and Veenhoven hold).

4.1.1 More advance than decline

Of the 199 series 67% showed a rise in happiness and 33% a decline, which resulted in a
ratio of 2.0. Likewise happiness rose in 62% of the 66 nations and declined in 38%, which
is a ratio of 1.6. See Table 2(a). This is clearly more than the ratio of about 1 that Easterlin
predicts.
The average yearly rise in happiness observed in the 199 series is +0.016. The average
rise in the 67 nations was +0.012.
These numbers may seem small at first sight, but result in a considerable improvement
in happiness in the long term. At this growth rate average happiness will rise one point on
a 0-10 scale in 70 years. Given that the actual range on this scale is between 2.5 and 8.5
(Veenhoven, 2012d), a one point rise equals a gain of 17%.
Ruut Veenhoven & Floris Vergunst 11 The Easterlin Illusion

4.1.2 In the long run

In his latest paper Easterlin argues that happiness rises in the short run only. Our data show
otherwise. Remember that we considered time-series of at least 10 years, so we deal with
long-term change in Easterlin’s use of that term. We can see from Table 2(b) that the
average change in happiness does not differ very much between the long and very long-
term and that the rise is slightly stronger in the longest term considered, that is, 40 years of
more.

4.2 Rise of happiness paralleled by economic growth

Now that we have established that average happiness has risen in most nations, the next
questions are about the relationship of this rise with economic growth. Question 2a deals
with whether there any parallel at all between increase of average happiness and GDP. If
so, question 2b is whether there is proportionality in the relationship: Has happiness
increased more in the nations were the economy has grown the most? Lastly, question 3 is
whether we are dealing with a general pattern or whether the correlation between rising
happiness and growth of the economy is contingent on specific situations.

4.2.1 More concordant than discordant combinations

Question 2a covers whether economic growth tends to be accompanied by rising
happiness. According to Easterlin, this is typically not the case and the term ‘paradox’ is
also used to denote this difference between common sense expectation and reality.
What is the most common pattern in the period observed here? Following the method
presented in Section 3.1.1, we counted the number of cases in which economic growth was
paralleled by rising happiness (concordant with common sense) and the number of cases in
which it was not7 (discordant) and next assessed the ratio of these. See Table 3(a). As one
can see, there are hardly any cases of long-term economic decline. Most of the variation is
in changes of average happiness. In this table, the number of concordant cases is 132 and
the number of discordant cases 64, which equals a ratio of 2.2. So ‘paradoxical’
combinations are clearly not the rule.

4.2.2 Greater rise of happiness in nations where the economy grew most
Question 2b, in extension to question 2a, deals with proportionality in the relationship
between increase of happiness and economic growth: has more economic growth come
together with a greater rise in happiness? This correlation was assessed in two ways.

More rise of happiness in fastest growing countries

We distinguished between slow and fast growing economies, using the median of
economic growth in our dataset. We next counted the cases of rising and declining
happiness in each of these categories, see Table 3(b). Again, we compared the number of
cases that fit the common sense expectation that more growth goes with more happiness to
the discordant situation. Once more the former outweighed the latter. The odds ratio was
2.14 and statistically significant.

Significant correlation
Considering the correlation between average economic growth in the past year and average
change of happiness we observed a correlation of +0.20. This correlation is significantly
different from zero at the 95% confidence level. A split-up of the data by length of the
time period considered showed similar correlations. See Table 3(c). Two of these three
similar correlations are statistically significant, but the third is not. Since the effect size is
Ruut Veenhoven & Floris Vergunst 12 The Easterlin Illusion

the same, this non-significance is due to the limited number of 18 observations. One-sided
tests yielded the same result.
The pattern of correlation is presented visually in the scatter plots shown in Figures
1(a), 1(b) and 1(c). In each of these figures one can see a modest correlation. There is no
clear pattern of non-linearity. So the Easterlin Paradox fails this test.

Size of the effect

The average effect of 1% economic growth on happiness in the next year is a gain of about
0.00335 points on a scale of 0-10. Whether this effect is strong of small will be discussed
in section 5.2.

4.3 Variation across situations

Question 3 deals with possible contingencies. Are we dealing with a general pattern or
does economic growth add to happiness only in specific conditions?

Little difference across time-span

The available data show a positive correlation between economic growth and happiness
over all three long-term time spans considered here. The correlation of the yearly change
rates is lowest over the shortest time span of 10-20 year periods: 0.003 compared to 0.006
over the 10-40 years periods and 0.005 in the series of 40 or more years. See Table 4(a).

Stronger in poor nations

In line with expectation, we found that the correlation between happiness and economic
growth is strongest in the poor nations. The correlation is quite strong in the 20 ‘lower
middle’ income nations in this data set and relatively small in the nations where the
income per capita is at the upper middle level and the high level [see Table 4(b)].

No less in bumpy growth

We expected that the effect of economic growth on happiness would be smaller in nations
that have gone though economic ups and downs than in nations where the economy has
grown more smoothly. This expectation was not supported by the data [see Table 4(c)].

4.4 Similar across measures of happiness

Finally, we checked whether the correlations between happiness and economic growth
differ across the measures of happiness used. To that end, we limited our examination to
cases for which at least 15 time-series were available, since the variation in the nations
involved might otherwise cloud the effect of the measures. There were four such cases in
the data set, all of which yielded yearly change correlations in the range of +0.0025 to
+0.034. Results not shown.

5 DISCUSSION

Our aim with this study was to check the claim that average happiness in nations tends to
remain at the same level in spite of economic growth. Our data do not support that claim;
we found that happiness tends to rise over time and to rise more in nations where the
economy has grown the most. This begs the question of why our findings differ from
Easterlin’s. The next question we must ask is about the observed effect size: Is this small
or substantial? We conclude that the long-term effect of economic growth on happiness is
substantial and this opens a new agenda for research.
Ruut Veenhoven & Floris Vergunst 13 The Easterlin Illusion

5.1 Why are our results different from Easterlin’s latest reading of the data?
Easterlin et al.’s (2010) latest analysis is based on data for 37 nations over time spans
varying from 12 to 34 years. The number of data points used is not reported. The analysis
reported in this paper draws on data from 67 nations and over periods running from a
minimum of 10 years to 46 years, which gave us 1,531 data points. This difference in size
of the available data pool is crucial. The law of greater numbers helped us to see a general
pattern, which Easterlin could not see.
Additionally, Easterlin uses responses to a question on financial satisfaction as the
dependent variable in his separate analysis of 17 Latin American nations. Yet, financial
satisfaction is not the same as life-satisfaction and is more likely to adjust to changed
income levels as has been shown by VanPraag and Kaptein (1973). We consistently used
the available data on happiness in the sense of life-satisfaction and in this respect our data
are better suited to the question to hand.

5.2 Is the effect of economic growth on happiness big or small?

As we have seen, 1 % economic growth was followed by a rise in average happiness in the
next year of 0.0034. This yearly gain in average happiness may appear small at first sight
and could as such be taken as a proof of the thesis that economic growth does not buy
greater happiness. Yet, economic growth is typically higher than 1% a year and small
effects amount to a considerable increase in the long-term. An annual 5% growth of the
economy will lead to a gain of a full point in average happiness on a scale 0-10 in 60
years, which equals 10% of the theoretical variation on a scale of 0-10 and 17% of the
actual variation between 2.5 and 8.5.
When expressed in effect of a $100 increase in GDP per capita the average effect of
economic growth on happiness is about 0.10 in poor countries and 0.003 in rich countries
[Table 4(b)] This means that in poor nations a gain of one point of happiness on scale 0-10
will require a $1,000 increase and in rich countries an increase of about $ 33,000.
The above figures are probably an underestimation of the real effects of economic
growth on happiness. One reason is that the law of big numbers may not have neutralised
all the random measurement error in both the measures of happiness and economic growth,
such as due to variation in data sources and data massage. This will have attenuated the
change coefficients. Another reason is that our perspective on the real effect of economic
growth is still blurred by external shocks such as the regime changes in the East European
nations after the fall of communism and the civil wars in Africa.
Seen in this perspective, the rise of happiness is comparable to the extension of
longevity in modern society, where a long process of piecemeal increments has resulted in
a doubling of our life time.
In the long-term perspective, it is also clear that the rise of happiness is unlikely to
continue forever, since an average happiness of about 8.5 is probably the maximum
possible in a country. So there will be a point where economic growth no longer adds to
greater happiness. Yet, economic development is also likely to contribute to longevity,
which seems to be less bound to a maximum. If so, economic growth will still result in a
rising number of happy life years for individuals (Veenhoven, 2005).

5.3 New research questions

The question raised by the Easterlin Paradox is why does economic growth not result in
greater happiness. Now we have seen that economic growth typically does add to average
Ruut Veenhoven & Floris Vergunst 14 The Easterlin Illusion

happiness in nations, the questions to ask are: how does economic growth add to happiness
and why does it not always do so? To answer these questions we can build on
the earlier research instigated by the Easterlin Paradox, such as on negative effects of
economic growth and adaptation processes, but take these as elements in wider a balance
of effects.
In exploring how economic growth adds to happiness a crucial question is to determine
to what extent the effect is due to increased consumption and to what extent it is due to
institutional changes that tend to go together with economic development, such as political
democracy and women’s emancipation. The recent study of Diener et al. (2012) suggests
that a great deal of the effect of economic growth on happiness can be found in material
comfort, but this is certainly not the last word.

6 CONCLUSION

Economic growth in nations does tend to go with rising happiness. Though there are cases
where happiness remains stable in spite of economic growth, these are exceptions rather
than the rule. The ‘Easterlin Paradox’ has become the ‘Easterlin Illusion’.
Ruut Veenhoven & Floris Vergunst 15 The Easterlin Illusion

REFERENCES

Bjornskov, C., Gupta, N. and Pedersen, P. (2008)

Analyzing trends in subjective well-being in 15 European countries, 1973-2002,
Journal of Happiness Studies, Vol. 9, No. 2, pp.317-330.

Brickman, P. and Campbell, D.T. (1971)

Hedonic relativism and planning the good society, in Appley, M.H. (Ed.): Adaptation
Level Theory, pp.287-302,
Academic Press, New York, USA.

Deaton, A. (2008)
Income, health and well-being around the world: evidence from the Gallup World Poll,
Journal of Economic Perspectives, Vol. 22, No. 2, pp.53-72.

Di Tella, R., MacCulloch, R.J. and Oswald, A.J. (2003)

The macroeconomics of happiness,
Review of Economics and Statistics, Vol. 85, No. 4, pp.809-827.

Diener, E., Tay, L. and Oishi, S. (2012)

Rising income and subjective wellbeing in nations,
Journal of Personality and Social Psychology, Vol. 104, No. 2, pp.267-276.

Easterlin, R.A. (1974)

Does economic growth improve the human lot?, in David, P.A. and Melvin, W.R. (Eds.):
Nations and Households in Economic Growth, pp.89-125,
Academic Press, New York, USA.

Easterlin, R.A. (1995)

Will raising the incomes of all increase the happiness of all?,
Journal of Economic Behavior and Organization, Vol. 27, No. 1, pp.35-47.

Easterlin, R.A. (2005)

Feeding the illusion of growth and happiness: a reply to Hagerty and Veenhoven,
Social Indicators Research, Vol. 74, No. 3, pp.429-443.

Easterlin, R.A. and Angelescu, L. (2009)

Happiness and Growth the World Over: Time-series Evidence on the Happiness-Income
Paradox, IZA Discussion Paper 4060.

Easterlin, R.A., Aggelescu-McVey, L., Switek, M., Sawangfa, O. and Smith-Zweig, J.

(2010)
The happiness-income paradox revisited,
Proceedings of the National Academy of Sciences, Vol. 107, No. 52, pp.22463-22468.

Fisher, C.S. (2008)

What wealth-happiness paradox? A short note on the American case,
Journal of Happiness Studies, Vol. 9, No. 2, pp.219-226.
Ruut Veenhoven & Floris Vergunst 16 The Easterlin Illusion

Graham, C. (2011)
Does money make you happier: why so much debate?,
Applied Research in Quality of Life, Vol. 6, No. 3, pp.219-239.

Hagerty, M.R. and Veenhoven, R. (2003)

Wealth and happiness revisited: growing wealth of nations does go with greater
happiness,
Social Indicators Research, Vol. 64, No. 1, pp.1-27.

International Monetary Fund (IMF) (2012)

[online] http://elibrary-data.imf.org (accessed 1-12013).

Lane, R. (2000)
The Loss of Happiness in Market Democracies,
Yale University Press, USA.

Lykken, D.T. (1999)

Happiness: What Studies on Twins Show us about Nature, Nurture and the Happiness Set-
point,
Golden books, New York, USA.

Phelps, C. (2001)
A clue to the paradox of happiness,
Journal of Economic Behavior and Organization, Vol. 45, No. 3, pp.293-300.

Schorr, J. (1993)
The Overworked American: The Unexpected Decline of Leisure,
Basic Books, New York, USA.

Schorr, J. (1999)
The Overspent American: Why We Want What We Don’t Need,
Basic Books, New York, USA.

Scitovsky, T (1976)
The Joyless Economy: The Psychology of Human Satisfaction,
Oxford University Press, New York, USA.

Stevenson, B. and Wolfers, J. (2008)

Economic Growth and Wellbeing: Reassessing the Easterlin Paradox,
Brooking Papers on Economic Activity, nr. 14282, Washington DC, USA.

Stevenson, B. and Wolfers, J. (2013)

Subjective well-being and income: is there any evidence of satiation?,
American Economic Review, Vol. 103, No. 3, pp.598-604.

Stouffer, S.A. (1949)

The American Soldier,
Princeton University Press, USA.
Ruut Veenhoven & Floris Vergunst 17 The Easterlin Illusion

Thurstone, L.L. (1927)

Psychophysical analysis,
American Journal of Psychology, Vol. 38, No. 3, pp.368-389.

VanPraag, B.M.S. and Kaptein, A. (1973)

Further evidence on the individual welfare function of income: an empirical investigation
in the Netherlands,
European Economic Review, Vol. 4, No. 1, pp.33-62.

Veenhoven, R. (1989)
National wealth and individual happiness, in Grunert, K.G. and Olander, M. (Eds.):
Understanding Economic Behavior, pp.9-32,
Kluwer Academic Press, The Netherlands.

Veenhoven, R. (1991)
Is happiness relative?,
Social Indicators Research, Vol. 24, No. 1, pp.1-34.

Veenhoven, R. (1993)
How the data are homogeneized Chapter 7 in ‘Happiness in nations: subjective
appreciation of life in 56 nations 1946-1992,
RISBO, Erasmus University Rotterdam [online]
http://worlddatabaseofhappiness.eur.nl/hap_nat/introtexts/intronat7.pdf (accessed 1-10-
2011).

Veenhoven, R. (2005)
Apparent quality of life: How long and happy people live,
Social Indicators Research, Vol. 71, Nos. 1-3, pp.61-86.

Veenhoven, R. (2007)
Quality of life in nations, as measured with how long and happy people live,
in Joop Stam and Ruut Veenhoven (Eds.): Quality of Life & Happiness of People in Japan
and The Netherlands, Vol. I, pp.16-31,
KIT Publishers, Amsterdam in cooperation with NIOD Encounters Series, ISBN 978 90
6832 4358.

Veenhoven, R. (2009a)
How do we assess how happy we are?, in Dutt, A.K. and Radcliff, B. (Eds.): Happiness,
Economics and Politics: Towards a Multi-disciplinary Approach, Chapter 3, pp.45-69,
Edward Elgar Publishers, Cheltenham UK, ISBN: 978-1-84844-093-7.

Veenhoven, R. (2009b)
International scale interval study: improving the comparability of responses to survey
questions about happiness,
in Moller, V. and Huschka, D. (Eds.): Quality of Life and the Millennium Challenge:
Advances in Quality-of-Life Studies, Theory and Research,
Social Indicators Research Series, Vol. 35, pp.45-58, Springer.
Ruut Veenhoven & Floris Vergunst 18 The Easterlin Illusion

Veenhoven, R. (2012a)
Bibliography of Happiness World Database of Happiness,
Erasmus University Rotterdam, Netherlands [online]

Veenhoven, R. (2012c)
Happiness in Nations World Database of Happiness,
Erasmus University Rotterdam [online]

Veenhoven, R. (2012d) Average Happiness in 149 Nations 2000-2009

World Database of Happiness,
Erasmus University Rotterdam, Netherlands [online]

Veenhoven, R. (2012e)
World Database of Happiness: Archive of Research Findings on Subjective Enjoyment of
Life,
Erasmus University Rotterdam, Netherlands [online]

Vergunst, E.F. (2011)

Economische groei en geluk in landen (Economic Growth and Happiness in Nations,
Thesis, Erasmus University Rotterdam.

World Bank (2012)

[online] http://data.worldbank.org (accessed 1-10-2012).
Ruut Veenhoven & Floris Vergunst 19 The Easterlin Illusion

NOTES

1 All this literature is listed in the Bibliography of Happiness, Section Fg02 ‘Happiness
and economic growth/decline’ (Veenhoven, 2012a).
2 The idea that average happiness in nations will not change over time also fits with the
‘set-point’ theory of happiness, which became popular some 10 years later (e.g.,
Lykken, 1999).
3 This collection includes only findings based on probability samples in the general
population, typically aged 15 years or more. Studies among specific subgroups in a
country are left out, such as people living in urban areas, working age people or
internet users.
4 We made an exception for the two variants of the question on life satisfaction used in
the Latino Barometro, that is, the questions coded O-SLU-g-sq-v-4-b and O-SLu-g-sq-
v-4-c, which differ slightly in the wording of the answer options. In this case we could
use the first data yielded by the happiness scale interval study (Veenhoven, 2009b) in
which native speakers rated the degree of happiness denoted by the answer categories
in intervals on scale 0 to 10. This enabled us to transform the observed frequency
distributions to a mean on a common scale.
5 An alternative to this bi-partition is to distinguish an additional intermediating category
of small change. This requires an arbitrary definition of how small a small change
should be and involves a loss of statistical power. Analyses of that kind showed the
same pattern but did not reach statistical significance.
6 In this context, Stevenson and Wolfers (2008, p. 18) rightly observe that “... absence of
evidence should not be confused with evidence of absence”.
7 The right top-case of rising happiness in spite of economic decline is not part of the
Easterlin Paradox, but mentioned for the sake of completeness.

Biographical notes:

Ruut Veenhoven is a Sociologist and Emeritus Professor of Social

Conditions for Human Happiness at Erasmus University Rotterdam in the Netherlands. He

is also an Extraordinary Professor at North-West University in South Africa. He is the
Director of the World Database of Happiness and Founding Editor of the Journal of
Happiness Studies. His research focuses on conditions for happiness and aims at more
informed choice in public and private matters. His recent publications are: Greater
Happiness for a Greater Number: Is that Possible? If So, How? (2011) and Has
Modernization Gone Too Far? Happiness and Modernity in 141 Contemporary Nations
(2013).

Floris Vergunst received his Bachelor in Economics and Business Economics from the
Erasmus University Rotterdam in 2011. Under the supervision of Prof. Veenhoven, he has
written his Bachelor thesis about the relationship between economic growth and happiness.
After Economics, he continued studying Sociology at Tilburg University. In 2013, he
graduated with honour on a thesis about the relation between religion and happiness. At
the moment, he is studying Philosophy at the Katholieke Universiteit Leuven in Belgium.
Ruut Veenhoven & Floris Vergunst 20 The Easterlin Illusion

Scheme 1a
Combinations of change in happiness and change in GDP per capita
Rise or decline in GDP

Change happiness Change GDP

growth decline

increase concordant discordant

decline discordant concordant

Scheme 1b
Combinations of change in happiness and change in GDP per capita
Great or small rise in GDP

Change happiness Growth GDP

great small

increase concordant concordant

decline discordant discordant

Concordant = in line with common sense, changes in happiness and GDP go together
Discordant = contrary to common sense, but in line with Easterlin’s thesis; no relation between change in
happiness and GDP
Ruut Veenhoven & Floris Vergunst 21 The Easterlin Illusion

Table 2a
Change of average happiness in nations
Frequency of rise versus decline

Pattern of change series nations

N % N %

Rise 133 67% 41 62%

Decline 66 33% 25 38%

Total 199 100% 66 100%

Ratio rise-decline 2.02 1.63

Table 2b
Change of average happiness in nations
Average yearly change in points on scale 0-10, split-up by length of period

Period series nations

N b N b

10- 20 years 114 +0.017 30.5 +0.010

20-40 years 67 +0.013 27.0 +0.009

> 40 years 18 +0.020 8.5 +0.030

Total 199 +0.016 67 +0.012

Ruut Veenhoven & Froris Vergunst 22 The Easterlin Illusion

Table 3a
Rise or decline of happiness and GDP
Combinations of change coefficients observed in 1991 series

Change happiness Change GDP

rise decline

rise 132 1

decline 64 2

Concordant/discordant ratio 2.1

Table 3b
Rise or decline of happiness by strong or modest economic growth
Combinations observed in 199 1 series

Change happiness Growth GDP

strong growth modest growth

> median < median

rise 74 59

decline 24 41

Odds Ratio = 2.14; 95% confidence interval +1.29 to +4.18

1
Case of median economic growth not included
 Ruut Veenhoven & Froris Vergunst 23 The Easterlin Illusion

Table 3c
Correlation between change in average happiness and growth GDP
in 199 time series

Time span Correlation coefficient 95% confidence of r

Δ Happiness with Δ GDP

10-20 years +0.20 +0.01 to +.037

21-40 years +0.21 +0.05 to +0.50

> 40 years +0.20 -0.03 to +0.61

All periods +0.20 +0.08 to +0.34

Ruut Veenhoven & Froris Vergunst 24 The Easterlin Illusion

Table 4a
Average effect of 1% growth GDP p.c. on average happiness on scale 0-10
Split-up by time span

Time span series

N b

10-20 years 114 0.0030

21-40 years 67 0,0063

> 40 years 18 0.0048

All 199 0.0034

Table 4b
Average effect of 1% growth GDP p.c. on average happiness on scale 0-10
Split-up by initial wealth of the nation

Average income per series

capita in nation

N b

Low 0 -

Lower middle 20 0.0117

Upper middle 56 0.0034

High 123 0.0029

Total 199 0.0034

Ruut Veenhoven & Froris Vergunst 25 The Easterlin Illusion

Table 4c
Average effect of 1% growth GDP p.c. on average happiness on scale 0-10
Split-up by instability of economic growth on a 0-1 scale

Economic development series

N b

Smooth: i < 0.1 57 0.0054

Medium: 0.1 < i < 0.25 80 0.0032

Bumpy: i > 0.25 62 0.0061

Total 199 0.0034

Ruut Veenhoven & Froris Vergunst 26 The Easterlin Illusion

Figure 1a
Economic growth and rising happiness in nations
Correlation in 114 series over 10 to 20 year periods
Ruut Veenhoven & Froris Vergunst 27 The Easterlin Illusion

Figure 1b
Economic growth and rising happiness in nations
Correlation in 67 series over 21 to 40 year periods
Ruut Veenhoven & Froris Vergunst 28 The Easterlin Illusion

Figure 1c
Economic growth and rising happiness in nations
Correlation in 18 series over periods of more than 40 years
Ruut Veenhoven & Froris Vergunst 29 The Easterlin Illusion

APPENDIX A

Example of a presentation of findings on average happiness in nations in the

collection ‘Happiness in Nations’ of the ‘World Database of Happiness’

Distributional findings on happiness in Argentina (AR)

Table A1
Measure type: 111C four-step verbal happiness (see online version for colours)
Ruut Veenhoven & Froris Vergunst 30 The Easterlin Illusion

Table A2
Measure type: 121C four-step verbal life satisfaction (see online version for colours)
Ruut Veenhoven & Froris Vergunst 31 The Easterlin Illusion

Table A3
Measure type: 122F ten-step numeral life satisfaction (see online version for colours)

Table A4
Measure type: 122G 11-step numeral life satisfaction (see online version for colours)
Ruut Veenhoven & Froris Vergunst 32 The Easterlin Illusion

Table A5

Measure type: 222 ten-item affect balance scale (Bradburn) (see online version for
colours)

Table A6
Measure type: 235 more days like yesterday (see online version for colours)
Ruut Veenhoven & Froris Vergunst 33 The Easterlin Illusion

Table A7
Measure type: 236 14-item yesterday’s affect balance (see online version for colours)

Table A8
Measure type: 31D 11-step numeral best-worst possible life (see online version
for colours)
Ruut Veenhoven & Froris Vergunst 34 The Easterlin Illusion

Table A9
Measure type: 411B 3-step feel happy (see online version for colours)

Source: Cite as: R. Veenhoven, Distributional findings on Happiness in

Argentina (AR), World Database of Happiness, Erasmus University Rotterdam, The
Netherlands [online]
http://worlddatabaseofhappiness.eur.nl (accessed 2012-10-07)
APPENDIX B
Ruut Veenhoven & Floris Vergunst 35 The Easterlin Illusion
Table B1 Datamatrix
Period Country OHL3 OHL4 OHL5 OSL2 OSL3 OSL4 OSL5 OSL7 OSL10 OSL11 CBW Happ/GDP GDP-growth Term Development Instability
2002-2011 Argentina 0,078 0,006 0,066 10-20 Upper mid. 0,491
1981-2005 Argentina 0,026 0,003 0,055 21-40 Upper mid. 0,687
1981-2006 Argentina 0,039 0,001 0,060 21-40 Upper mid. 0,733
1975-2008 Australia 0,011 0 0,070 21-40 High 0,15
1981-2005 Australia -0,005 -0,001 0,054 21-40 High 0,178
1981-2005 Australia -0,001 -0,003 0,054 21-40 High 0,178
1975-2011 Australia 0,001 0,072 21-40 High 0,185
1995-2011 Austria -0,027 -0,001 0,042 10-20 High 0,289
1990-1999 Austria 0,187 0,018 0,049 10-20 High 0,314
1990-2006 Austria -0,029 -0,002 0,050 10-20 High 0,428
1990-2000 Belarus 0,072 -0,559 -0,034 10-20 Upper mid. 0,243
1990-2000 Belarus -0,09 -0,317 -0,034 10-20 Upper mid. 0,243
1981-1999 Belgium 0,011 0,001 0,050 10-20 High 0,133
1989-2008 Belgium -0,032 -0,002 0,059 10-20 High 0,27
1975-1986 Belgium -0,054 -0,004 0,057 10-20 High 0,833
1973-2011 Belgium -0,008 0 0,077 21-40 High 0,098
1981-2006 Belgium 0,016 0,002 0,052 21-40 High 0,154
2002-2011 Bolivia 0,046 0,036 0,073 10-20 Lower mid. 0,116
1997-2007 Bolivia -0,05 0,055 0,031 10-20 Lower mid. 0,912
1990-2006 Brasil 0,054 0,024 0,059 10-20 Upper mid. 0,92
1990-2006 Brasil 0,022 0,001 0,059 10-20 Upper mid. 0,92
1960-2011 Brasil 0,046 0,009 0,093 > 40 Upper mid. 0,249
2001-2011 Bulgaria 0,05 0,009 0,139 10-20 Upper mid. 0,053
2002-2011 Bulgaria 0,021 0,004 0,155 10-20 Upper mid. 0,055
1990-2006 Bulgaria 0,029 -0,008 0,047 10-20 Upper mid. 0,734
1990-2006 Bulgaria 0,019 0,014 0,047 10-20 Upper mid. 0,734
1968-1977 Canada -0,009 -0,001 0,112 10-20 High 0,053
1981-2000 Canada 0,01 -0,002 0,042 10-20 High 0,174
1982-2000 Canada -0,004 0 0,038 10-20 High 0,203
1997-2010 Chile 0,009 0,007 0,060 10-20 Upper mid. 0,334
1990-2005 Chile 0,017 0,007 0,078 10-20 Upper mid. 0,383
1990-2005 Chile -0,025 -0,015 0,078 10-20 Upper mid. 0,383
1990-2007 China 0,002 -0,002 0,120 10-20 Upper mid. 0,088
1997-2011 China -0,005 0,001 0,144 10-20 Upper mid. 0,16
1990-2009 China -0,019 -0,006 0,136 10-20 Upper mid. 0,198
1990-2010 Costa Rica -0,041 -0,016 0,057 21-40 Upper mid. 0,08
1995-2007 Croatia 0,019 0,003 0,138 10-20 High 0,209
1962-2011 Croatia 0,018 -0,011 0,070 > 40 High 0,19
2001-2011 Cyprus 0,033 0,002 0,070 10-20 High 0,075
2001-2011 Czech 0,026 0,002 0,115 10-20 High 0,075
1990-2006 Czech 0,047 0,002 0,095 10-20 High 0,177
1981-1999 Denmark 0,003 0 0,055 10-20 High 0,081
1975-1986 Denmark -0,026 -0,006 0,064 10-20 High 0,51
1972-2006 Denmark 0,019 0,002 0,085 21-40 High 0,072
1973-2011 Denmark 0,014 0,001 0,078 21-40 High 0,084
1962-2011 Dom. Republic 0,081 0,081 0,081 > 40 Upper mid. 0,177
1997-2007 Ecuador -0,05 -0,016 0,062 10-20 Upper mid. 0,483
Ruut Veenhoven & Floris Vergunst 36 The Easterlin Illusion
Table B1 Datamatrix (continued)
Period Country OHL3 OHL4 OHL5 OSL2 OSL3 OSL4 OSL5 OSL7 OSL10 OSL11 CBW Happ/GDP GDP-growth Term Development Instability
1960-2011 Egypt -0,012 -0,107 0,064 > 40 Lower mid. 0,173
1991-2010 El Salvador -0,04 -0,03 0,073 10-20 Lower mid. 0,013
1997-2006 England -0,006 0 0,068 10-20 High 0,166
1975-1986 England 0,023 0,004 0,088 10-20 High 0,305
1973-2011 England 0,008 0,001 0,076 21-40 High 0,074
1975-2011 England 0,002 0,001 0,073 21-40 High 0,074
1981-2006 England -0,002 0 0,069 21-40 High 0,075
1981-2006 England -0,003 0 0,069 21-40 High 0,075
1990-1999 Estonia -0,038 0,071 0,156 10-20 High 0,036
2001-2011 Estonia 0,091 0,006 0,128 10-20 High 0,153
1990-2006 Estonia 0,051 0,008 0,151 10-20 High 0,162
1956-2011 Finland 0,01 0,002 0,081 > 40 High 0,103
1972-2006 Finland 0,01 0,001 0,089 21-40 High 0,104
1981-2005 Finland 0 0 0,064 21-40 High 0,247
1975-1986 France -0,003 0 0,064 10-20 High 0,546
1973-2011 France 0,016 0,002 0,074 21-40 High 0,083
1975-2011 France 0,032 0,003 0,064 21-40 High 0,092
1981-2006 France 0,017 0,002 0,050 21-40 High 0,167
1981-2006 France 0,011 0,001 0,050 21-40 High 0,167
1991-2009 Germany 0,009 0,001 0,056 10-20 High 0,447
1997-2006 Germany 0,001 0 0,013 10-20 High 0,814
1990-2010 Germany 0,001 0,001 0,046 21-40 High 0,34
1981-2011 Greece -0,012 -0,002 0,057 21-40 High 0,151
2002-2011 Guatamala -0,15 -0,104 0,055 10-20 Lower mid. 0,044
1997-2009 Guatamala 0,05 0,054 0,055 10-20 Lower mid. 0,238
1997-2007 Honduras 0,005 0,004 0,077 10-20 Lower mid. 0,037
2002-2011 Honduras -0,12 -0,106 0,059 10-20 Lower mid. 0,042
1981-1999 Hungary -0,076 -0,048 0,054 10-20 High 0,074
2001-2011 Hungary -0,065 -0,007 0,104 10-20 High 0,138
1981-2006 Hungary -0,006 -0,002 0,076 21-40 High 0,234
1981-1999 Iceland 0,004 0,001 0,056 10-20 High 0,192
1981-1999 Iceland -0,003 0 0,056 10-20 High 0,192
1990-2006 India 0,027 0,072 0,050 10-20 Lower mid. 0,226
1962-2011 India 0,044 0,087 0,067 > 40 Lower mid. 0,272
1975-2007 India 0,064 0,315 0,056 21-40 Lower mid. 0,184
1981-1999 Ireland 0,028 0,003 0,088 10-20 High 0,069
1975-1986 Ireland 0,06 0,016 0,084 10-20 High 0,216
1981-2006 Ireland -0,006 0 0,094 21-40 High 0,13
1973-2011 Ireland 0,002 0 0,095 21-40 High 0,153
1961-2011 Israel 0,04 0,008 0,069 > 40 High 0,085
1975-1986 Italy 0,053 0,011 0,084 10-20 High 0,188
1973-2011 Italy 0,019 0,002 0,079 21-40 High 0,071
1975-2011 Italy 0,027 0,003 0,073 21-40 High 0,075
1975-2009 Italy 0,011 0,001 0,081 21-40 High 0,088
1981-2006 Italy 0,024 0,003 0,067 21-40 High 0,156
1981-2005 Italy 0,008 0,001 0,069 21-40 High 0,175
1988-2005 Japan -0,01 -0,001 0,049 10-20 High 0,623
1964-2011 Japan 0,007 0,001 0,098 > 40 High 0,099
Ruut Veenhoven & Floris Vergunst 37 The Easterlin Illusion

Table B1 Datamatrix (continued)

Period Country OHL3 OHL4 OHL5 OSL2 OSL3 OSL4 OSL5 OSL7 OSL10 OSL11 CBW Happ/GDP GDP-growth Term Development Instability
1962-2011 Japan 0,023 0,002 0,100 > 40 High 0,099
1978-2002 Japan -0,021 -0,001 0,086 21-40 High 0,136
1975-2007 Japan 0,022 0,002 0,075 21-40 High 0,159
1981-2005 Japan 0,026 0,002 0,065 21-40 High 0,241
1981-2005 Japan 0,013 0,001 0,065 21-40 High 0,241
1981-2011 Korea 0,072 0,009 0,092 21-40 High 0,08
1981-2005 Korea 0,081 0,014 0,099 21-40 High 0,093
1981-2005 Korea 0,035 0,006 0,099 21-40 High 0,093
1980-2007 Korea 0,028 0,004 0,108 21-40 High 0,096
1981-2001 Korea 0,017 0,003 0,103 21-40 High 0,13
2001-2011 Latvia 0,018 0,001 0,134 10-20 Upper mid. 0,193
1990-2006 Latvia 0,049 0,011 0,068 10-20 Upper mid. 0,383
1990-1999 Latvia -0,068 0,1 0,015 10-20 Upper mid. 0,975
2001-2011 Lithuania 0,04 0,004 0,132 10-20 Upper mid. 0,135
1990-2006 Lithuania 0,065 0,011 0,076 10-20 Upper mid. 0,338
1990-1999 Lithuania -0,03 0,113 0,022 10-20 Upper mid. 0,968
1975-1986 Luxembourg 0,038 0,008 0,050 10-20 High 0,707
1975-2004 Luxembourg 0,009 0,001 0,082 21-40 High 0,092
1973-2011 Luxembourg 0,009 0 0,092 21-40 High 0,145
2001-2011 Malta -0,011 -0,001 0,063 10-20 High 0,062
1975-2011 Mexico 0,017 0,009 0,077 21-40 Upper mid. 0,146
1975-2007 Mexico 0,23 0,01 0,084 21-40 Upper mid. 0,176
1981-2005 Mexico 0,045 0,032 0,070 21-40 Upper mid. 0,248
1981-2005 Mexico 0,017 0,016 0,070 21-40 Upper mid. 0,248
1996-2006 Moldavia 0,023 -0,008 0,069 10-20 Lower mid. 0,649
1996-2006 Moldavia 0,189 0,25 0,069 10-20 Lower mid. 0,649
1975-1986 Netherlands 0,015 0,008 0,059 10-20 High 0,712
1973-2011 Netherlands 0,007 0,001 0,078 21-40 High 0,103
1977-2011 Netherlands -0,005 0 0,062 21-40 High 0,113
1981-2008 Netherlands 0,022 0,002 0,057 21-40 High 0,114
1981-2008 Netherlands 0,001 0 0,057 21-40 High 0,114
1974-2009 Netherlands 0,012 0,001 0,082 21-40 High 0,121
1997-2007 Nicaragua -0,076 -0,207 0,030 10-20 Lower mid. 0,149
1990-2000 Nigeria 0,16 1,198 0,029 10-20 Lower mid. 0,953
1990-2000 Nigeria 0,026 0,438 0,029 10-20 Lower mid. 0,953
1962-2011 Nigeria 0,01 0,012 0,076 > 40 Lower mid. 0,597
1981-1996 Norway -0,019 -0,001 0,067 10-20 High 0,095
1972-2007 Norway -0,018 -0,001 0,094 21-40 High 0,109
1962-2011 Panama 0,042 0,032 0,064 > 40 Upper mid. 0,112
1997-2007 Paraguay -0,055 -0,082 0,002 10-20 Lower mid. 0,672
2002-2011 Peru -0,009 0 0,101 10-20 Upper mid. 0,045
1997-2007 Peru -0,013 -0,007 0,038 10-20 Upper mid. 0,589
1996-2005 Peru 0,015 0,002 0,032 10-20 Upper mid. 0,995
1991-2000 Poland 0,028 0,01 0,082 10-20 High 0,05
1990-2007 Poland -0,011 -0,005 0,099 10-20 High 0,096
1990-2007 Poland 0,028 0,08 0,099 10-20 High 0,096
2001-2011 Poland 0,066 0,031 0,106 10-20 High 0,108
1962-2011 Poland 0,027 0,007 0,086 > 40 High 0,173
Ruut Veenhoven & Floris Vergunst 38 The Easterlin Illusion

Table B1 Datamatrix (continued)

Period Country OHL3 OHL4 OHL5 OSL2 OSL3 OSL4 OSL5 OSL7 OSL10 OSL11 CBW Happ/GDP GDP-growth Term Development Instability
1990-1999 Portugal -0,011 -0,002 0,088 10-20 High 0,166
1990-2006 Portugal 0,036 0,005 0,078 10-20 High 0,169
1985-2011 Portugal -0,02 -0,002 0,086 21-40 High 0,062
1990-2006 Romania -0,003 0,009 0,070 10-20 Upper mid. 0,467
1990-2005 Romania -0,015 0,031 0,055 10-20 Upper mid. 0,5
1990-2003 Romania -0,018 -0,003 0,021 10-20 Upper mid. 0,692
2002-2011 Russia 0,09 0,01 0,210 10-20 Upper mid. 0,115
1990-2005 Russia 0,046 0,037 0,031 10-20 Upper mid. 0,892
1990-2005 Russia 0,056 0,073 0,031 10-20 Upper mid. 0,892
1992-2005 Russia 0,128 -0,001 0,032 10-20 Upper mid. 0,981
1996-2006 Serbia -0,034 -0,018 0,123 10-20 Upper mid. 0,834
1996-2006 Serbia 0,047 0,017 0,123 10-20 Upper mid. 0,834
1990-1999 Slovakia -0,015 -0,004 0,113 10-20 High 0,04
2001-2011 Slovakia 0,116 0,008 0,108 10-20 High 0,074
2002-2011 Slovakia 0,061 0,005 0,123 10-20 High 0,09
1990-2006 Slovakia 0,07 0,011 0,115 10-20 High 0,129
2001-2011 Slovenia -0,01 0 0,073 10-20 High 0,104
1990-2007 Slovenia 0,067 0,006 0,062 10-20 High 0,197
1992-2006 Slovenia 0,111 0,013 0,061 10-20 High 0,228
1962-2011 Slovenia 0,014 0,003 0,060 > 40 High 0,154
2002-2011 South-Africa -0,038 -0,006 0,106 10-20 Upper mid. 0,114
1983-2002 South-Africa -0,044 0,059 -0,001 10-20 Upper mid. 0,85
1981-2007 South-Africa 0,03 0,015 0,048 21-40 Upper mid. 0,655
1981-2007 South-Africa 0,015 0,041 0,048 21-40 Upper mid. 0,655
1983-2004 South-Africa -0,075 0 0,018 21-40 Upper mid. 0,924
1985-2011 Spain 0,007 0 0,081 21-40 High 0,135
1981-2007 Spain 0,015 0,002 0,069 21-40 High 0,137
1981-2007 Spain 0,023 0,003 0,069 21-40 High 0,137
1995-2011 Sweden 0,024 0,001 0,051 10-20 High 0,229
1972-2006 Sweden 0,001 0 0,073 21-40 High 0,111
1981-2006 Sweden -0,015 -0,001 0,050 21-40 High 0,222
1990-2007 Switzerland 0,007 0,001 0,037 10-20 High 0,471
1990-2007 Switzerland -0,023 -0,002 0,037 10-20 High 0,471
1995-2006 Taiwan -0,038 -0,01 0,034 10-20 High 0,415
1995-2006 Taiwan 0,002 0 0,034 10-20 High 0,415
2002-2011 Turkey 0,08 0,009 0,108 10-20 Upper mid. 0,089
2001-2011 Turkey 0,067 0,006 0,103 10-20 Upper mid. 0,111
1990-2007 Turkey 0,061 0,025 0,104 10-20 Upper mid. 0,305
1990-2000 Turkey -0,03 -0,026 0,096 10-20 Upper mid. 0,394
1996-2006 Ukrain 0,133 0,131 0,071 10-20 Lower mid. 0,644
1996-2006 Ukrain 0,188 0,159 0,071 10-20 Lower mid. 0,644
1997-2007 Uruguay -0,092 0,02 0,013 10-20 Upper mid. 0,563
1991-2004 USA 0,049 -0,001 0,040 10-20 High 0,009
1991-2008 USA 0,074 0,005 0,042 10-20 High 0,986
1959-2011 USA 0,009 0,001 0,058 > 40 High 0,04
1957-2010 USA 0,001 0 0,058 > 40 High 0,042
1959-2007 USA 0,005 0 0,062 > 40 High 0,044
1946-2002 USA -0,001 0,001 0,064 > 40 High 0,045
Ruut Veenhoven & Floris Vergunst 39 The Easterlin Illusion
Table B1 Datamatrix (continued)
Period Country OHL3 OHL4 OHL5 OSL2 OSL3 OSL4 OSL5 OSL7 OSL10 OSL11 CBW Happ/GDP GDP-growth Term Development Instability
1946-1990 USA -0,002 -0,007 0,073 > 40 High 0,075
1981-2006 USA 0,006 0 0,052 21-40 High 0,008
1981-2006 USA -0,016 -0,001 0,052 21-40 High 0,008
1975-2008 USA 0,008 0,001 0,060 21-40 High 0,011
1973-2008 USA 0,011 0,001 0,062 21-40 High 0,012
1968-2000 USA -0,002 0 0,067 21-40 High 0,016
2002-2011 Venezuela 0,078 0,007 0,128 10-20 Upper mid. 0,138
1997-2007 Venezuela 0,046 0,009 0,080 10-20 Upper mid. 0,553

Missende data GDP: Egypte 1959, Croatia 1961-1989, Polen 1961-1984, USA 1945-1959, Finland 1955-1959, Estonia 1989-1994, Czech 1989, Lithuania 1989, Moldavia 1989, Serbia 1995-1996, Belarus 1989.