Beyond GDP
Measuring social progress
in Europe
Koen Decancq
Erik Schokkaert
Introduction
• Development of measures that go “beyond GDP” is
booming in recent decades:
o Sen-Nussbaum approach to human capabilities;
happiness literature.
o Human Development Index (UNDP); “Better Life
Initiative” (OECD).
o Stiglitz-Sen-Fitoussi report (2009).
Europe and “quality of life”
• Europe 2020 strategy is inherently multidimensional.
• Communication European Commission (2009): “Beyond
GDP”.
• Initiatives launched by the European Statistical System on
(multidimensional) “Quality of life”-indicators, focusing on
EU-SILC (2013 ad hoc-module).
• This paper: how should we proceed? A specific proposal.
Caveats
• We focus on the measurement of the well-being of the
present generation.
• CAVEAT 1: sustainability is an essential challenge, but
should be measured by separate indicators (and should
act as a constraint).
• CAVEAT 2: development of lists of specific policy
indicators is useful, but a different issue. Operational
targets often refer to “inputs” and not to directly relevant
“outputs”.
Structure
1. Basic principles.
2. A specific proposal: equivalent income.
3. Application: well-being and social progress in Europe
between 2008 and 2010.
Basic principles
Principle 1: focus on individual well-being
The ultimate goal of policy and the ultimate criterion to
evaluate social progress is the well-being of individuals
making up a society.
• Recall: sustainability is a restriction to be imposed on
present generations.
Principle 2: focus on outcomes
Information must be collected on the different dimensions of
life that are important for the well-being of individual citizens.
This information should be about outcomes, and not about
inputs.
• Well-being is not fully determined by income or material
consumption. Other dimensions of life are essential (e.g.
health, quality of social interactions and of the natural
environment, safety).
Collective characteristics are included!
• Although there are many different lists, they are largely
overlapping.
Principle 3: accounting for cumulative
deprivation
“well-being”
income
health
individual 1
100
10
55
individual 2
10
100
55
average
55
55
10/1
10/1
income
health
individual 1
100
100
100
individual 2
10
10
10
average
55
55
10/1
10/1
ratio
ratio
10
1/1
“well-being”
10/1
Accounting for cumulative deprivation requires that one first
constructs an index of well-being at the individual level and
then aggregates these well-being indices across individuals.
Principle 4: respect for individual ideas about
a good life
The weighting scheme applied to construct the measure of
individual well-being should respect the individual ideas about
what is a good life.
• This discards the use of objective “synthetic” indicators,
such as the Human Development Index:
• Reluctance towards “indicator aggregation” is easily
understandable if one aims at defining trade-offs at the
aggregate level. Weights are arbitrary.
• At the level of individual evaluation, however, different
dimensions of life are commensurable. In daily life, we
continuously trade off life dimensions.
Principle 5: what about life satisfaction or
happiness?
• There seems to be a general presumption that subjective
satisfaction measures are attractive precisely because
they satisfy the principle of individual sovereignty, i.e. they
do respect preferences.
o Richard Layard: “If we accept the Marxist idea of ‘false
consciousness, we play God and decide what is good
for others, even if they will never feel it to be so” (2005,
p. 121).
• Is this interpretation correct?
Influence of social background
• PERSON 1
• Low income, poor job, ill.
• PERSON 2
• High income, prestigious
job, now and then
somewhat stressed.
• All by all, relatively
• Dissatisfied with life. His
satisfied with life. Comes
from a deprived social
background and lives in a
deprived neighbourhood.
father did much better.
15
Adaptation
o
“A person who is ill-fed, undernourished, unsheltered and ill
can still be high up in the scale of happiness or desirefulfillment if he or she has learned to have ‘realistic’ desires
and to take pleasure in small mercies” (Sen, 1985).
• Much evidence on adaptation in the empirical literature:
o
o
Countries with higher rates of HIV prevalence do not report
poorer life satisfaction, yet individuals care about HIV
(Deaton, 2008).
Individuals who have lost a limb may, after adaptation,
recover a good satisfaction score – but still express a strong
aversion to disability (Loewenstein and Ubel, 2008).
• Conflict conflict between “respecting preferences” and
using life satisfaction as a measure of well-being.
• Ethical issue: how do we treat differences in aspirations,
adaptation etc.
• Yet, “it would be odd to claim that a person broken down
by pain and misery is doing very well” (Sen, 1985).
Happiness or (subjective life satisfaction) may be one of the
important dimensions of life, but it should not be seen as an
encompassing measure of individual well-being.
Principle 6: inequality aversion
Justice requires accounting for inequality in individual wellbeing. This can be done in a natural and flexible way by
introducing an inequality aversion parameter in the analysis.
Social welfare = M (1 - I )
“AVERAGE” WELL-BEING
INEQUALITY MEASURE
ZERO INEQUALITY
GINI
LARGER
WEIGHT
TO THE POOR
A specific proposal:
equivalent income
Equivalent incomes
• Fix reference values for all the non-income dimensions.
• DEFINITION: The equivalent income of an individual is the
hypothetical income that, if combined with the reference
value on all non-income dimensions, would place the
individual in a situation that he/she finds equally good as
his/her actual situation.
An example: income and health
To “estimate” A’ and B’ we need information about
willingness to pay for perfect health.
Equivalent income = actual income – WTP.
Choice of reference values
• THIS IS AN ETHICAL OR POLITICAL, NOT A
PSYCHOLOGICAL QUESTION!
• When can we say that someone with a larger income is
necessarily better off? If the two individuals are “equally
ill”? Not really, because the “richer” individual may care
more about her poor health.
• But what if they both are in perfect health? Would it then
be reasonable for the rich to claim that he is not better off
because he cares less about being healthy?
• Combine this choice of reference values with “respect for
individual opinions about the good life”.
Pros and cons
• Equivalent income = actual income minus the welfare loss
incurred on the non-income dimensions (measured as
willingness-to-pay).
• Satisfies all our basic principles.
• Measurable in money terms, can be introduced in any
social welfare/inequality measure.
• Yet, NOT money fetishism. Just a methodology that allows
us to capture the (personal) trade-offs.
Pros and cons
• Less intuitive than happiness or HDI – but these
approaches do not satisfy our basic principles.
• Choice of reference values: room for debate.
• Information needed about “preferences” (or WTP), yet
techniques are well-known.
o Behaviour.
o Contingent valuation surveys (environment, health).
o Derive information about willingness-to-pay from life
satisfaction questions.
Pros and cons
• Less intuitive than happiness or HDI – but these
approaches do not satisfy our basic principles.
• Choice of reference values: room for debate.
• Information needed about “preferences” (or WTP), yet
techniques are well-known.
o Behaviour.
o Contingent valuation surveys (environment, health).
o Derive information about willingness-to-pay from life
satisfaction questions.
Equivalent income derived from satisfaction
data
Social progress in
Europe: an illustration
with pooled data
Data
• European Social Survey, 2008 and 2010. (SILC does not
contain a question on life satisfaction).
• 18 countries: 15 EU-members, Switzerland, Norway, the
Russian Federation. About 52,000 individual observations.
• Dimensions:
Estimating preference differences
• Assumption: preferences do not differ between different
countries.
• Different groups within the countries have different
preferences:
o The higher educated give a (relatively) larger weight to
income.
o Health is less important to the young.
o Unemployment is relatively less important for women,
social interactions are (relatively) more important for
women than for men.
Estimation results
REFERENCE
GROUP
young
female
higher
educated
log income (per
capita)
0.371***
(0.021)
0.014
(0.010)
0.037
(0.023)
0.027**
(0.010)
self-assessed
health
0.661***
(0.018)
-0.064**
(0.020)
0.002
(0.018)
-0.053**
(0.020)
unemployment
-0.840***
(0.080)
0.030
(0.081)
0.222**
(0.075)
0.017
(0.085)
social
interactions
0.143***
(0.010)
-0.001
(0.011)
0.019+
(0.011)
-0.006
(0.012)
personal safety
0.224***
(0.021)
0.023
(0.021)
-0.060**
(0.021)
-0.016
(0.022)
Controls: household size, education, education squared, gender, age, age squared,
marital status, religious, urban, ethnic minority, time, country.
Income, equivalent income, happiness (2010)
Income
(NO, CH)
Equivalent
income
(NO, CH)
Happiness
(DK, CH)
DE
28986 (6)
3188 (10)
7.26 (9)
DK
28162 (7)
6938 (4)
8.35 (1)
FR
25779 (10)
3529 (9)
6.34 (15)
ES
22282 (11)
3182 (11)
7.30 (8)
GR
19388 (13)
2564 (13)
5.71 (17)
(RU, EE)
(RU, HU)
(GR, RU)
From income to equivalent income (2010)
from income
to equivalent
income
health
unemployment
social interactions
safety
(DK, NO, SE)
DK
(1)
-75%
-52%
-3%
-38%
-20%
ES
(7)
-86%
-67%
-6%
-39%
-31%
FR
(9)
-86%
-67%
-3%
-44%
-29%
GR
(11)
-87%
-45%
-5%
-61%
-44%
DE
(13)
-89%
-73%
-1%
-48%
-30%
(EE, HU, RU)
Inequality (2010)
Gini coefficient
(income)
(CZ, SE)
Gini coefficient
(equivalent income)
(NO, DK)
CZ
0.27 (1)
0.73 (10)
DK
0.28 (3)
0.65 (2)
HU
0.30 (6)
0.77 (17)
SI
0.32 (9)
0.75 (14)
CH
0.34 (14)
0.66 (3)
GB
0.36 (16)
0.72 (9)
GR
0.36 (17)
0.75 (13)
ES
0.38 (18)
0.74 (12)
(GR, ES)
(HU, EE)
Social well-being (2010)
Income ( = 0)
(NO, CH)
Income ( = 5)
(NO, SE)
Equivalent
income ( = 5)
(NO, CH)
GB
29794 (5)
11262 (9)
282 (7)
DE
28986 (6)
12754 (7)
175 (10)
DK
28162 (7)
13828 (5)
590 (4)
BE
27477 (8)
13299 (6)
375 (6)
ES
22282 (11)
8668 (13)
146 (11)
GR
19388 (13)
7716 (14)
110 (12)
CZ
16729 (14)
8983 (11)
89 (14)
(RU, EE)
(EE, RU)
(RU, HU)
Yearly growth rates (2008-2010)
income growth
(CH, PL)
welfare growth (=5)
happiness growth
(CH, RU)
(HU, EE)
CH
+ 7.35% (1)
+9.69% (1)
+2.23%
(6)
DE
+ 0.09% (3)
- 4.51% (9)
+4.46%
(3)
BE
- 0.55% (4)
+ 4.54% (4)
+3.33%
(5)
DK
- 1.73% (8)
-4.53% (10)
-2.00%
(16)
ES
- 2.24% (11)
-12.04% (17)
-0.01%
(15)
GR
- 5.81% (17)
-22.92% (18)
-5.78%
(18)
EE
- 8.60% (18)
-7.24% (14)
+5.16%
(2)
(GR, EE)
(ES, GR)
(CZ,GR)
-15%-14%-13%-12%-11%-10%-9% -8% -7% -6% -5% -4% -3% -2% -1% 0% 1% 2% 3% 4% 5% 6% 7% 8% 9% 10% 11% 12% 13%
CH
RU
BE
PL
HU
SI
SE
NO
DE
DK
FR
GB
FI
Income
Income + inequality
NL
EE
CZ
Equivalent income
equivalent income +inequality
ES
GR
Social progress in Europe:
exploration of intercountry
differences in preferences
Direct effects for some typical countries
POOLED
FR
DE
GB
CH
log income
0.371***
0.673***
0.530***
0.189**
0.205***
health
0.661***
0.614***
0.703***
0.469***
1.109***
unemployment
-0.840***
-0.672*
-0.883***
-1.088***
-1.357*
social
interactions
0.143***
0.108**
0.170***
0.179***
0.049
personal
safety
0.224***
0.215**
0.263***
0.274***
0.189+
52137
3334
4620
3812
2584
N
Results
Equivalent income (2010)
Gini equivalent income
POOLED
SPECIFIC
POOLED
SPECIFIC
FR
3529
(9)
7797
(3)
0.70
(8)
0.50
(1)
DE
3188
(10)
5230
(6)
0.74
(11)
0.64
(3)
GB
5324
(5)
2688
(11)
0.72
(9)
0.82
(16)
CH
7706
(2)
5100
(7)
0.66
(3)
0.81
(14)
What to do?
Conclusion
1. We strongly believe in the basic principles. Debate should
turn on their ethical foundation.
2. The equivalent income is an interesting concept, but there
may be other approaches.
3. Our empirical illustration is only meant to be an
illustration.