European Journal of Political Economy 26 (2010) 137–146

Contents lists available at ScienceDirect

European Journal of Political Economy

j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / e j p e

The effects of age and job protection on the welfare costs of inflation
and unemployment
Leonardo Becchetti a,⁎, Stefano Castriota b, Giovanni Osea Giuntella c
a
Department of Economics, Università di Roma Tor Vergata, via Columbia 2, 00133, Rome, Italy
b
Department of Economics and Euricse, Università di Trento, Via Inama 5, 38100, Trento, Italy
c
Department of Economics, Boston University, 270 Bay State Road, 02215, Boston, Massachusetts, United States

a r t i c l e i n f o a b s t r a c t

Article history: We extend the happiness literature on the welfare costs of inflation and unemployment by
Received 20 May 2008 looking at age and job market characteristics. Our findings show that the relative welfare cost of
Received in revised form 27 July 2009 unemployment versus inflation is higher than one, and much higher in intermediate age
Accepted 5 August 2009
cohorts and in low job protection countries. The potential role of our findings in explaining the
Available online 21 August 2009
heterogeneous behaviour of CBs under different job market settings is discussed and compared
with alternative explanations based on other institutional or structural differences in
JEL classification:
economies and in their reactions to shocks.
E5
© 2009 Elsevier B.V. All rights reserved.
E6

Keywords:
Phillips curve
Unemployment/inflation trade-off
Happiness
Employment protection
Aging population

1. Introduction

Evaluating the relative cost of unemployment versus inflation is of foremost importance for economists and policymakers.1 In
order to maximize the population's welfare, it is essential to know which of the two phenomena has the greatest impact on
people's lives. The effectiveness of economic policies is usually evaluated on the basis of welfare indicators which arbitrarily assign
certain weights to the two bads.2 A useful convention for analysis may be the so-called Misery Index3 which calculates countries'
well being levels as a simple unweighted sum of the unemployment and inflation rates. Its implicit (and strong) assumption is that
a 1% rise in inflation and unemployment generates the same welfare loss.

⁎ Corresponding author. Tel.: +39 06 72595706; fax: +39 06 2020500.

E-mail address: becchetti@economia.uniroma2.it (L. Becchetti).
1
The unemployment–inflation trade-off is probably one of the longest-debated and most controversial issues in economics. Is the Phillips curve really vertical in
the long run? Is it linear or non-linear? For the most recent contributions and discussion on “grease” and “sand” effects in the price and wage setting mechanisms
see Akerlof et al. (1996), Wyplosz (2001) and Dickens (2001).
2
For robust evidence that inflation and unemployment are actually two bads not just for policymakers but also for individuals see Di Tella and MacCulloch
(2006, 2007) and Blanchflower (2007).
3
The Misery Index was first used by Robert Barro in the 1970s, although some commentators credit it to Arthur Okun. It is simply the sum of unemployment and
inflation rates.

0176-2680/$ – see front matter © 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.ejpoleco.2009.08.001
 138 L. Becchetti et al. / European Journal of Political Economy 26 (2010) 137–146

One of the most influential papers in the happiness literature (Di Tella et al., 2001) shows that the Misery Index underestimates
the relative cost of unemployment. Using a two-stage approach the authors demonstrate that policymakers should trade-off a 1%
reduction in unemployment with a 1.66% increase in inflation in order to maintain life-satisfaction constant. This finding
contradicts the Misery Index hypothesis that the two bads have to be considered as perfect substitutes. A main contention of the
paper is that the cost of unemployment is higher also because it does not affect unemployed individuals alone.
In fact, the cost of unemployment is given by the sum of two components: the psychological cost of being unemployed (which
affects only jobless people)4 and the disutilities due to the cost of supporting the unemployed and the fear of becoming
unemployed (which affect the non unemployed population).
Our paper extends this literature in two directions by testing whether the inflation/unemployment trade-off changes according
to age and degree of employment protection in different countries. Using individual observations from the Eurobarometer Survey
(1975–2002) we find that the relative cost of unemployment versus inflation is markedly higher in central age classes and in
countries with lower employment protection (EPL). In the final part of the paper we wonder whether the different behaviours of
economic institutions in the US, UK and continental Western Europe may be partly explained by the differing worries about
unemployment observed among citizens—even though the differences which emerge when comparing, for instance, the main
targets declared in the statutes of the ECB and the Federal Reserve are not always supported by robust empirical evidence (Sardoni
and Wrai, 2005; Cobham, 2006).
With regard to such differences the EU found consensus for the creation of an independent Central Bank with a clear anti-
inflationary stance and no explicit consideration of an unemployment target.5 By contrast, the statute of the Federal Reserve lists
six main monetary policy objectives, does not fix any clear inflation target, and asserts that monetary policy should both sustain
economic growth and fight price increases. Central Banks aside, to be noted is that the UK government has based its decision not to
enter the EMU also on the fear that it would not be possible to pursue an active employment policy autonomously.6
This interpretation is only one among the several considered in the literature to explain the different behaviours of the two
Central Banks. Sahuc and Smets (2008) refer to it when they write that the explicit dual mandate of the Federal Reserve may lead to a
larger weight on the stabilization of output around its sustainable path and a stronger response to developments in the output gap. This
could in turn explain why policy rates move more strongly in response to the business cycle.
However, the authors provide two additional explanations. First, the size, composition and voting mechanisms of the two CBs
may determine differences in the speed of reactions to shocks. Second, economies and shocks are markedly different in the two
areas.7 Without ignoring the relevance and validity of these two additional explanations, the empirical findings of our paper may
be helpful for checking whether the statutory differences between the two CBs are consistent with data on differences in inflation
and unemployment costs in terms of psychological wellbeing within different institutional settings.

2. Data, methodologies and econometric findings

Our source is the Eurobarometer Survey,8 which contains information on self-reported life satisfaction of more than 634,000
individuals from 1975 to 2002. With respect to Di Tella et al. (2003) we extend the analysis beyond 1992 and include new
countries (Norway from 1990 to 1996, Finland from 1993, Sweden and Austria from 1994). Macroeconomic data are extracted as
three-year moving averages centered in t − 1 in order to reduce possible measurement errors and because it takes time for people
to react to changes in the macroeconomic environment. The source of unemployment rates is the OECD Center for Economic
Performance dataset, while data on inflation are from the World Bank's World Development Indicators. Table 1 contains a detailed
description of the variables used, while Table A1 in the appendix provides summary statistics.

4
See Frey and Stutzer (2002), Frey (2008), Clark and Oswald (1994), Winkelmann and Winkelmann (1998), Feather (1990) and Darity and Goldsmith (1996).
5
De Grauwe (2005) cites two main reasons for the ECB's approach, which he considers relatively more conservative and concerned about inflation. The first is
the emphasis placed in the 1980s on the central bank's independence (Barro-Gordon, 1983) and the call for a more conservative central banker (see Rogoff,
Q2 1985). The second is the role played by Germany in shaping the EMU and the Eurosystem.
6
Gordon Brown, as Chancellor of the Exchequer, in 1997 set out five economic tests on which any decision about UK membership of the EMU should be based:
i) Cyclical convergence: are business cycles and economic structures compatible so that we and others could live comfortably with euro interest rates on a
permanent basis?; ii) Flexibility: if problems emerge is there sufficient flexibility to deal with them?; iii) Investment: would joining EMU create better conditions
for firms making long-term decisions to invest in Britain?; iv) Financial services: what impact would entry into EMU have on the competitive position of the UK's
financial services industry, particularly the City's wholesale markets?; v) Employment and growth: in summary, will joining EMU promote higher growth,
stability and a lasting increase in jobs? (Brown, 1997). The Executive HM Treasury Report (1997) concluded as follows: “We need to demonstrate sustainable
and durable convergence before we can be sure that British membership of EMU would be good for growth and jobs. Joining before such convergence is secured
would risk harming both”.
7
More specifically, if the interest rate sensitivity of various demand components in the United States is different from that in the Euro area, this will affect the
size of the interest rate changes required to maintain price stability, even in face of similar shocks. Furthermore, the Euro area economy may be more rigid in
presence of economic shocks and therefore require a more cautious monetary policies response to news. A final point is that the size and source of the shocks
hitting both economies are different (this is particularly evident when we compare productivity growth in the two areas). To sum up, if economic structures and
sensitivity to shocks are different, even a common objective function would generate heterogeneous behaviours in the two areas.
8
Our database is not a panel dataset like the German Socioeconomic Panel (GSOEP) and the British Household Survey Panel (BHPS). As a consequence, it is
impossible to measure fixed effects which proxy time invariant inherited traits affecting individual happiness, net of the impact of the variables under
observation in this paper (Ferrer-i-Carbonell and Frijters, 2004; Clark et al., 2006). However, consider that the specific focus of our research—an inquiry into the
effects of the unemployment status under different job market conditions which requires sufficient variation in country-year data—is possible only with the
Eurobarometer.
 L. Becchetti et al. / European Journal of Political Economy 26 (2010) 137–146 139

Table 1
Description of the variables used.

Variable Source Description

Life satisfaction Eurobarometer Self-declared life-satisfaction level from 1 (not at all satisfied) to 4 (very satisfied)
Unemployed Eurobarometer Dummy variable (DV) which takes value 1 if the respondent is unemployed, 0 otherwise
Self-employed Eurobarometer DV which takes value 1 if the respondent is self-employed, 0 otherwise
Retired Eurobarometer DV which takes value 1 if the respondent is retired, 0 otherwise
Student Eurobarometer DV which takes value 1 if the respondent is student, 0 otherwise
Home Eurobarometer DV which takes value 1 if the respondent is responsible for home and is not working,
0 otherwise
Male Eurobarometer DV which takes value 1 if the respondent is male, 0 otherwise
Age Eurobarometer Age of the respondent in years
Age squared Eurobarometer Square of the respondent's age in years
Middle education Eurobarometer DV which takes value 1 if the respondent has 15–18 years of education, 0 otherwise
High education Eurobarometer DV which takes value 1 if the respondent has more than 18 years of education,
0 otherwise
Married Eurobarometer DV which takes value 1 if the respondent is married, 0 otherwise
Divorced Eurobarometer DV which takes value 1 if the respondent is divorced, 0 otherwise
Separated Eurobarometer DV which takes value 1 if the respondent is separated, 0 otherwise
Widowed Eurobarometer DV which takes value 1 if the respondent is widowed, 0 otherwise
Income 2nd quartile Eurobarometer DV which takes value 1 if the respondent belongs to the 2nd income quartile,
0 otherwise
Income 3rd quartile Eurobarometer DV which takes value 1 if the respondent belongs to the 3rd income quartile,
0 otherwise
Income 4th quartile Eurobarometer DV which takes value 1 if the respondent belongs to the 4th income quartile,
0 otherwise
GDP World Bank GDP per capita in 2000 constant international US $
Inflation World Bank Inflation rate, three-year moving average
Unemployment OECD Unemployment rate, three-year moving average
EPL OECD Employment Protection Level

Our target is to calculate the Trade-Off Indexes of inflation versus unemployment for different age groups and for countries
with different EPL levels. More specifically, following Di Tella et al. (2001), the Trade-Off Index (TOI) expresses the ratio between
the social cost of a 1% change in unemployment and the social cost of a 1% change in inflation as9:

Cost of Unemployed + Cost of Unemployment ðγ + φÞ

TOI = =
Cost of Inflation ψ

where γ is the regression coefficient of unemployed individuals in the life-satisfaction estimate, φ the coefficient of the
unemployment rate and ψ that of the inflation rate. As age cohorts we consider the 25th (28 years old) and the 50th (41 years old)
percentiles of the age of sample respondents at the beginning of the investigation period. We divide the remaining 50% of the
sample by taking into account the 65-year threshold, which should discriminate between the active population and retired
workers. This allows us to measure the cost of unemployment also for persons aged over 64 who are still participating in the labour
market, and not only for retired people, for whom this cost should obviously be lower. Hence, the definition of the over-64 class
helps us to proxy the cost of unemployment for individuals close to retirement age or already retired.10 However, as well-known,
this is not the case of all professional activities and of all investigated countries. In fact, only around 60% of the subjects in our
sample aged over 64 declared that they were retired.
The institutional framework split is based on the index of job protection provided by the OECD Employment Outlook (1999).11
This index captures the strictness of employment protection laws. As thresholds delimiting low and high job protection groups we
take respectively the 35th and 65th percentiles of the OECD indicator.12 Basing our taxonomy on this index we classify Ireland, UK,
France (in the 1980s), the Netherlands, Finland (in the 1990s) and Denmark (from 1984) in the low job protection group. By
contrast, Italy, Greece, Spain, Portugal and France (in the 1990s) belong to the high protection group.13 We focus on this index

9
The index implies that we are evaluating the trade-off between a 1% increase in inflation and a 1% increase in unemployment. This is preferred to the
alternative of weighting the first coefficient at the numerator for the unemployment rate, which would make the index no longer a marginal rate of substitution
and dependent on the unemployment level.
10
Consider that also retired workers may suffer from an increase in the general level of unemployment owing to intergenerational altruism within or outside
their families and to concerns about a fiscal effect of higher unemployment levels on their pensions.
11
Chapter 2, Table 2.6, page 67. The index varies from 0 to 30: higher values imply higher job protection. Data are available for three periods: 1956–1984
(source: Lazear 1990), 1985–1990 (source: OECD), 1991–2002 (source: OECD, 1999, 2004). These sources are used also by Nickell et al. (2005). Using as
reference the labour standard index (OECD 1989–1994), which includes measures of working time, fixed-term contracts, employment protection, minimum
Q3 wages and employees representation rights, we obtain very similar classifications (see also Nickell, 1997).
12
We cautiously omit country/years with central values of the indicator from the two subgroups in order to reduce the noise created by measurement errors on
the indicator itself.
13
A sensitivity analysis performed using the Labour Standard Index (1989–1994) provided by OECD (1999) excludes France from the sample, but the substance
of our results remains unchanged. Estimates are omitted for reasons of space and are available upon request.
140 L. Becchetti et al. / European Journal of Political Economy 26 (2010) 137–146

Table 2
One-stage life-satisfaction equations, by age.

One-stage, by (sub) sample One-stage, with slope dummies

1 2 3 4 5 6 7 8 9 10
General <29 29–41 42–64 >64 <29 29–41 42–64 >64 Complete

Unemployed − 0.5 −0.49 −0.55 − 0.51 − 0.25 − 0.5 − 0.5 −0.5 −0.5 − 0.5
(0.01) (0.01) (0.02) (0.02) (0.05) (0.02) (0.02) (0.02) (0.02) (0.02)
U − 1.91 − 1.55 − 2.10 −2.64 −1 − 1.97 − 1.86 − 1.81 − 2.04 − 1.61
(0.45) (0.58) (0.56) (0.48) (0.46) (0.45) (0.45) (0.46) (0.46) (0.49)
Π −1.4 − 1.75 − 1.9 −1.3 −0.76 − 1.39 − 1.45 − 1.37 − 1.4 − 1.42
(0.31) (0.36) (0.38) (0.34) (0.3) (0.36) (0.31) (0.31) (0.30) (0.30)
DV u 28 0.32
(0.11)
DV π 28 −0.06
(0.13)
DV u 29_41 − 0.19 − 0.51
(0.09) (0.10)
DV π 29_41 0.19 0.13
(0.10) (0.12)
DV u 42_64 − 0.3 − 0.54
(0.08) (0.17)
DV π 42_64 − 0.06 − 0.03
(0.07) (0.15)
DV u 65 0.67 0.19
(0.16) (0.25)
DV π 65 0.02 0.01
(0.18) (0.24)

(Pseudo) R2 0.09 0.09 0.1 0.1 0.09 0.09 0.09 0.09 0.09 0.09
Number of obs. 404,578 74,479 110,623 149,329 70,147 404,578 404,578 404,578 404,578 404,578
a
Trade-Off Index 1.72 1.18 1.39 2.42 1.63 1.51 2.02 1.82 1.33
(0.40) (0.36) (0.33) (0.60) (0.68) (0.40) (0.55) (0.42) (0.38)

Results from multinomial ordered-probit regressions. Country–year-level clustered standard errors are in parentheses. Inflation and unemployment rates are
three-year moving averages. All regressions include country and year dummies (France and 1975 are the base), time trend, GDP per capita (in 2000 constant US$)
and personal characteristics (see Table 1). DV u (age class): slope dummy on the unemployment variable for individuals belonging to the given age class; DV π (age
class): slope dummy on the inflation variable for individuals belonging to the given age class; 28: individuals younger than 29; 29–41: individuals in the 29–41 age
class; 42–64: individuals in the 42–64 age class; 65 individuals older than 64.
a
The Trade-Off Indexes for column 10 are 1.51 (0.40), (<29); 2.02 (0.54), (29–41): 1.82 (0.42), (42–64); 1.35, (0.37) (>64). Standard Errors are in parentheses.

because we expect higher job protection to reinforce insiders' position and thereby reduce the more general population's fear of
unemployment.
We use two different methodologies for our analysis. First, we use ordered-probit models with standard errors clustered at
country–year level to regress the individual happiness levels on country and year dummy variables, the set of individual standard
controls listed in Table 1 (among which the dummy variable for being unemployed) and three macroeconomic variables (namely
the GDP per capita in international US$ and three-year averages of unemployment and inflation rates). In this first exercise we run
regressions on the full sample of individuals and use slope dummy variables for the two macroeconomic variables by age and EPL
subgroups in order to capture the different effects of the two bads on individuals with different characteristics. Therefore, the
equation we calculate is the following:

K L
Hijt = αj + λt + ∑ βk MICROkijt + ∑ θl MACROljt + εijt
k=1 l=1

where Hijt is the happiness level of individual i (i = 1,…, n) living in country j (j = 1,…, m) in period t, αj and λt are respectively
country and year dummy variables, MICROkijt is the set of personal controls listed in Table 1, and MACROljt is the set of
macroeconomic variables which include the slope dummy variables for age and EPL subgroups. The second methodology consists
in running separate regressions by age and EPL subgroups without using slope dummy variables.
Tables 2 and 3 show that the results obtained using the two different approaches exhibit an important regularity: the impact of
the general level of unemployment on life satisfaction seems to be highly influenced by age cohorts and employment protection,
since the Trade-Off Index is highest for the central age groups and for low EPL countries.
As further robustness checks we use five additional methodologies, which confirm our results—see Table A2a (Robustness
Check 1), Table A2b (RC 2, 3 and 4) and Table A3. (RC 1) Using the first we replicate the two-stage procedure performed by Di Tella
et al. (2001), the only difference being that we do not include the variable number of children because this is not available after
1988. Since the objective of our analysis is to study the effect of age cohort and job market characteristic splits, in the first stage we
perform one regression for each category. We then calculate the average prediction error for each country–year by subgroup and
use these data to perform separate second-stage regressions by sample split. The four remaining robustness checks are intended to
test whether our findings on age cohort and job protection persist when the two variables affecting costs of inflation and costs of
 L. Becchetti et al. / European Journal of Political Economy 26 (2010) 137–146 141

Table 3
One-stage life-satisfaction equations, by JP subgroups.

One-stage, by (sub) sample One-stage, with slope dummies

1 2 3 4
General Low EPL High EPL Complete

Unemployed − 0.50 −0.52 − 0.39 − 0.47

(0.01) (0.02) (0.03) (0.02)
U − 1.91 −3.80 0.39 0.25
(0.45) (1.30) (0.74) (0.70)
Π − 1.40 −2.28 − 1.99 − 1.49
(0.31) (0.82) (0.43) (0.43)
DV u Low EPL − 2.41
(0.96)
DV π Low EPL 0.85
(0.45)

(Pseudo) R2 0.09 0.08 0.03 0.10

Number of obs. 404,578 136,849 128,406 256,485
Trade-Off Index 1.72 1.88 ⁎ ⁎⁎
(0.40) (0.33) (0.37) ⁎⁎

Results from multinomial ordered-probit regressions. Standard errors adjusted for country–year clusters are in parentheses. Inflation and unemployment rates are
three-year moving averages. All regressions include country and year dummies (France and 1975 are the base), time trend, GDP per capita (in 2000 constant US$)
and personal characteristics; DV u Low EPL: unemployment slope dummy variable for individuals in low job protection countries; DV π Low EPL: inflation slope
dummy variable for individuals in low job protection countries. ⁎Unemployment is not significant in the estimate. ⁎⁎The Trade-Off Indexes for column 4 are 4.08
(2.95) for low EPL and.15 (0.46) for high EPL.

unemployment are considered jointly. (RC 2) Using the second methodology we run four one-stage separate regressions by age
groups in which membership of the two (low and high) employment protection groups is introduced as a slope dummy variable.
(RC 3) By means of the third methodology we run a single one-stage overall sample regression with both age and job protection
slope dummies. (RC 4) With the fourth methodology, age and employment protection slope dummies in the single overall sample
regression are interacted. (RC 5) With the fifth, we check the extent to which the influence of social norms14 on the importance of
work (see Stutzer and Lalive, 2004) affects our results. To test this latter point we use data from the World Value Survey on the
responses to the statement work is a duty towards society and find that the correlation between this statement and EPL is positive
and equal 0.28. Social norms therefore have a certain role.
In a second battery of checks (Table A3 in the appendix) we control whether our findings are robust to changes in the
unemployment and inflation indicators. To this end, we consider (RC 6 and RC 10) the prime age (25–54) male unemployment
rate15 and alternative inflation measures such as (RC 7) the OECD GDP deflator and (RC 8) the GDP Consumer Price Index. We do
not find significant differences with respect to our main results. As an additional control (RC 9), we further introduce the return on
the stock market index to proxy for wealth effects. All these robustness checks confirm that the effects of age and job protection on
the cost of unemployment are remarkably stable and independent from each other. For the sake of brevity, here we report only the
TOI values of these last estimates: the full estimate results are omitted for reasons of space and are available upon request.

2.1. Discussion of our results

It is rather difficult in most economic relationships to identify the causality nexus in the correlation between two variables. In
our case, we may seek to interpret the results by presuming that regulation affects psychological costs of inflation and
unemployment, and not vice versa. But note that, in a reverse causality nexus, when the costs of unemployment are higher, we
would expect people to demand greater employment protection. By contrast, what we observe in Table 3 is exactly the opposite;
and aside from the reverse causality hypothesis, it is also puzzling, because we would expect higher job market flexibility to reduce
the cost of the unemployment status in low job protection countries. Our findings show that the opposite holds. The puzzle
persists in the various robustness checks outlined in the previous paragraph.
Possible explanations of the puzzle are habituation effects (see, for example, Winkelmann and Winkelmann, 1998 and Clark
et al., 2001) among the relatively higher share of long-term unemployed, stronger informal economies in high job protection
countries, and the role of social norms.16 The informal economy explanation is supported by empirical evidence. Using data from

14
The well-known definition of a social norm is “a behavioural regularity; that is […] based on a socially shared belief how one ought to behave; which triggers
Q4 […] the enforcement of the prescribed behavior by informal social sanctions” (Fehr and Gachter, 2000).
15
Our choice is to avoid institutional distortions in the definition of the employment, unemployment and non participation status in the older age cohorts.
16
In high EPL countries, unemployed people retain this status for long time. Hence they become accustomed to it and end up by suffering less. Conversely, in
low EPL countries greater job market flexibility implies a higher share of short-term unemployment, so that the jobless often find new jobs before they have
become accustomed to their unemployment status.
142 L. Becchetti et al. / European Journal of Political Economy 26 (2010) 137–146

Table 4
Population age and EPL index in selected countries.

Country Average population age % Pop. > 65 EPL index U% Π% U/Π

Germany 42.6 17.8 20 8.77 1.41 6.22

Italy 42.2 19.3 23 10.34 2.75 3.75
France 39.1 16.5 21 10.68 1.56 6.83
EMU 41.2 17.2 17.8 9.86 2.40 4.12
UK 39.3 15.9 2 6.04 2.62 2.30
United States 36.5 12.3 1 5.06 2.46 2.06

Average population age is from the CIA World Facts Book (2006). The percentage of population over 65 is from the World Bank's WDI (2003). The EPL is from the
OECD Employment Outlook (1999) and refers to the average situation in the late 1990s. Unemployment and Inflation rates are averages over the period 1995–2004.

Schneider (2002) we find a positive correlation between EPL and the size of the shadow economy (0.43), the latter variable being
positively correlated with the high EPL dummy and negatively correlated with the low EPL dummy (0.61 and −0.41 respectively).
To determine the extent to which this matters in our estimates, we split the sample into three groups: high EPL/high shadow
economy group (6000 observations), high EPL/low shadow economy group (46,842 observations), and low EPL low shadow
economy groups (we have no low EPL high shadow economy countries). The individual costs of unemployment for the three
groups are respectively −00.47, −0.34 and −0.39. When controlling for the size of the shadow economy, the relationship
between EPL and the individual cost of unemployment is as expected, but the difference in TOIs between high and low EPL
countries persists.
Another related interpretation receiving important empirical support is the one based on the role of social norms. The point is
very well illustrated by Stutzer and Lalive (2004), who show in an empirical paper on Swiss cantons the positive and significant
correlation between unemployment stigma and the share of votes against unemployment benefits in a federal referendum and, in
turn, the significant relationship between the latter and the psychological cost of unemployment. Stutzer and Lalive also highlight
that the stronger the norm, the more rapidly the unemployed find new jobs. To return to our findings, if the reasoning by Stutzer
and Lalive (2004) works, countries with stronger social norms and unemployment stigma have both higher costs of
unemployment and lower EPL protection for different reasons: people must cope on their own, unemployment is associated
with a stigma, and its duration must be as short as possible (which implies higher job flexibility and less employment protection).
We have three final remarks. First, although we are interested in capturing the effect of age per se (since retired people and students
have the same voting rights as working people), we have run one-stage regressions with only employed persons in the sample, the
results remaining unchanged. Second, it would be interesting to check for assets, and for family rather than personal income, since the
impact of unemployment may diminish considerably if the spouse has a high salary. Unfortunately, the Eurobarometer does not
provide information on these aspects. Third, although most of the countries with high employment protection provide higher benefits
aimed at supporting and re-skilling unemployed workers, in some countries like Denmark the so-called flexicurity approach combines
low firing costs with generous benefits aimed at re-skilling and re-employing displaced workers.
The natural extension of this would be to disentangle the effect of employment protection from that of social security.
However, since in Western Europe only Denmark seems to be characterised by both flexibility and social security (if we adopt a
broader classification, the phenomenon is still limited to very few country–years), at the moment we do not have sufficient
information to test this new phenomenon.

3. Age, employment protection and economic policies

Are political and economic institutions aware of the different psychological costs of unemployment in low and high job protection
countries, and do their statutory principles reflect these concerns? Some useful information in this regard can be obtained by
examining Charters and observing the behaviour of different Central Banks, and also how the UK government decided on whether or
not to join the MU (see footnote 5). In the debate on monetary policy strategies, economists often contrast the differing attitudes of the
Fed and the ECB toward price stability. Whilst the Federal Reserve Act17 gives the same importance to price stability and employment,
the European Central Bank Statute18 stipulates that low inflation should be the Bank's fundamental priority.19

17
Federal Reserve Act (1913), Section 2A—Monetary Policy Objectives: “The Board of Governors of the Federal Reserve System and the Federal Open Market
Committee shall maintain long run growth of the monetary and credit aggregates commensurate with the economy's long run potential to increase production,
so as to promote effectively the goals of maximum employment, stable prices, and moderate long-term interest rates.”
18
Protocol on the Statute of the European System of Central Banks and of European Central Bank. Objectives and tasks of the ECB (1992). Article 2: “In accordance
with Article 105(1) of this Treaty, the primary objective of the ESCB shall be to maintain price stability. Without prejudice to the objective of price stability, it
shall support the general economic policies in the Community with a view to contributing to the achievement of the objectives of the Community as laid down in
Article 2 of this Treaty. The ESCB shall act in accordance with the principle of an open market economy with free competition, favouring an efficient allocation of
resources, and in compliance with the principles set out in Article 4 of this Treaty.”
19
Since 2003, the ECB has adopted the following definition of price stability: inflation is “below but close to 2% over the medium term” (ECB, 2003, p. 51).
Before 2003, the ECB defined it as a yearly increase in the harmonized index of consumer prices (HICP) below 2%, where “below 2%” could be anything (even a
negative inflation rate). Galí et al. (2004) see this change of definition as a possible “preparatory move before an eventual increase in the target inflation rate”.
 L. Becchetti et al. / European Journal of Political Economy 26 (2010) 137–146 143

Before drawing conclusions we must nonetheless consider the time when, and the different cultural environments
in which, the two Statutes were written, although nothing has prevented the FED from changing its Statute over
time. Furthermore, empirical analyses based on the Taylor's rule fail to capture this difference in the recent behaviour
of the two Central Banks (see Sardoni and Wrai, 2005; Cobham, 2006), even though analysis of the minutes
seemingly reveals the FED's greater concern for active employment policies. Many factors shape institutions around
the world, their statutes, and their priorities. The demographic structure and the level of employment protection may
help explain, among other things, why some central banks have developed a stronger antinflationary stance than
others.
Table 4 shows that the European countries which contributed most to shaping the ECB and its policies (countries
which are characterised by older populations and higher employment protection) displayed higher unemployment rates
and a higher unemployment/inflation ratio with respect to the UK and the US over the period 1995–2004. 20 The United
Kingdom, with its younger population, and especially its very low EPL index, records a much lower unemployment rate
and a slightly higher inflation than the EMU countries on average. A greater fear of unemployment may have played
some role in the decision taken by the British authorities not to join the EMU.21 Finally, the United States are
characterised by much younger population, a lower share of over-65s with respect to EMU countries, and an EPL index
close to zero.

4. Conclusions

Policymakers should bear in mind that the relative costs of inflation versus unemployment for voters are markedly
heterogeneous across age cohorts and institutional settings, especially if job market regulation is considered.
Furthermore, authorities should take into account that in the case of inflation and unemployment, two radically
different things are being compared. On the one hand there is an illness which directly affects only some individuals
(unemployment) but creates additional costs and engenders a fear of contagion in many others. On the other hand there
is a pervasive factor (inflation) which affects everyone (although individuals may have different capacities to insulate
themselves against it). From this perspective, age and job protection are two fundamental variables explaining the
heterogeneity of individuals' relative preferences on inflation and unemployment, since the probabilities of contagion by
the unemployment illness for those actually employed are markedly different when the two above mentioned factors
vary.
In light of this intuition, we have shown that the negative effect of unemployment on individual happiness is much lower for
those individuals who, because of age class and institutional framework, are less likely to be affected by the illness (or have
relatively reduced consequences from the infection). This implies that an average marginal rate of substitution between
unemployment and inflation, calculated on the entire sample population (1.91 with a 91% higher weight on unemployment),
conceals dramatically different rates of substitution according to different age cohorts and domestic job market rules. Such
heterogeneity may push the index well above two for low job protection countries, and to one and below for high job protection
ones.
Our findings well match the variability of institutional designs and policymakers' behaviour in countries which differ in
demographics and labour market rules. Although we have illustrated this point in the paper by comparing CBs statutes in the EU,
UK and US, we are aware that this does not necessarily imply a causal link from labour market rules to psychological inflation/
unemployment trade-offs and CBs' behaviours. The alternative of a social norm which stigmatizes unemployment, and
simultaneously explains the higher cost of unemployment and the lower EPL protection in some countries, is also consistent with
what we have observed. Finally, when possible explanations for the heterogeneous behaviour of CBs are considered, it should be
borne that differences in economic fundamentals and in reactions to shocks may generate heterogeneous behaviours even in the
presence of a common objective function.

Acknowledgments

We wish to thank Michele Bagella, Carlo Borzaga, Marcello De Cecco, Rafael Di Tella, Zeno Enders, Armin Falk, Andrew Oswald,
Paolo Paesani, Filipa Sa, Marika Santoro, participants to the seminars at the Universities of Rome, Trento and Bonn and to the
happiness conferences held in Siena and Cassino, and two anonymous referees for their useful comments and suggestions. The
usual disclaimer applies.

20
The main reason for focusing on the considered period is to investigate what happens after the start of the ECB. As it is well known, many other
factors may affect observed levels of inflation in different countries in the pre-ECB period. Inflation in the seventies and the beginning of the eighties is
strongly dependent from the oil shock and the different capacities of European countries of sterilizing its effect or not in wage contracts (in Italy the
trade union strength obtained wage indexation on imported inflation thereby fuelling inflation). This is another reason for looking at the more recent
period.
21
See footnote 5.
144 L. Becchetti et al. / European Journal of Political Economy 26 (2010) 137–146

Appendix A

Table A1
Summary statistics.

Variable Observations Mean Std. dev. Min Max

Life satisfaction 516,697 3.05 0.75 1 4

Unemployed 936,845 0.06 0.24 0 1
Self-employed 936,845 0.01 0.12 0 1
Retired 936,845 0.19 0.4 0 1
Student 936,845 0.09 0.29 0 1
Home 1,094,032 0.15 0.36 0 1
Male 936,845 0.48 0.5 0 1
Age 904,325 43.48 17.96 15 99
Age Sq. 904,325 2213 1694 225 9801
Middle education 745,901 0.44 0.5 0 1
High education 745,901 0.3 0.46 0 1
Married 831,965 0.61 0.49 0 1
Divorced 831,965 0.04 0.2 0 1
Separated 831,965 0.01 0.11 0 1
Widowed 831,965 0.08 0.27 0 1
Income 2nd quartile 662,817 0.36 0.48 0 1
Income 3rd quartile 662,817 0.24 0.42 0 1
Income 4th quartile 662,817 0.18 0.38 0 1
GDP 936,845 1.94 0.66 0.61 4.6
Inflation 936,845 0.04 0.04 − 0.01 0.25
Unemployment 921,325 0.09 0.04 0.01 0.23
EPL 1,056,200 13.52 6.57 1 25
Informal economy 776,619 17.41 5.21 8.6 29
LTU 913,610 0.44 0.13 0.14 0.76

Table A2a
Robustness check 1, two-stage regressions.

2 3 4 5 6 7
<29 29–41 42–64 >64 Low EPL High EPL

Unemployed (1st stage) − 0.33 − 0.36 − 0.35 −0.15 − 0.35 − 0.29

(0.01) (0.01) (0.01) (0.03) (0.01) (0.01)
Unemployment rate (2nd stage) − 1.27 − 1.20 − 1.17 − 1.23 −2.70 − 0.41
(0.29) (0.29) (0.29) (0.29) (0.74) (0.58)
Inflation rate (2nd stage) − 0.90 − 0.91 − 0.94 −0.98 − 2.14 − 1.15
(0.21) (0.21) (0.21) (0.21) (0.60) (0.35)
Trade-Off Index 1.79 1.71 1.62 1.41 1.43 0.61

Standard errors are in parentheses. The coefficients of the unemployed status are the result of first-stage OLS regressions controlling for personal characteristics,
year and country dummy variables. The coefficients of unemployment and inflation rates (three-year moving averages) are from the second-stage OLS estimates.
The dependent variable of the second-stage is the regression-corrected life-satisfaction level from the first stage. All second-stage regressions include country and
year dummies (France and 1975 are the base), time trend, GDP per capita (2000 constant US$) and personal characteristics. The Trade-Off Index shows the ratio
between the social cost of unemployment and the social cost of inflation. Since the TOI are the result of coefficients from different (first- and second-stage)
regressions, it is not possible to calculate their standard errors.

Table A2b
Robustness checks 2–4: alternative methodologies.

Trade-Off Indexes

Age group Low EPL High EPL

Rob. check 2
Four separate <29 4.10 (2.53) 0.23 (0.23)
Regressions run 29–41 3.02 (1.41) 0.49 (0.22)
By age groups, with EPL 42–64 5.07 (2.84) 1.08 (0.34)
Slope dummy variables >64 2.74 (1.72) − 0.29 (0.71) a

Rob. check 3
One single regression <29 2.56 (1.03) 0.47 (0.25)
With EPL and age slope 29–41 3.58 (1.73) 0.77 (0.27)
Dummy variables 42–64 3.05 (1.18) 0.73 (0.24)
>64 2.35 (1.00) 0.36 (0.24)
 L. Becchetti et al. / European Journal of Political Economy 26 (2010) 137–146 145

Table A2b (continued)

Trade-Off Indexes

Age group Low EPL High EPL

Rob. check 4
One single regression <29 2.26 (0.71) 0.30 (0.28)
Run with EPL and age 29–41 4.16 (2.42) 0.54 (0.27)
Interacting slope 42–64 4.40 (2.43) 0.69 (0.26)
Dummy variables >64 − 18.49 (75.94) b 0.45 (0.25)

Reported TOIs are given by the sum of the coefficients of the base and the slope dummy variable. Standard errors are in parentheses.
a
Not significant.
b
Trade-Off index in this case explodes because the inflation coefficient is close to zero, but is not significant.

Table A3
Robustness checks 6–9: alternative regressors.

Trade-off across age cohorts Full sample <29 29–41 42–64 >64

(6) U prime age male by age 1.19 0.76 1.05 1.62 1.12
(0.27) (0.20) (0.21) (0.42) (0.56)
(7) GDP deflator (OECD) 2.24 1.47 2.13 3.20 1.50
(0.68) (0.50) (0.81) (1.06) (0.51)
(8) CPI World Bank 1.12 0.54 0.87 1.90 0.88
(0.32) (0.29) (0.28) (0.59) (0.45)
(9) Stock market index 1.46 1.14 1.23 1.93 1.30
(0.29) (0.30) (0.28) (0.39) (0.40)

Trade–off across EPL groups Full Sample Low EPL High EPL

(10) U prime age male by EPL subs. 1.19 2.63 − 0.11

(0.27) (0.64) (0.40) a
(11) Continuous EPL by EPL subsample 2.43 1.89 0.40
(0.56) (0.38) (0.77)

Reported TOIs are given by the sum of the coefficients of the base and the slope dummy variable. Standard errors are in parentheses.
a
Not significant.

References

Akerlof, G., Dickens, W., Perry, G., 1996. The macroeconomics of low inflation. Brookings Papers on Economic Activity 1, 60–76.
Barro, R., Gordon, D., 1983. Rules discretion and reputation in a model of monetary policy. Journal of Monetary Economics 12, 101–121.
Blanchflower, D., 2007. Is unemployment more costly than inflation. NBER Working Paper 13505.
Brown, G., 1997. Speech to the Royal Institute for International Affairs. HMSO, London. www.hm-treasury.gov.uk.
Clark, A., Oswald, A., 1994. Unhappiness and unemployment. Economic Journal 104, 648–659.
Clark, A., Georgellis, Y., Sanfey, P., 2001. Scarring: the psychological impact of past unemployment. Economica, London School of Economics and Political Science
68, 221–241.
Clark, A., Frijters, P., Shields, M., 2006. Income and happiness. Evidence, explanations and economic implications. Paris Jourdan Science Economiques, Working
Paper 24, 1–65.
Cobham, D., 2006. Using Taylor rules to assess the relative activism of the European Central Bank, the Bank of England and the Federal Reserve Board. CDMA
Conference Paper Series 0602. Centre for Dynamic Macroeconomic Analysis.
Darity, W., Goldsmith, A., 1996. Social psychology, unemployment and macroeconomics. Journal of Economic Perspectives 10, 121–140.
De Grauwe, P., 2005. Economics of Monetary Union. Oxford University Press, New York.
Di Tella, R., MacCulloch, R., 2006. Some uses of happiness data. Economics Journal of Economic Perspectives 20, 25–46.
Di Tella, R., MacCulloch, R., 2007. Happiness, contentment and other emotions for Central Banks. NBER Working Paper 13622.
Di Tella, R., MacCulloch, R., Oswald, A., 2001. Preferences over inflation and unemployment: evidence from surveys of happiness. American Economic Review 91,
335–341.
Di Tella, R., MacCulloch, R., Oswald, A., 2003. The macroeconomics of happiness. Review of Economics and Statistics 85, 809–827.
Dickens, W., 2001. Comments on Charles Wyplosz Paper: do we know how low should inflation be? Why price stability? First ECB Central Banking Conference.
European Central Bank, Frankfurt.
European Central Bank, 2003. The ECB's monetary policy strategy. Press Release, 8th of May 2003, www.ecb.int.
European System of Central Banks, 1992. Protocol on the Statute of the European System of Central Banks and of European Central Bank. Objectives and tasks of the
ECB. www.ecb.int.
Feather, N., 1990. The psychological impact of unemployment. Springer, New York.
Federal Reserve Act (as amended by acts of 1978, 1988 and 2000), 1913. www.federalreserve.gov.
Fehr, E., Gachter, S., 2000. Cooperation and punishment in public goods experiments. American Economic Review 90, 980–994.
Ferrer-i-Carbonell, A., Frijters, P., 2004. How important is methodology for the estimates of the determinants of happiness? Economic Journal 114, 641–659.
Frey, B., 2008. Happiness: A Revolution in Economics. CES, The MIT Press, Cambridge.
Frey, B., Stutzer, A., 2002. What can economists learn from happiness research. Journal of Economic Literature 40, 402–435.
Galí, J., Gerlach, S., Rotemberg, J., Uhlig, H., Woodford, M., 2004. The Monetary Policy Strategy of the ECB Reconsidered: Monitoring the European Central Bank.
Centre for Economic Policy Research, London.
HM Treasury, 1997. U.K. Membership of a Single Currency: An Assessment of the Five Economic Tests. HMSO, London. www.hm-treasury.gov.uk.
Lazear, E., 1990. Job security provisions and employment. Quarterly Journal of Economics 105, 699–726.
Nickell, S., 1997. Unemployment and labour market rigidities: Europe versus North America. Journal of Economic Perspectives 11, 55–74.
146 L. Becchetti et al. / European Journal of Political Economy 26 (2010) 137–146

Nickell, S., Nunziata, L., Ochel, W., 2005. Unemployment in the OECD since the 1960s. What do we know? The Economic Journal 115, 1–27.
OECD, 1999. Employment Outlook. OECD Publishing, Paris.
OECD, 2004. Employment Outlook. OECD Publishing, Paris.
Rogoff, K., 1985. The optimal degree of commitment to an intermediate monetary target. The Quarterly Journal of Economics 100, 1169–1189 (November, MIT
Press).
Sahuc, J.G., Smets, F., 2008. Differences in interest rate policy at the ECB and the Fed: an investigation with a medium-scale DSGE model. Journal of Money, Credit
and Banking 40, 505–521.
Sardoni, C., Wrai, R., 2005. Monetary policy strategies of the European Central Bank and the Federal Reserve Bank of the U.S. InThe Levy Economics Institute of Bard
College. Working Paper 431.
Schneider, F., 2002. Size and measurement of the informal economy in 110 countries around the world. Rapid Response Unit, The World Bank, Washington, DC.
Stutzer, A., Lalive, R., 2004. The role of social work norms in job searching and subjective well-being. Journal of the European Economic Association 2, 696–719.
Winkelmann, L., Winkelmann, R., 1998. Why are the unemployed so unhappy? Evidence from panel data. Economica 65, 1–15.
Wyplosz, C., 2001. Do we know how low should inflation be? In: Garcia Herrero, A., Gaspar, V., Hoodguin, L., Morgan, J., Winkler, B. (Eds.), Why Price Stability?
European Central Bank.