Carol Graham University of Maryland/ The Brookings Institution

The Brookings Institution, Washington, D.C. www.brookings.edu

Can Happiness Research Contribute to Development

Economics and Policy?Presented to the World Bank

February 13, 2007

Carol GrahamUniversity of Maryland/

The Brookings Institution


Why Happiness Economics?

• New method combining tools and methods of economists with those typically used by psychologists

• Method captures broader elements of welfare than do income data alone

• Method is uniquely well suited for analyzing questions where revealed preferences do not provide answers, for example the welfare effects of institutional arrangements individuals are powerless to change (like inequality or macroeconomic volatility) and/or behaviors that are driven by norms or by addiction and self control problems (alcohol and drug abuse, smoking, obesity)

• While economists traditionally have shied away from reliance on surveys (e.g. what people say rather than what they do), there is increasing usage of data on reported well being (happiness):

• a) consistent patterns in the determinants of well being across large N samples across countries and across time

• b: econometric innovations help account for error and bias in survey data (AND with the error that exists in all kinds of data!!)


Why NOT Use Happiness Surveys

• Biases in the way people answer surveys (question ordering/random events)• Adaptation – at individual and country levels• a) individual level: some psychologists believe that people ALWAYS adapt to

their set point, even after extreme events like divorce or spinal cord injuries; THUS if a poor peasant, who has adapted to his/her condition and/or has low aspirations due to lack of information reports he/she is happy, what can development economists do with this information?

• b) country level: Easterlin paradox - average happiness levels have not increased as rich countries get richer and make improvements in other areas such as health, education; BUT………………

• how long does it take individuals to adapt to these extreme changes? Equilibrium could be a LONG time away…..Do we not care about several years of very low well being levels caused by certain phenomena?

• most relevant information is about individual well being; country level averages do not tell us much and it is difficult to control for error/cultural traits, etc. Do we really care if Nigerians are happier than Ghanaians just because they have a tendency to respond in a more cheerful manner?

• WITHIN countries, wealthier, healthier, and more educated people are happier than poorer, less healthy, and less educated ones and have more time to enjoy those lives


What can this approach contribute to development? To the debate on globalization, poverty, and inequality?

• Major discrepancies between positive assessments of the benefits of globalization for the poor and for poor countries by economists, and the more negative assessments of the typical lay person on the street

• In some contexts, these assessments are positive; in others they are very negative and are even associated with widespread public frustration, political protest, and civic unrest

• Our research – based on well being surveys of thousands of individuals in developing countries world-wide finds that – counter to the received wisdom - public frustration is NOT most evident among the poor, but rather among upwardly mobile, low and middle income individuals

• Frustration is associated with mobility, inequality, and insecurity – e.g. with dynamic trends - rather than with static poverty

• Our approach is based on analyzing the dynamics of poverty and inequality, rather than using the usual static measures and focus on equilibrium

• It is also based on comparing perceptions of well being to objective, income-based measures



Relative Economic Mobility Matrices

United States, 1979-1989

1989 Q

1979 QBottom quintile II III IV Top quintile Total

Bottom quintile 61 24 9 5 1 100

II 23 33 28 14 3 100

III 8 25 30 26 11 100

IV 5 13 23 33 26 100

Top quintile 3 5 11 23 59 100

Total 100 100 100 100 100 100

Source: Mishel at al. (1999)

Peru, 1991-2000

2000 Q

1991 QBottom quintle II III IV Top quintile Total

Bottom quintile 45 25 19 6 5 100

II 25 25 23 14 13 100

III 16 23 22 20 19 100

IV 11 18 18 32 21 100

Top quintile 3 9 18 28 42 100

Total 100 100 100 100 100 100




Peru, 2000

Whole Sample

(N=500)Frustrated Achievers

(N=128)Non-Frustrated Achievers

(N=160) DifferenceFA v. NFA

Mean Std. Dev. Mean Std. Dev. Mean Std. Dev.

Age 52.95 15.29 55.67 15.09 49.49 14.90 **

Area (Urban=1) 0.86 0.35 0.93 0.26 0.78 0.42 ***

Gender (Male=1) 0.53 0.50 0.51 0.50 0.57 0.50

Education 8.02 4.66 8.03 4.52 8.12 4.68

Equivalence HouseholdExpenditure '00

8922 7314 9885 6144 10809 9957

Coefficient of Variation(91, 94, 96, 00)

0.42 0.19 0.43 0.19 0.48 0.19 **

Economic Satisfaction 2.91 0.80 2.53 0.78 3.21 0.64 **

Prospect of UpwardMobility

3.29 1.03 3.03 1.13 3.54 0.89 **

Significance in the difference of the means between Fas and non-FAs: * at the 10%, ** at the 5%, *** at the 1% level.


Whole Sample

(N=2289)Frustrated Achievers

(N=217)Non-Frustrated Achievers

(N=90) Difference FA v NFA

Mean Std. Dev. Mean Std. Dev. Mean Std. Dev.

Age 54.47 15.40 51.58 13.97 50.37 16.09

Gender (Male=1) 0.21 0.41 0.24 0.43 0.25 0.43

Education 8.41 2.35 8.62 2.13 8.91 1.97

Equivalence Household Income '98

2698 2935 4753 5964 6114 5574 ***

Coefficient of Variation (95, 96, 97, 98)

0.56 0.38 0.64 0.62 0.55 0.23 ***

Life Satisfaction 1.91 1.00 1.82 0.88 2.45 1.25 ***

Economic Ladder Question

2.93 1.48 3.00 1.56 3.72 1.52 ***

Prospect of Upward Mobility

2.06 1.00 2.07 1.02 2.58 0.97 **

Pro-Democracy Attitude 0.53 0.54 0.45 0.53 0.70 0.51 ***

Satisfaction with Market Reform Process

0.74 0.55 0.65 0.59 0.96 0.57 ***

Fear of Unemployment 3.96 1.37 4.15 1.22 3.57 1.56 ***

Restrict the Rich 3.22 0.79 3.16 0.82 2.89 0.94 **

Russia, 1998

Significance in the difference of the means between Fas and non-FAs: * at the 10%, ** at the 5%, *** at the 1% level.



Happiness by Age LevelLatin America, 2000

18 26 34 42 50 58 66 74 82 90 98

years of age

leve

l of

hap

pin

ess


Table 2: Happiness in Latin America, 2001

Dependent Variable: Happiness

Independent variables Coef. z

Age -0.025 -4.21

Age squared 0.000 4.72

Male -0.002 -0.07

Married 0.056 1.63

Log wealth index 0.395 10.56

Years of education -0.003 -0.64

Minority -0.083 -2.49

Student 0.066 1.01

Retired -0.005 -0.06

Homemaker -0.053 -1.04

Unemployed -0.485 -7.54

Self employed -0.098 -2.33

Health (self-reported) 0.468 24.58

Pseudo R2 0.062

Number of obs. 15209

* Ordered logit estimation; country dummies included but not shown.

Source: Latinobarometro, 2001. Author's calculations.


Table 3: Happiness in Russia, 2000



Age -0.067 -7.42


Male 0.152 2.80

Married 0.088 1.40

Log equivalent income 0.389 11.48

Education Level 0.015 0.96

Minority 0.172 2.46

Student 0.199 1.59

Retired -0.378 -3.97

Housewife 0.049 0.33


Self employed 0.537 2.23

Health index 0.446 3.82

Pseudo R2 0.033

Number of obs. 5134

* Ordered logit estimation

Source: Graham, Eggers, Sukhtankar (forthcoming).


Table 4: Happiness in the U.S., 1972 - 1998



Age -0.025 -5.20


Male -0.199 -6.80

Married 0.775 25.32

Log income 0.163 9.48

Education 0.007 1.49

Black -0.400 -10.02

Other race 0.049 0.59

Student 0.291 3.63

Retired 0.219 3.93

Housekeeper 0.065 1.66


Self employed 0.098 2.29

Health 0.623 35.91

Pseudo R2 0.075

Number of obs. 24128

* Ordered logit estimation; year dummies included but not shown.

Source: GSS data, Author's calculations.


Dependent Variable: Log equivalence income, 2000 (OLS)

Independent variables coef t coef t coef t

Age -0.0133 -3.00 -0.0132 -2.97 -0.0146 -3.25

Age2 0.0001 3.18 0.0001 3.15 0.0002 3.52

Male 0.0102 0.42 0.0102 0.42 -0.0004 -0.02

Married 0.2053 7.84 0.2054 7.84 0.2050 7.84

Education level 0.0301 4.51 0.0301 4.51 0.0296 4.44

Minority 0.1213 3.98 0.1227 4.03 0.1216 4.00

Student -0.0336 -0.34 -0.0301 -0.31 -0.0367 -0.38

Retired -0.1906 -4.85 -0.1899 -4.83 -0.1659 -4.18

Housewife -0.2488 -3.90 -0.2492 -3.90 -0.2388 -3.73

Unemployed -0.3450 -8.16 -0.3435 -8.12 -0.3426 -8.07

Self-employed 0.1415 1.46 0.1411 1.46 0.1284 1.33

Health index 0.0601 1.11 0.0588 1.09 0.0559 1.04

Log-equiv income 1995 0.2420 18.11 0.2429 18.12 0.2244 15.69

Log-equiv income 1995, poor 0.0094 2.60

Log-equiv income 1995, rich 0.0180 4.36

Unexplained happiness, 1995 0.0298 2.64 0.0634 2.32 0.0269 2.38

Unexp. happiness, 1995, 2nd quintile -0.0436 -1.14

Unexp. happiness, 1995, 3nd quintile -0.0361 -0.95

Unexp. happiness, 1995, 4th quintile -0.0626 -1.71

Unexp. happiness, 1995, 5th quintile -0.0229 -0.65

Constant 5.8325 36.35 5.8234 36.19 5.9365 34.62

number of observations 4457 4457 4457

adjusted R-squared 0.1335 0.1333 0.1518

“Poor" is defined as bottom 40% of the income distribution in 1995; “Rich" is the top 20%. “Unexplained happiness” is the residual of basic happiness regression using only 1995 data. Independent variables are from 2000 unless otherwise noted.

The effects of happiness on income


Happiness Gap in Honduras and Chile

Happiness Gap = wealth gap * coefficient ÷ 4

Calculated Happiness Gap

Poor Rich

Chile wealth gap -2.489 2.521

Honduras wealth gap -2.142 3.261

Chile-Honduras difference 0.347 0.740

difference * coefficient / 40.43% 0.93%

= Honduran happiness differential

RICHPOOR

Average Chilean wealth: 7.8

Average Honduran wealth: 4.8

Poor Hondurans: wealth = 2.6Poor Chileans: wealth = 5.3

Rich Hondurans: wealth = 8.0Rich Chileans: wealth = 10.3

Honduran gap: 3.3

Chilean gap: 2.5

Honduran gap: 2.1

Chilean gap: 2.5


Conclusions• Happiness economics can give us a novel insights into the development process, which

complements those provided by income-based measures, accepting that there is a margin for error and for the role of personality traits driving results (which can, at times, be corrected for)

• Research highlights unhappiness/frustration related to volatility, insecurity, and inequality, even among upwardly mobile respondents. Some of this unhappiness is the result of rising expectations related to development. It may also reflect genuine grievances created or exacerbated by globalization related trends, such volatility, inequality, and a marked increase in information about the living standards of others

• We also have evidence that happiness and unhappiness are in turn linked to future outcomes in the labor market and in the health arena. There may also be a link to political behavior, although we only have information on attitudes rather than actual voting

• Most recent research is testing whether these attitudes – and in particular low expectations for the future and related unhappiness – can result in higher discount rates and less willingness to invest and save for the future, in the income, health, and political (e.g. support for reform) arenas (topic for future research)

• Other new research finds that happiness/optimism is highest among the poorest respondents in Africa – exploring whether it is due to selection bias (need to be optimistic to survive in such extreme contexts) )or whether it merely reflects individuals’ ability to adapt own expectations downwards but to maintain optimism for their children (again, topic for future research)


Relevance to World Bank Work?

• The discrepancies that we find in respondents’ subjective assessments of well being and objective income based measures provide policy relevant information. Why are the FA’s frustrated, for example, if they have upward mobility?

• Discrepancies can help explain why “rational” policies which seem to work can be rejected by taxpayers and consumers (because of hyperbolic discounting, risk aversion, horizontal inequalities, perceived inequalities, etc. – Russian unemployment findings, for example)

• Does happiness vary across regions? Surprisingly little in its main determinants but some minor differences show important differences among the regions: unemployment rates and fear of reform in Russia; optimism among the poor in Africa; relatively lower happiness rates in general in Eastern Europe even in places where incomes have increased (status changes vs income gains, etc).

• Some findings merely an insight into human psychology and adaptation to changing circumstances; some provide some hints about how to make economic reforms more politically palatable


Applications to Policy - Caveats

• Key to success of happiness measures as a survey instrument is lack of definition of happiness – it is left up to the respondent

• At the same time, how we define happiness will matter a great deal to its relevance to policy, and that entails normative judgements

• Happiness as contentment – relevance to policy? Happiness as defined by Kenny and Kenny as contentment, welfare, and dignity (smacks of Sen) seems a more appropriate policy tool

• Cardinality vs ordinality: do we care more about making someone who is unhappy happy than making a happy person happier? Surveys do not attach cardinal weights, but should policy?

• Happy peasant problem – rich person with high expectations who is miserable versus poor peasant with no information that is very happy

• Policies that can make people happy in the short term, like inflationary spending, may be very bad for their longterm welfare (unemployment in Russia, for example). At same time, particularly do to risk aversion and hyperbolic discounting, most reform policies will cause unhappiness in the short term. Can that be a gauge to policy?

• National well being indicators as a way to track and compare happiness across countries and time – complements to GNP?

Documents

Carol Graham University of Maryland/ The Brookings Institution