The Economics of Happiness and Health

1

• The Economics of Happiness and Health
• Andrew Oswald
• IZA and Warwick
• I would like to acknowledge that much of this
  work is joint
• with coauthors Chris Boyce, Andrew Clark, Nick
  Powdthavee,
• David G. Blanchflower, and Steve Wu.

2

• This week Id like to propose a number of
  ideas.

3
1

• Happiness data offer us interesting potential
  as proxy-utility data.
• u u(y, z, ..)

4
Regression equations

• Mental well-being f(Age, gender, education
  level, income, marital status, friendship
  networks, region, year)

5
We now know

• There is a lot of regularity in these
  regression-equation patterns, across countries
  and well-being measures.
• Fairly robust to panel estimators and different
  methods.
• Progress can be made on causality.

6

• One potentially important implication

7
If this form of function can be estimated (and K,
L, M are life events)

• Happiness a bK cL dM eY
• where Y is income,

8
If this form of function can be estimated (and K,
L, M are life events)

• Happiness a bK cL dM eY
• where Y is income, then we may be able to use
  such equations to calculate the implied dollar
  value of the happiness from life events K, L, M.

9

• Monetary equivalences
• A life satisfaction equation
• Life satisfaction B1income B2Event error
• Marriage - 100,000 (Blanchflower and Oswald,
  2004), Neuroticism - 314,000 (Boyce et al., in
  press), Widowhood (175,000-496,000), Health
  limiting daily activities (473,000) (Powdthavee,
  van den Berg, 2011)

10
2

• The next 20 years are likely to see economists
  work more and more with physiological and
  hard-science data.

11
3

• Biomarker data will (slowly) be used more and
  more in economics.

12
4

• Empirically, there are strong relative effects
  on utility

13
4

• Empirically, there are strong relative effects
  on utility
• u u(y, y)
• eg. if y is others incomes.

14
5

• A crucial role in social-science behaviour is
  played by the second derivative, v?, of the
  function
• utility v(relative status) ..

15
In humans (I shall argue)

• Concavity of v(.) leads to imitation and herd
  behaviour
• Convexity of v(.) leads to deviance.

16
6

• The Stiglitz Commissions ideas will eventually
  take hold.

17

• Stiglitz Report 2009
• Measures of .. objective and subjective
  well-being provide key information about peoples
  quality of life. Statistical offices worldwide
  should incorporate questions to capture peoples
  life evaluations, hedonic experiences in their
  own survey. P.16. Executive Summary of
  Commission Report.

18

• So........

19
Could we perhaps learn
20
..how to make whole countries happier?
21

• Preferably not like this

22
Germany 4 England 1
23
Germany 4 England 1
24
Useful introductions

• Relative Income, Happiness and Utility An
  Explanation for the Easterlin Paradox and Other
  Puzzles (Andrew Clark, Paul Frijters and Mike
  Shields), Journal of Economic Literature, 2008.
• The Happiness Equation (Nick Powdthavee), Icon
  Books, 2010.

25

• This is a good time for general questions if
  people would like to ask some?

26

• Now lets think about how human beings report
  their feelings (for example, in a survey).

27

• First, they have genuine feelings inside
  themselves (about how happy they are, say).

28

• Second, they make a decision about how to report
  those feelings.

29

• There are then two processes going on inside a
  person.

30

• Human feelings
• Human reporting

31

• Lets think of the example of money and peoples
  well-being.

32
Assume

33
Assume

• People get true happiness, h, from income, y.
  Call it h(y).

34
Assume

• People get true happiness, h, from income, y.
  Call it h(y).
• They give a number for this, which is their
  reported happiness, r. Call it r(h).

35
The Reporting Function

36
The Reporting Function

• Write R(y) which is reported happiness as a
  function of income.
• This is what is studied in well-being regression
  equations.

37

• Now think of the function-of-a-function rule
  in calculus.

38
By definition

• R(y) r(h(y))

39
By definition

• R(y) r(h(y)) so
• R'(y) r'(h) h'(y) gt 0
• where y is income.

40
(No Transcript)
41
In the cross-section, income is positively
correlated with happiness
Take America in 1994 for example
42
From Deaton-Kahneman in PNAS 2010
43
Now lets think of the second derivative

44
The first derivative earlier was

• R'(y) r'(h) h'(y)

45
The first derivative earlier was

• R'(y) r'(h) h'(y)
• where y is income, r is reported happiness, h
  is actual happiness.

46
Think of the second derivative

• The curvature of reported happiness is

47
Think of the second derivative

• The curvature of reported happiness is
• R?(y) r?(h) h'(y) h'(y)
• r'(h) h?(y)

48

• But if R?(y) is found to be negative that does
  not prove that h?(y) is negative.
• R is reported happiness
• h is true happiness

49

• Hence there are lots and lots of papers in the
  literature that get this wrong.

50
Reiterating why

• The curvature of reported happiness is
• R?(y) r?(y) h'(y) h'(y)
• r'(h) h?(y)

51

• Even if the estimated happiness function
  itself is concave, we cannot be certain that true
  happiness is concave.

52

• All social scientists (and many medical
  scientists) need to know more about the reporting
  function.

53

• So is there any way to make progress on this
  tricky issue?

54
Height as an example
55
113 Men and 106 Women

• The respondents were asked to record how tall
  they felt, using a continuous un-numbered line
  with the words very short written at the
  left-hand end to very tall at the right-hand
  end.

56
113 Men and 106 Women

• The respondents were asked to record how tall
  they felt, using a continuous un-numbered line
  with the words very short written at the
  left-hand end to very tall at the right-hand
  end.
• Numbers were coded 110 afterwards.

57

• Then we looked at the correlation between
  feelings of being tall and actual true height.

58
(No Transcript)
59

• How well correlated are feelings of height and
  actual height?

60
Feelings of height and actual height in 113 men
61
Feelings of height and actual height in 106 women
62

• These plots are consistent with a linear
  reporting function.

63

• Much more research on the reporting function
  r(.) will be required in the future.

64
Evidence from Neuroscience

• Positive feelings correspond to brain activity in
  the left-side of the pre-frontal cortex, above
  and in front of the ear
• Negative feelings correspond to brain activity in
  the same place in the right side of the brain

65
Happy and Sad Pictures
66
The Brain Responses to Two Pictures(MRI Scan)
Source Richard Davidson, University of Wisconsin
67
The types of statistical sources

• General Social Survey of the USA
• British Household Panel Study (BHPS)
• German Socioeconomic Panel
• Australian HILDA Panel
• Eurobarometer Surveys
• Labour Force Survey from the UK
• World Values Surveys
• NCDS 1958 cohort
• BRFSS

68
From the U.S. General Social Survey (sample size
40,000 Americans approx.)

• Taken all together, how would you say things are
  these days - would you say that you are very
  happy, pretty happy, or not too happy?

69
An alternative DRM approach

• A study by Daniel Kahneman and his colleagues on
  1,000 working women in Texas (see Kahneman et al,
  2003)
• These women were asked to divide the previous day
  into 15 episodes. They were then asked what they
  were doing in each episode, and who were they
  doing it with.

70
Happiness in Different Activities
71
Happiness while Spending Time with Different
People
The average reported feelings across 1,000 people
correspond well with activities predicted to be
good for us, as well as activities predicted to
be bad for us
72

• So how has the modern work on the economics of
  happiness proceeded?

73

• Here is a modern US happiness equation
  (courtesy of David Blanchflower, Dartmouth
  College and NBER)

74

• Could you turn to the NBER Blanchflower-Oswald
  paper on international happiness?

75
(No Transcript)
76
Some cheery news

77
Some cheery news

• In Western nations, most people are pretty happy
  with their lives.

78
Some cheery news

• In Western nations, most people are pretty happy
  with their lives.

79
Some cheery news

• In Western nations, most people are pretty happy
  with their lives.

80
Some cheery news

• In Western nations, most people are pretty happy
  with their lives.

81
The distribution of life-satisfaction levels
among British people
Source BHPS, 1997-2003. N 74,481
82
Exogenous shocks and happiness

• New work looks at
• Genes
• Lottery wins
• 9-11s effects
• Deaths of children
• Sporting results
• Movements in air pollution

83
Other work on happiness as causal

• John Ifcher and Homa Zarghamee, forthcoming in
  the AER, on happiness leading to different rate
  of time discount.
• Oswald, Proto, Sgroi on happiness leading to
  higher productivity.
• These randomly assign happiness.

84

• Is modern society going in a sensible direction?

85
This is an empirical question

• "Does Economic Growth Improve the Human Lot?"
  Richard Easterlin
• in Paul A. David and Melvin W. Reder, eds.,
  Nations and Households in Economic Growth Essays
  in Honor of Moses Abramovitz, New York Academic
  Press, Inc., 1974.

86

• We will focus on it tomorrow.

87

• Lets return for a moment to the
  microeconomics of human well-being

88

• What have we learned?

89
Big effects

• Unemployment
• Divorce
• Marriage
• Bereavement
• Friendship networks
• Health
• No effects from children but for
  grandchildren Nick Powdthavee

90

• There is also an intriguing life-cycle pattern

91
The pattern of a typical persons happiness
through life
92
Arthur Stone, Angus Deaton, et al (2010)
93
Overall well-being
94
Quadratic Life-Satisfaction in the US

• Steve Wu on BRFSS 2010 data
• age -.0030621agesq .0000419
• Again the U-shape.

95
A life satisfaction U-shape in age also exists in
many developing nations

• In World Values Survey data, there is a U-shape
  and it reaches its minimum at

96
A life satisfaction U-shape in age also exists in
many developing nations

• In World Values Survey data, there is a U-shape
  and it reaches its minimum at
• Brazil 37
• China 46
• El Salvador 48
• Mexico 41
• Nigeria 42
• Tanzania 46

97
Obviously life is a mixture of ups and downs
98

• Much of the recent research follows people
  through time.
• eg. Andrew Clarks work

99
The unhappiness from bereavement
100

• Human beings also bounce back from, say,
  disability.
• Work with N. Powdthavee, Journal of Public
  Economics, 2008

101
Life-Satisfaction Path of Those Who Entered
Disability at Time T and Remained Disabled in T1
and T2BHPS data 1996-2005
102
However, there is a downside to that
adaptability (eg. marriage)

103
However, there is a downside to that
adaptability (eg. marriage)

104
Is there income adaptation?

• Maybe.
• The joy of having higher income may also wear
  off

105
Source Di Tella et al (2008), German
Socio-Economic Panel
106
And should you invest in a baby?
107
Happiness and children
108
But people do not seem to adapt to joblessness

109
The evidence suggests that when a person is made
unemployed
110
The evidence suggests that when a person is made
unemployed

• 20 of the fall in mental well-being is due to
  the decline in income
• 80 is due to non-pecuniary things (loss of
  self-esteem, status..).

111

• An important question in a modern society is
  the impact of divorce.

112
Divorce (eventually) makes people happier
113
Divorce (eventually) makes people happier
114

• Points or questions?

115

• What about money and happiness?

116
A key social-science fact

117
A key social-science fact

• The data show that richer people are happier and
  healthier.

118

• But some general economists have low
  life-satisfaction when they hear about this
  research.

119

• The tradition of economics has been to ignore
  what people say about the quality of their own
  lives.

120

• The tradition of economics has been to ignore
  what people say about the quality of their own
  lives.
• Many are opposed to the idea of measuring
  happiness.

121
I always liked the retort

122
I always liked the retort

• If molecules could talk, would physicists refuse
  to listen?
• A. Blinder

123
I always liked the retort

• If molecules could talk, would physicists refuse
  to listen?
• A. Blinder

124
So how could we move forward?

125
So how could we move forward?

• Brain-science correlates as a validation

126
So how could we move forward?

• Brain-science correlates as a validation
• Physiological correlates as a validation

127
(No Transcript)
128
A brain-science approach (Urry et al
Psychological Science 2004)
129

• But, for a sceptic, there is a major
  difficulty.

130

• Biological data only validate well-being
  scores in so far as they are unambiguously
  measures of utility or happiness.

131
A killer question

• Can we devise a test in the economists spirit
  that shows, once and for all, a match between
  subjective well-being data and objective
  well-being data?

132

• Yes.

133

• I would like to give you the flavour of the
  argument in Oswald-Wu in Science in 2010.

134

• Are objective and subjective data on
  quality-of-life correlated?

134
135

• We can exploit neo-classical economic theory
  to assess the validity of well-being data.

136

• Think not about people but about places.

137
Joint work with Steve Wu

• New data from the Behavioral Risk Factor
  Surveillance System (BRFSS)
• 1.3 million randomly sampled Americans
• 2005 to 2008
• A life-satisfaction equation

138

• Then we go to the compensating-differentials
  literature dating back to Adam Smith, Sherwin
  Rosen, Jennifer Roback, etc.
• The most recent is Gabriel et al 2003.

139
Gabriel painstakingly takes data on

• Precipitation
• Humidity
• Heating Degree Days
• Cooling Degree Days
• Wind Speed
• Sunshine
• Coast
• Inland Water
• Federal Land
• Visitors to National Parks
• Visitors to State Parks
• Number of hazardous waste sites

140
and

• Environmental Regulation Leniency
• Commuting Time
• Violent Crime Rate
• Air Quality-Ozone
• Air Quality-Carbon Monoxide
• Student-teacher ratio
• State and local taxes on property, income and
  sales and other
• State and local expenditures on higher education,
  public welfare, highways, and corrections
• Cost-of-living

141

• Then there are 2 ways to measure human
  well-being or utility across space.
• Subjective and objective

142

• Gabriels work assigns a 1 to the state with
  the highest imputed quality-of-life, and 50 to
  the state with the lowest.

143

• So we need to uncover a negative association
  in order to find a match.

144

• And there is one.

145
One Million Americans Life Satisfaction and
Objective Quality-of-Life in 50 States
146
To conclude across US states

• There is a match between life-satisfaction
  scores and the quality of life calculated using
  (only) non-subjective data.

147
Some ideas to end

148
My hunch

149
My hunch

• The methods of the economics of happiness and
  mental well-being will slowly enter public life.

150
Other important applications
151
Other important applications

• The valuation of environmental amenities

152
Other important applications

• The valuation of environmental amenities
• The valuation of health states

153
Other important applications

• The valuation of environmental amenities
• The valuation of health states
• The valuation of emotional damages for the courts.

154
Conventionally

• Economics is a social science concerned with the
  efficient allocation of scarce resources

155

• We owe this definition to Lionel Robbins of
  the London School of Economics.
• For a long time, it served us well.

156

• But perhaps the time has come to think
  differently and to define economics differently.

157
An alternative definition
158
An alternative definition for 2011

• Economics is a social science concerned with the
  best way to allocate plentiful resources to
  maximize a societys well-being and mental health.

159
Looking ahead

• Policy in the coming century may need to
  concentrate on non-materialistic goals.

160
Looking ahead

• Policy in the coming century may need to
  concentrate on non-materialistic goals.
• GNH not GDP.

161
And the next research area?
162
(No Transcript)
163
Thank you.
164

• The Economics of Happiness and Health
• Andrew Oswald
• Research site www.andrewoswald.com
• I would like to acknowledge that much of this
  work is joint
• with coauthors Chris Boyce, Andrew Clark, Nick
  Powdthavee,
• David G. Blanchflower, and Steve Wu.