happiness and utility: evidence from new surveys daniel j. benjamin ori heffetz miles kimball alex...

12
Happiness and Utility: Evidence from New Surveys Daniel J. Benjamin Ori Heffetz Miles Kimball Alex Rees-Jones 1

Upload: griffin-mills

Post on 17-Dec-2015

212 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Happiness and Utility: Evidence from New Surveys Daniel J. Benjamin Ori Heffetz Miles Kimball Alex Rees-Jones 1

Happiness and Utility:Evidence from New Surveys

Daniel J. BenjaminOri Heffetz

Miles KimballAlex Rees-Jones

1

Page 2: Happiness and Utility: Evidence from New Surveys Daniel J. Benjamin Ori Heffetz Miles Kimball Alex Rees-Jones 1

Motivation• Our question: How does revealed preference

relate to measures of happiness?• In hypothetical scenarios with two options, we

measure both choice and predicted SWB. – Since we relate choice to predicted SWB, we are

not focusing on misprediction of SWB.– We find discrepancy between choice and various

common survey measures of subjective well-being (SWB).

• Discrepancy larger for felt happiness than for measures of overall life happiness / satisfaction.

• Discrepancy smaller for individuals with higher measured willingness to pay (WTP) for a better mood. 2

Page 3: Happiness and Utility: Evidence from New Surveys Daniel J. Benjamin Ori Heffetz Miles Kimball Alex Rees-Jones 1

Survey Designs• ~1500 patients waiting at an endocrinologist’s office in

Denver, CO, mostly for ongoing health maintenance (primarily diabetes).

• Receptionist handed out anonymous surveys.• Demographics:

– Gender: 21% male, 64% female, 15% no response– Mean age: 49– Race: 72% White/Caucasian, 4% Black/African-American, 4%

Hispanic/Latino, 3% other, 17% no response– Median income: ~$100k

• Also 222 Cornell undergraduates in WTP for felt happiness surveys.

3

Page 4: Happiness and Utility: Evidence from New Surveys Daniel J. Benjamin Ori Heffetz Miles Kimball Alex Rees-Jones 1

Choice Scenario: Sleep QuestionSay you have to decide between two new jobs. The jobs are exactly the same

in almost every way, but have different work hours and pay different amounts.

Option 1: A job paying $80,000 per year. The hours for this job are reasonable, and you would be able to get about 7.5 hours of sleep on the average work night.

Option 2: A job paying $140,000 per year. However, this job requires you to go to work at unusual hours, and you would only be able to sleep around 6 hours on the average work night.

Hypothesis: People will recognize that they’d be happier with more sleep, but choose the higher paying job due to responsibility or desire for status.

4

Page 5: Happiness and Utility: Evidence from New Surveys Daniel J. Benjamin Ori Heffetz Miles Kimball Alex Rees-Jones 1

5

Career Sleep Concert Commute Friends Relative Income

1 6 1 6 1 6 1 6 1 6 1 6Art Income More

SleepMore

IncomeDuty Fun More

IncomeShorter

CommuteFriends Income Absolutely

RicherRelatively

Richer

Choice 4.35(0.07)

3.08(0.09)

4.39(0.08)

3.00(0.08)

3.40(0.09)

3.30(0.09)

SWB 3.89(0.09)

2.73(0.08)

4.53(0.08)

3.22(0.08)

2.91(0.08)

3.39(0.08)

p-value 0.001 0.001 0.022 0.088 0.000 0.430

N 422 425 422 425 422 420

Page 6: Happiness and Utility: Evidence from New Surveys Daniel J. Benjamin Ori Heffetz Miles Kimball Alex Rees-Jones 1

Do Responses Make Sense?• Both choice and SWB monotonic in income difference.• Smaller deviations for individuals with stronger

preference for happiness.

6

$42K $60K $80K $100K0%

20%

40%

60%

80%

100%

Artist Question Responses$40K + Big Impact vs More

IncomeOption 2: choice + more satisfaction

Choice ≠ Satisfaction

Option 1: choice + more satisfaction

Income in Option 2

Page 7: Happiness and Utility: Evidence from New Surveys Daniel J. Benjamin Ori Heffetz Miles Kimball Alex Rees-Jones 1

How Well Do SWB Measures Predict Choice?

• We estimate (via non-linear least squares) that deviations between choice and SWB for overall life satisfaction / happiness are about 60% the size of deviations for felt happiness.

• Of our SWB measures, life satisfaction predicts choice the best.

SWB measure R-squared of choice on SWB

Life Satisfaction 0.69

Lifetime Happiness 0.59

Felt Happiness 0.56

7N~850 for each regression

Page 8: Happiness and Utility: Evidence from New Surveys Daniel J. Benjamin Ori Heffetz Miles Kimball Alex Rees-Jones 1

Preference (WTP) for Felt Happiness

Suppose you fell into a medium-serious depression. You can still function, but in terms of your mood you feel down most of the time and feel much less cheerful than you normally do. This has been going on for two months with no end in sight.

Your doctor tells you, and you confirm by checking things out yourself, that there is only one treatment that will get you back to feeling normal in terms of your mood. This treatment is fast, effective, and has no side effects, but only remains effective as long as you continue it.

8

Page 9: Happiness and Utility: Evidence from New Surveys Daniel J. Benjamin Ori Heffetz Miles Kimball Alex Rees-Jones 1

Freq. Response6% Wouldn’t accept even if it were free.12% Would pay a maximum of 0% of income7% Would pay a maximum of 1% of income19% Would pay a maximum of 2% of income16% Would pay a maximum of 5% of income13% Would pay a maximum of 10% of income9% Would pay a maximum of 20% of income3% Would pay a maximum of 33% of income7% Would pay a maximum of 50% of income8% No response2% Inconsistent responses

9Mean among consistent responses: 10%.

Page 10: Happiness and Utility: Evidence from New Surveys Daniel J. Benjamin Ori Heffetz Miles Kimball Alex Rees-Jones 1

WTP for Felt Happiness Predicts Current Felt Happiness

10

Felt happiness in last 24 hours

WTP (% income) for depression treatment

WTP (in time) for meditation treatment

WTP (% income) for magnetic treatment

Happiness in last 24 hours 1WTP (% income) for depression treatment 0.1466 1

WTP (in time) for meditation treatment

0.1892 0.1446 1WTP (% income) for magnetic treatment

0.0625 0.5849 0.1980 1

Page 11: Happiness and Utility: Evidence from New Surveys Daniel J. Benjamin Ori Heffetz Miles Kimball Alex Rees-Jones 1

WTP for Felt Happiness Predicts Current Felt Happiness

11

Felt happiness in last 24 hours

WTP (% income) for depression treatment

WTP (in time) for meditation treatment

WTP (% income) for magnetic treatment

Happiness in last 24 hours 1WTP (% income) for depression treatment 0.1466 1

WTP (in time) for meditation treatment

0.1892 0.1446 1WTP (% income) for magnetic treatment

0.0625 0.5849 0.1980 1

N = 412

Page 12: Happiness and Utility: Evidence from New Surveys Daniel J. Benjamin Ori Heffetz Miles Kimball Alex Rees-Jones 1

Planned Studies• Between-subjects design

– Currently underway.• Nationally-representative sample (N = 1,000)

– Current felt happiness, sleep scenario, and WTP for felt happiness on 2009 Cornell National Social Survey (CNSS).

• What else matters besides happiness?– Measure other attributes of options, such as

health and autonomy.– Estimate utility function with these arguments and

compare their importance with happiness.12