Love And Money: Experimental Tests Of Theories Of The Household.

1
Love And Money Experimental Tests Of Theories Of
The Household.

• Ian Bateman, Tara McNally and Alistair
  MunroUniversity of East Anglia, Norwich, UK
• Royal Holloway College, University of London.
• http//personal.rhul.ac.uk/UQTE/003/Couples.htm

2
What are we testing?

• Theories of the household, notably
• Income pooling does the source of income affect
  choices?
• Pareto efficiency
• Unitary models
• (theories of risky choice)
• Altruistic behaviour

3
What do we do?

• We
• Recruit established couples
• Separate them
• Make them do choices separately
• Most of these choices involve gambles
• E.g. 20 for me versus a 60 chance of 40 for my
  partner
• We make them predict their partners choices
• Re-unite couples and face them with joint
  choices.
• Pick one of their 36 choices at random and make
  them play it out.

4
How do we run the experiment?
5
Example task test
6
Does information matter?

• Treatment 1 (n76),
• separated couples in different rooms,
• no revelation of answers to partner
• private and separate payouts
• in sealed envelopes.
• These facts pre-announced

• Treatment 2 (n40)
• Separated couple in same room
• All answers available to partner
• Payments made in front of partner
• These facts pre-announced

7
Within and between task tests for consistency
with Income Pooling
8
Income Pooling. Typically

• cannot reject income pooling for joint choice
• Pattern of deviations consistent with inequality
  aversion
• For separated choice
• Reject income pooling for separated choice
• Pattern of deviations consistent with greater
  weight placed on own payoffs.
• Information regime has little impact.

9
Pareto efficiency

• Across a wide range of tasks couples who make the
  same choice separately, deviate from that choice
  in joint deliberation
• Deviation is systematic away from risky choices.

10
Summing up

• Strong rejection of unanimous decisions
• Rejection of income pooling in separated choice
• Widespread deviation from Pareto efficiency
• Couples are more risk averse than the individual
  partners
• Results are robust to information disclosure rule
• Though some evidence that women are less risk
  averse with information disclosure
• Suggestion is that standard household models are
  inappropriate.
• More details on this sequence of experiments at
• http//personal.rhul.ac.uk/UQTE/003/Couples.htm