Monday, March 15, 2010

Tomorrow's Seminar: Uwe Sunde

March 16th:  1pm: Geary Seminar Room 

Uwe Sunde (University of St. Gallen). 

"On the stability of risk and time preferences"


Martin Ryan said...

Hat tip to Mark McGovern for forwarding on a recent IZA paper by Meier and Sprenger (2010) which shows that time preference is not a perfectly stable characteristic.

According to Borghans,
Duckworth, Heckman and ter Weel (2008), less is known about the stability of `preferences in
economic decision-making' compared to the stability of personality traits. Frederick et al. (2002)
state that no longitudinal study has measured the mean-level stability of time preference over the life cycle. Borghans and Golsteyn (2008) is the first study to report that time preference has
a U-shaped relation with age: adolescents have very high time preferences; then at 36 years of
age time preference reaches its minimum, after which it increases again.

Risk aversion increases
dramatically after adolescence but remains relatively stable thereafter (Borghans and Golsteyn,
2008).There is a peak in sensation-seeking (which is linked to
risk-attitudes) during adolescence (Spear, 2000). Sahm (2007) shows that risk aversion is not a
perfectly stable characteristic over the lifecycle, and that it increases with age.

Kevin Denny said...

Would you not expect discount rates to rise as you age since "you can't take it with you".
A common marker that has been used for high discount rates is smoking:
high discount rate=>smoking=>die young
But it could be the other way round:
smoking=>die young=>high discount rate

Martin Ryan said...


That's an interesting hypothesis; nothwithstanding the possible incentive to leave a behind a bequest for one's children.

Borghans and Golsteyn (2008) report that time preference reaches its minimum at 36 years of age, after which it increases again.

Of course, formally (though hard to disentangle in practice), time preference is only one component of the discount rate (as described by Frederick, Lowenstein and O'Donoghue; 2002); the other being risk-preference.

According to Borghans and Golsteyn (2008), risk aversion increases
dramatically after adolescence but remains relatively stable thereafter.

So the literature is pointing towards a stable risk-preference after adolescence, and increasing present-orientation after 36 years of age. This is not necessarily at odds with your hypothesis; retirement is another life-cycle development that could have a major impact on discounting: "I now have all this spare time, and I've never been closer to kicking it!"

P.S. In the discussion after today's seminar, Uwe mentioned that he hasn't found much to back up smoking as a marker of time-preference.