Sam Bowles talk TODAY at Harvard on morals and incentives

91 views
Skip to first unread message

Aaron Swartz

unread,
Oct 19, 2010, 10:50:50 AM10/19/10
to Bowles Reading Group, Stephen R Laniel
For those in the area, Samuel Bowles is giving a talk today at Harvard.

Talk by Sam Bowles
Tuesday, October 19 from 12:00 - 1:30 p.m.
1305 William James Hall
33 Kirkland Street

"Why do economic incentives (sometimes) crowd out pro social motives?"

 Samuel Bowles, Santa Fe Institute

Psychologists have long known, and economists recently documented in
behavioral experiments the fact that explicit incentives such as

subsidies, taxes, and fines sometimes  crowd out social preferences
that internalize the effects of ones' actions on others or diminish
ethical

or intrinsic motivation to contribute to the public good. As a result
incentives may be less effective than predicted by economic models or

even have the opposite of the intended effect. Yet sometimes the very
same incentives work exactly as economic models predict (their effect
on

behavior being additive to that of social preferences). And in yet
other cases (not common), crowding in occurs; economic incentives and

pro-social preferences interacting synergistically (as complements
rather than as substitutes, in economics).

Devising adequate incentives to promote the public good requires an
understanding of why these non-additive crowding effects occur, and
how

policy design may take account of them. Evidence from behavioral
experiments indicates that crowding may take two forms: categorical
(the

effect on values depends only on the presence or absence of the
incentive) or marginal (the effect depends on the extent of the

incentive). There are also two mechanisms underlying the non-additive
effects of incentives and pro-social preferences. The incentive may

provide information or a cue to appropriate behavior in which cases
the individual's preferences are situation-dependent (state-dependent,
in

economics).  Or the incentive may alter the process by which
individuals learn new preferences which then persist across different
situations (in

which case preferences are endogenous.)  Using this 2x2 framework I
will provide evidence that incentives per se are not the proximate
cause of

the crowding phenomenon.

--

Samuel Bowles

Santa Fe Institute      Dipartimento di Economia Politica

1399 Hyde Park Rd       Universita di Siena

Santa Fe,NM,87501,USA   Piazza San Francesco,7

Phone: (505) 984 8800   Siena 53100, Italy

Fax:(505) 982 0565

http://www.santafe.edu/~bowles

Reply all
Reply to author
Forward
0 new messages