Special issue in honor of Harris Schlesinger: New developments in the study of risk preferences
Following a long illness, Harris Schlesinger died on September
Special issue in honor of Harris Schlesinger: New developments in the study of risk preferences
Cary Deck 0 1 2
Sebastian Ebert 0 1 2
Andreas Richter 0 1 2
Sebastian Ebert 0 1 2
Andreas Richter 0 1 2
0 Institut für Risikomanagement und Versicherung, LMU München , Schackstraße 4/III, Zi. 305, 80539 München , Germany
1 Frankfurt School of Finance and Management , Adickesallee 32-34, 60322 Frankfurt am Main , Germany
2 Culverhouse College of Commerce, University of Alabama , 361 Stadium Dr., Tuscaloosa, AL 35487 , USA
-
European Group of Risk and Insurance Economists (EGRIE) and was president in
2006–2007. Harris also served as President of the American Risk and Insurance
Association (1997–1998), a member of its Board of Directors (1991–1998), and
held editorial positions on its flagship journal, the Journal of Risk and Insurance.
Given the importance of Harris Schlesinger’s contribution to the risk and insurance
economics community, it was decided that both the Geneva Risk and Insurance
Review and the Journal of Risk and Insurance would jointly dedicate special issues
in his honor in appreciation of Harris’s impact on the academic community.
This special issue of the Geneva Risk and Insurance Review in honor of
Harris Schlesinger features five articles. A major theme in Harris’s work is behavior
under risk within the expected utility model. In the first article of this issue,
Gollier
and Kimball (2018)
provide several fundamental results on this matter. The authors
study two general types of risk comparisons, which are referred to as being
“contingent” and “non-contingent” (on a reference lottery), respectively. The authors
illustrate the economic relevance of each risk comparison, clarify their relationship, and
propose a general solution concept to problems involving these comparisons using
the “basis” approach. While the basis approach has often been used implicitly, the
current paper makes a methodological contribution by making the basis approach
itself as well as the range of its application explicit.
Ambiguity is a condition where the probabilities of possible events are not
uniquely assigned. Harris contributed to our understanding of the effects of
ambiguity on optimal behavior
(Osaki and Schlesinger 2014a, b)
and of its experimental
measurement including higher-order ambiguity preferences
(Baillon et al. 2018)
.
When decision-makers are averse to ambiguity, they value information that reduces
it
(Snow 2010)
, although the process of acquiring information may actually increase
ambiguity, for example, due to the uncertainty of which (genetic) test results will be
obtained
(Hoy et al. 2014)
.
Nocetti (2018)
presents a model that facilitates the
distinction between posterior uncertainty over the correct probability distribution and
uncertainty over the message that will be received from a given message service.
This allows to integrate the existing results by identifying attitudes towards both
sources of ambiguity separately. In such a more general model, the comparative
statics of the value of information under ambiguity resemble the comparative statics of
the value of information under risk to a much larger degree.
“A well-known result in the insurance literature states that a risk averter who is
offered insurance at a ‘fair’ price will choose to fully insure.” Doherty and
Schlesinger (1990, p. 243). But
Schlesinger and Schulenburg (1987)
and
Doherty and
Schlesinger (1990)
show that this holds only when there is no risk of default by the
insurer. In those papers, Harris and his respective coauthors assume that the
individual’s behavior is characterized by expected utility theory. Harrison and Ng (2018)
use experiments to examine this setting. They find that allowing for non-expected
utility preferences by participants suggests that welfare loss is not as significant as
when all participants are assumed to have expected utility preferences.
Courbage et al. (2018
) contribute an article on higher-order risk preferences,
an area of research that was significantly advanced by Harris. The authors derive
new properties of the Nth-order utility premium that was also examined in one of
Harris’s most recent articles,
Ebert et al. (2017
). The Nth-order utility premium
measures the pain associated with facing the passage of one risk to a more severe
risk. The authors analyze the superadditivity of the Nth-order utility premium and
relate it to concepts such as mixed risk aversion
(Caballé and Pomansky 1996)
, a
preference for combining good with bad
(Eeckhoudt et al. 2009)
, or risk
apportionment
(Eeckhoudt and Schlesinger 2006)
.
In the final article of this issue,
Bostian and Heinzel (2018)
also investigate
higher-order risk effects. But, unlike most articles on higher-order risk
preferences, the authors study them in a recursive utility framework rather than in the
classical expected utility model. This (...truncated)