MBS 93-35
Joint Receipt and Thaler's Hedonic Editing Rule
Peter C. Fishburn and R. Duncan Luce
We consider a rule of "hedonic editing" suggested by R. H. Thaler and
others to describe how people evaluate the joint receipt of two separate
quantities of a real variable x. Let U be a continuous and increasing utility
function on x. We refer to x r 0 as a gain, x s 0 as a loss, fix U(0) =
0, and denote by x&y the joint receipt of x and y. The hedonic editing
rule says that U(x&y) = max{U(x+y), U)x) + U(y)} so that U(x&y)
is the larger of the utility of the integrated sum of and y, and the sum
of the utilities of x and y considered separately. The paper explains structures
of U constrained by hedonic editing. Two main cases are analyzed. Case
(I) assumes that U is concave in gains and convex in losses. Case (II)
assumes that is concave separately in gains and in losses. Each main case
divides into six subcases according to the limiting relations among the
slopes of at q0 and ql. these partition the behavior of U in the mixed
(x>0, y<0) joint-receipt region into two subregions of integration and
segregation. The paper also axiomatizes the cases with assumptions about
(R,&,r) from which a suitable U can be constructed. Each main case
uses a few axioms satisfied by all its subcases. Special axioms are then
invoked for the different subcases.