Skip to main content icon/video/no-internet

Introduction

Utility is always subjective and refers to the amount of personal or institutional satisfaction of an alternative. The personal satisfaction associated with the outcome of good food, a holiday trip or a sports car depends on taste, interests and needs. Institutional satisfaction associated with the outcome of employment strategies or treatment selection predominantly depends on economic considerations. It is assumed that both individuals and institutions try to maximize their utility. Choosing alternatives according to their utility lies at the heart of all summative evaluation procedures.

Prescriptive Utility Theory

Utility Function

The measurement of utility has a long history in economics, starting with the measurement of preferences on an ordinal scale. Since von Neumann and Morgenstern (1947) developed the axiomatic foundation of utility, it can be measured on an interval scale. It is derived from personal preferences for either a sure thing or a gamble (Figure 1a). The decision maker is presented with a choice between obtaining either outcome o for sure or a gamble that returns with probability p a better outcome (o∗) and with complementary probability (1-p) a worse outcome (o∗). The gamble is written as [p o∗, (1-p) o∗] or simply [p o∗, o∗] (the square denotes a decision, the circle a chance node).

With o∗ and o∗ as best and worst outcomes and all outcomes oi in-between, the indifference probability πi, where the decision maker is indifferent between the sure thing and the gamble, can be determined as oi ∼ [πi o∗, o∗]. Pairs (o∗, πi) of these indifferences constitute the utility function (Figure 2a). Above the function the gamble is preferred, below the function the sure thing is preferred. For outcomes with the following preference order o∗ ϕo1 ϕo2ϕ … oi …ϕo∗ the corresponding indifference probabilities with 1.0 = π∗ > p > p >… >p, >… > π∗ = 0 (as well as their linear transformations) are the utilities of the corresponding outcomes. Whenever the preference oi ϕoj holds, the utility measurement of oi exceeds oj and πi > πj.

Axiomatic Foundation

The perceived utility of outcomes is the basis for expressing preferences. These preferences are analysed in order to measure true internal utilities. If the expressed preferences with respect to outcomes of alternatives (objects, commodities, events, strategies) and gambles meet the following six axioms, utility can be measured on an interval scale.

Ordering

A decision maker should be able to compare outcomes and either prefer one to the other or be indifferent.

Transitivity

If one outcome is preferred to a second, and this second is preferred to a third, the first should also be preferred to the third.

Dominance

A rational decision maker should never accept a dominated gamble whose best outcome matches the outcome of the sure thing and should always accept a gamble whose worst outcome matches the outcome of the sure thing.

None

Figure 1. The choice between a sure thing and a gamble.

None

Figure 2. Utility functions.

Cancellation

Preferences for gambles must not depend on identical and equally probable outcomes.

Invariance

As long as the probabilities with which certain outcomes are obtained are the same, the form in which gambles are presented must not affect preferences.

...

  • Loading...
locked icon

Sign in to access this content

Get a 30 day FREE TRIAL

  • Watch videos from a variety of sources bringing classroom topics to life
  • Read modern, diverse business cases
  • Explore hundreds of books and reference titles

Sage Recommends

We found other relevant content for you on other Sage platforms.

Loading