In this section, we formulate a dynamic random utility model of happiness and
behavior that is an application of the dynamic framework proposed earlier.
The model is formulated in the context of the experiment described in the previous
section.
Figure 12.2 shows the dynamic modeling framework applied to the mode choice
experiment. We distinguish between two periods: pretreatment and posttreatment. In
the pretreatment period, since participants commute habitually by car and are not
making new decisions, they have a remembered utility from the car. This utility is
reflected by the pretreatment car happiness measure. In the posttreatment period,
the participants decide whether they will switch to public transportation or keep
commuting by car. The utility concept that is relevant in this decision-making period
is decision utilities of car and public transportation. These utilities are reflected by the
posttreatment car and public transportation happiness measures, respectively, and by the choice. Participants select the mode that maximizes their utility. The choice here
refers to the choice of mode on a given day in the posttreatment week.
Remembered and decision utilities are affected by explanatory variables. The
remembered utility of car is correlated with the decision utility of car and of public
transportation, and the decision utilities of car and public transportation are also
correlated. The correlations among the three happiness measures are captured
through the correlations among the utilities. Note that although the model
framework does not explicitly represent a causal relationship from remembered to
decision utility (as in the dynamic framework presented earlier) but rather shows a
correlation structure, the correlation relationship can be viewed as a special case of
the causal structure.
Structural Model
The structural model is a specification of the utility equations of car and public
transportation. For car, we specify pretreatment (remembered) and posttreatment
(decision) utilities. For public transportation, we specify only a posttreatment
(decision) utility equation as public transportation becomes relevant to the
participants only after they try it and decide if they want to switch to it. These
utilities are given by Eqs. (12.1)–(12.3).
high income dummy