Four studies of economic behavior : integrating revealed and stated preferences data

This thesis estimates equivalence scales for Indonesia, proposes a new type of test of the life-cycle model, shows how important economic and demographic characteristics of Dutch households evolve over age, cohorts and time, and studies the decision problem to when and how much to work, save and consume. The global theme of this thesis is applied micro-econometrics where economic theory is combined with survey data of individuals to understand behavior and opinions. However, a large part of this thesis is devoted to a relatively new and important topic in economics: combining revealed preferences and stated preferences data in one unifying economic framework. Traditionally, economics has relied on the theory of revealed preferences to draw conclusions about individual preferences and interests. If a child is offered a choice between an apple and a chocolate bar, it reveals its preference by choosing one instead of the other. Whereas this methodology is incredibly powerful, in some cases it just does not suffice. 1. Actions may be obstructed: borrowing constraints prevent individuals to reveal their preferences for borrowing money. 2. Preferences do not predict action more generally: opinions, memories, ideas. 3. Observed actions may be consistent with many competing theories of behavior. 4. People make inconsistent choices: people smoke or drink too much, or do not going to the gym, whereas in fact they would like to cut down on bad behavior and do more sports. In these and perhaps other situations stated preferences data can be useful. By asking the right questions, economists may be able to tease out preferences for borrowing, the most appropriate theory of behavior, or whether people actually would like to go to the gym or not. This thesis contributes to the important research agenda that incorporates different kinds of data to ultimately reach a better truth. It employs direct questions on economic wellbeing of households in combination with observed patterns of demand for food and nonfood to estimate equivalence scales, a key tool for making welfare comparisons between households. Further, this thesis employs questions about changes in economic wellbeing of individuals in an empirical test of the life-cycle model. The test has two clear advantages over conventional tests based on revealed preferences data. First, the test is able to discriminate between different theories that predict the same, or almost the same, behavior. Second, the test does not lose power from the simultaneous estimation of model parameters and test statistics. Finally, this thesis offers a few leads for developing new ways of interpreting satisfaction data in economic science