It is a capital mistake to theorize before one has data.Sherlock Holmes, in “A Scandal in Bohemia”
Following Holmes's advice means plunging into a sea of data: more than 17,000 enactments and over 400 vetoes. Each veto is unique. Many have fascinating stories. How can one deal with this overwhelming complexity?
It is possible to approach a data set in the spirit of a natural historian. Exploratory data analysis becomes the social scientist's equivalent of a collecting jar, magnifying lens, and scalpel. These tools, along with some simple pretheoretical notions, allow us to search for structure in the data, to reduce the enormous complexity in hundreds of events to a few memorable, reliable patterns that capture much of the variation in the data. The goal of this chapter, in short, is to transform bewildering complexity into bewildering simplicity.
Veto bargaining is a dynamic process. You can no more study veto bargaining by counting the aggregate vetoes per time period, than you can study price bargaining by counting the customers who leave a shop without purchases. What is needed is a different kind of data, event histories, longitudinal data in which discrete events (vetoes, override attempts, and repassages) may occur repeatedly. Such data identify episodes of veto bargaining and track what happens in each episode. Constructing a set of event histories of veto bargaining means, in practice, identifying each bill in a bargaining episode and detailing its fate.
Identifying the Bills
I start with the 434 vetoes of public bills cast between the beginning of the Truman administration in 1945 and the end of the Bush administration in 1992.