Majority Leadership in the U.S. Senate: Balancing Constraints By ...

differences in activities and styles before and after. Research of interactions is the only way to “measure” leadership by interactions.

One line of inquiry of this research focuses on federal aid to states and the majority leader's particular capability to funnel funds to his state or pet projects. Data on distributive benefits, specifically earmarks in annual appropriations bills, are available for many years from the watchdog group Citizens Against Government Waste.²⁴ While readily accessible, numbers and raw dollar amounts are unhelpful in linking the majority leader to such expenditures. There are, in most cases, increases in federal outlays to state governments about the time that state gains a majority leader. However, pinpointing the hand of the majority leader in those increases is a difficult, if not impossible, task. Other factors, such as inflation, the efforts of the leader's state colleagues in the Senate or House, or even authorizations appropriated years earlier, may be the proximate cause for the increase. Therefore, reliance on archival collections is the key to this project; these archives left a paper trail of the leader's case and the project activities to follow. From these, I was able to see which projects kept the leader's attention, why they did so, how he used his influence within and without the Senate to advance the causes, and with what outcomes. The result is a broad picture of the politics of leadership, especially its demands and constraints, and what it means for representation.

Of course, archives suffer from problems with both validity and reliability. No collection is exhaustive, so that gaps in record keeping may yield a biased sample. Moreover, a researcher must take special care to collect evidence in a standardized manner, so that findings are generalizable rather than anecdotal. And yet, given the small universe of Senate majority leaders (N = 21), any study of Senate majority leadership must rely on anecdotes, and so does this book. But it does so as systematically as possible, looking at similar or dissimilar events across tenures of leaders. Variations from collection to collection inhibit, to an extent, research. Records are kept differently within each office, and one collection may be heavy on administrative files while another is replete with project files. Still, an examination of several archives, like this work undertakes, ensures that no one collection is assumed typical. That is, systemization