Using survey data from more than 2400 members of the lower houses of 49 state legislatures, this article examines the influence of a set of variables on the amount of time that state legislators spent on the service activities of casework and public resource allocations ("pork"). This research finds that state legislators distinguish between casework and pork activities, rank casework ahead of pork, and that the only common variable predicting both types of service activities is members' perceptions of their constituents' preferences; otherwise, a different set of variables predicts each type of service activity.
State legislators perform the representational activities of lawmaking, service activities, and communicating with constituents (Eulau and Karps 1978: 55-74; Jewell 1982: 18-19). Of these, lawmaking is most important (Frantzich 1979; Patterson 1996; A. Rosenthal 1981: 70; Whistler and Ellickson 1999). However, service activities have become increasingly important; in the early 1950s, one-fourth of state legislators called service activities important (Wahlke et al. 1962: 304), by the 1990s-depending on the state-upwards of two-thirds of state legislators express the belief that constituency services are important (A. Rosenthal 1998: 16).
Legislative service activities are of two types: casework and public resource allocations (pork). The latter consist of revenues for projects/programs collected from the public treasury and spent in a legislator's district for the benefit of the district (e.g., roads, traffic signs, universities expanded, parks, prisons, etc.) (Jewell 1982: 137-40); the former are specific actions taken by a legislator on behalf of a constituent or group of constituents (e.g., helping with eligibility for benefits, public jobs, government regulations, and even marital problems) (151 59). In principle, the difference is a public good provided for whoever benefits (pork) versus a specific benefit for a particular person or set of persons (casework); but in practice it is often difficult to determine the difference (135).
A recent review of state legislative research (Moncrief, Thompson, and Cassie 1996: 316-17) concluded that studies explaining variation in constituency service, both casework and resource allocation, have remained curiously absent since the topic was first broached by Jewell (1982) in the early 1980s. With one exception (Whistler and Ellickson 1999), studies have been limited to a small number of states and to investigating only the casework component (Freeman and Richardson 1994, 1996; Nelson 1991; Thomas 1992) or the pork component (Thompson 1986). In this work, we focus on both types of service activities (casework and pork) across the states.
State legislators vary as to the amount of constituency service they perform (Freeman and Richardson 1996; Jewell 1982: ch. 6). There is, however, no research that investigates whether state legislators perceive a difference between the two types of constituency services (casework versus public resource allocation) and analyzes an explanatory model for each type. In the present research, we examine whether state legislators perceive a difference between casework and allocation of public resources (pork), and how much time they spend on each. We employ regression analysis to test the relative influence of variables that research indicates may explain the amount of time that legislators spend on casework and on public resource allocations. The data are from a survey of over 2,400 members of the lower houses in 49 state legislatures.'
LITERATURE AND RESEARCH EXPECTATIONS
For both conceptual and organizational reasons, we group into four categories the variables that research indicates may explain legislators' constituency service. These four categories are: individual legislators' characteristics, district circumstances, institutional features within state legislatures, and various statelevel variables. …