“…Using correlated random effects statistical models that account for unobservable attributes of legislators, we analyze a unique data set of 30,005 coded provisions and associated roll‐call votes in the House and Senate from 199 major laws enacted during the 93d to 111th Congresses (1973–2010). Within‐member designs such as ours are common in the contemporary literature on pork‐barrel politics (Alexander, Berry, and Howell ; Berry and Fowler ). Beyond ideology, committee membership, seniority, prior electoral margins, and other characteristics that have been found to be significant in studies of pork‐barrel politics (see Grose , 202–03), legislators represent different groups of constituents, which exert different pressures on their votes, suggesting the importance of legislator‐level designs to account for this heterogeneity.…”