THE PARTISANSHIP OF NEW MEMBERS IN THE 103RD AND 104TH HOUSES

Objective. We examine the party unity scores of new members in the 103 rd and 104th Houses to determine whether the behavior of the large fre shmen classes elected in 1992 and 1994 conforms to traditional expecta tions about new members' roll call behavior. Methods. We conduct bivar iate and multivariate tests on party support of new and returning memb ers for each party. We also investigate the effects of racially tailor ed redistricting on party support in the 103rd House. Results. New mem bers of both parties are frequently more supportive of their respectiv e party positions than are returning members. After controlling for tw o powerful electoral factors, we find that first-term Democrats have d istinctively high unity scores on party votes and key votes in the 103 rd House only. This distinctiveness is partially a function of the ext remely high levels of party unity among black and Hispanic members ele cted from new majority-minority districts. Republican newcomers are di stinctive on general party votes in both Houses and on key votes in th e 104th House. Conclusions. Even in the context of highly partisan con gresses, new members, especially those in the majority, are notably lo yal to their party.