Article Studying legislative party politics in data scarce environments: A new empirical approach Vineeta Yadav Pennsylvania State University, USA Abstract Efforts to study the dynamics of legislative and party politics in developing countries often confront a serious obstacle – lack of public data on legislative actions. This paper proposes a new feasible approach for obtaining legislative information indirectly – analyzing the perceptions and lobbying behaviors of business interest groups who are experienced, knowledgeable and highly invested in staying informed on policymaking processes. I build on extant literatures on political parties and lobbying to show how we can systematically exploit group insights to study legislative parties in data-scarce environments and complement existing approaches in data-rich cases. I then demonstrate and evaluate this approach by using original data from a 2006 survey of 158 groups in a data-rich case, Brazil, to mediate existing scholarly debates regarding the sources of legislative unity of Brazilian parties. Keywords agenda setting, amendments, Brazil, developing democracies, interest groups, legislative institutions, legislative party unity, legislative voting, party strength, policymaking Party unity in the legislature increases party ability to shape policy, build a brand and win votes and thus to deliver on pro- mises to party members and voters (for example see Bowler et al., 1999; Carey, 2008; Cox, 2000; Heller and Mershon, 2009; Laver and Benoit, 2003; Martin and Vanberg, 2011; Samuels and Shugart, 2010). Understanding what strengthens or undermines legislative party unity has therefore been important for understanding policymaking and the formation and survival of governments and coalitions. Yet, we still lack an empirically-based understanding of the relative impor- tance of various determinants of party unity in many develop- ing country democracies, including for example India, Indonesia and South Africa, due to their lack of data on basic legislative actions such as voting and the sponsorship of bills and amendments. This prevents us from applying empirical tools commonly used in US and Western European congres- sional studies to these developing countries. Even in relatively data-rich developing countries, for example Brazil, the limited number of roll-calls limits the ability to use voting data to distinguish between alternative ex-ante legislative bargaining processes leaving important questions regarding legislative party unity unresolved. In this paper, I discuss and demonstrate how we can use a pre- viously unexploited, rich source of political insight – the perspective of informed, engaged business interest groups (BIGs) with considerable knowledge of and experience with the legislative policy process – to help bridge this data gap and move this research agenda forward. An extensive literature on interest groups argues that BIGs identify which policymakers to lobby based on the ideologi- cal preferences of various policymakers, their relative legisla- tive policymaking influence and their relationships with the executive. 1 A substantial literature on political parties and legislative politics argues these three factors influence party unity as well (see Bowler et al., 1999; Carey, 2008; Cox and McCubbins, 2005; Samuels and Shugart, 2010). I exploit this Paper submitted 20 November 2014; accepted for publication 21 November 2014 Corresponding author: Vineeta Yadav, Department of Political Science, Pennsylvania State University, 331 Pond Lab, University Park, PA 16803, USA. Email: vyadav@psu.edu Party Politics 1–13 ª The Author(s) 2015 Reprints and permission: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/1354068814564815 ppq.sagepub.com at PENNSYLVANIA STATE UNIV on February 6, 2015 ppq.sagepub.com Downloaded from