Bridging Disciplinary Boundaries (January 11 - 14, 2007)

Because social welfare policymaking continues to devolve to the state level, social workers need to learn how to most efficiently and effectively influence this process. The purpose of this study was to determine: (a) whether a 50-state cross-sectional analysis of state level political party voting behavior significantly predicts state level Medicaid program expenditures, and (b) if statistical significance were found, whether this finding was substantively significant. Medicaid is a redistributive program which transfers resources or goods from one group to another group that cannot pay for their basic health care because of poverty, old-age, and/or disability. Medicaid expenditures are a product of a political process, which requires bargaining on the part of a variety of entities. Key players in the state level political process include a state's governor, senate, and assembly. Each of these key players is usually affiliated with a political party. Most previous research has grouped together the two legislative branches as one variable. This study operationalizes state actor political affiliation into three categories: a state's governorship, senate, and assembly in order to test political party influences on state Medicaid expenditures. These three variables indirectly measure a state citizenry's political affiliation and political ideology to support Medicaid, as it is expressed in voting patterns for governor, state's senators and representatives. These political actors then directly influence the Medicaid budget process. However, state voters choose their government officials based not only on political preferences, but also on other unique local factors. To address these and other construct validity threats, this study includes a variable that attempts to identify how a state's political characteristics are related to public support of health care for the poor as it is expressed by a state's popular vote for the national Presidency. States were categorized into "regions" based on previous research examining regional effects on welfare policies. This study hypothesized that state level political parties significantly differ in their commitment to support Medicaid expenditures. A 50-state analysis of covariance (ANCOVA) was conducted to predict state level Medicaid expenditures by political party influence while holding a state's financial and demographic influences constant. Results from ANCOVA indicated significant interactions (a) between political affiliation of governorship and majority party affiliation of Assembly, and (b) between political affiliation of governorship and a state's Presidential election voting pattern in predicting a state's Medicaid expenditures. Follow-up tests were conducted to evaluate significant simple effects. Findings after consideration of 'substantive significance' were: (a) states where Democratic governors worked with Democratic Party controlled Assemblies are more likely to have higher Medicaid expenditures; (b) when controlling for financial and demographic influences, significant differences were observed revealing a "regionalism" effect in support for Medicaid expenditures. Observed regional and political party influences on support for healthcare for the poor indicate fundamental state and regional political and cultural differences that need to be considered by social workers advocating for vulnerable populations dependent on state level welfare programs. These findings inform strategic efforts by social workers to effect positive changes in social welfare policy.