Research That Matters (January 17 - 20, 2008)

Purpose: Based on a state/culture theoretical frame (Steinmetz, 1999), the dialectical relationship between state culture and political context, as expressed by publicly funded health policy formation, is studied in this paper. In order to simultaneously examine direct, indirect, and total effects of state level cultural, social, and political factors on Medicaid program funding, a cross-sectional study of the 50 states in 2004 was developed.

Method: State level public support for indigent basic health care was measured using total Medicaid expenditures as a dependent variable. State level cultural, social, and political predictor variables were measured using political affiliation of major political actors, progressiveness of tax structure, and political willingness as expressed by more or less provision of Medicaid optional services and eligibility. In addition, influence of state level (a) health care interest groups, (b) financial capacity, and (c) Medicaid need demographics were included in the research model. A path model as a special case of a structural model was developed using created composite variables from multiple-predictor variables, and tested with EQS 6.0. Using SEM gave this study an ability to evaluate goodness-of-fit. Since chi-square test is directly influenced by sample size, which in this study also comprised the total population of 50 states, and as there is no way to increase sample size for this study, CFI, RMSEA, and 90% of confidence interval of RMSEA were evaluated instead of chi-square test result. Post-hoc modifications were incorporated for research models that were inadequately specified. Based on LM test, only theoretically meaningful paths were added. In order to yield a parsimonious model, Wald test was conducted.

Findings: In predicting total Medicaid expenditures (R2 = .47), Medicaid need had the largest impact (-.41); followed by political characteristics (.33); and financial capacity (.26). In predicting total Medicaid expenditures, there were two mediating variables: (a) political willingness acted as a mediating variable for the impact of political characteristics (.34), health care interest group influence (-.22), and Medicaid needs (-.40), and (b) progressiveness of taxation acted as a mediating variable for the impact of political characteristics (.16). The direct effects on total Medicaid expenditures for (a) political characteristics was .26, (b) Medicaid needs was -.36, and (c) financial capacity was .26. The direct effects of political willingness and progressiveness of taxation were not significant.

Implications: State level cultural, social, and political factors influencing health care policy formation for the poor are described in this paper. Social welfare policy knowledge in the context of health care policy formation is required for social workers to develop strong public advocacy skills. Such skills are needed when attempting to most effectively and efficiently utilize scarce social work resources in the complex and rapidly changing political context of social welfare policy making today. These findings inform strategic efforts by social workers to effect positive changes in social welfare policy. Creative methodologies for conducting secondary data analysis in social work policy research are discussed. The circumstances under which a structural model for path analysis with SEM is appropriately utilized is discussed.