Methods: Probability-based survey data from two Central Asian countries, Kazakhstan (N=1,505) and Kyrgyzstan (N=1,500) is used. Bivariate analysis is utilized to compare groups supporting and opposing government programs for families with children and the working poor by sociodemographic characteristics in each country. In multivariate models, the relationships between support for government programs and government performance indicators are examined using a logistic regression.
Results: Results indicate that trust for government is related to higher odds of supporting welfare programs for the poor and redistribution through higher taxes in both countries: trust in the government in Kazakhstan (OR=1.68) and trust in the president in Kyrgyzstan (OR=2.07) are associated with increased support for programs for the poor. In both countries, trust in the president is associated with increased willingness to pay more taxes (Kazakhstan with OR=1.51 and Kyrgyzstan with OR=2.08, respectively). In Kazakhstan, a perception of corruption as a priority governance issue is associated with increased willingness to pay more taxes to support the needy (OR=1.39). Results also find that personal experience of unofficial payments is related to lower odds of supporting programs for families with children and the poor. Making any unofficial payments in the last year is associated with reduced support for programs for families by 43% in Kazakhstan and 52% in Kyrgyzstan. In Kyrgyzstan, making unofficial payments show reduced support for programs for the poor by 49%.
Implications: The results indicate that citizens' trust in government and experiences of corruption are significant predictors of their support for welfare programs for the needy. Our results show a corrosive influence of corruption on support for social programs and the future development of welfare states in Central Asia. Addressing corruption in public sector will be important for maintaining public support for welfare programs.