Do State SME Programs Work ? Evidence from Loan Subsidies and Credit Guarantees in Kazakhstan

This paper evaluates the employment and sales effects of two widely used financial support instruments for small and medium-sized enterprises, interest rate subsidies and credit guarantees, using administrative program data from Kazakhstan matched to the universe of firms. Utilizing staggered intervention rollouts and a difference-in-differences design, the analysis reveals significant differences across program designs and local labor market conditions. Interest rate subsidies, despite their large fiscal costs, fail to improve firm performance: beneficiary firms experience a 10 percent decline in employment and no significant increase in sales. Fully subsidized credit guarantees show no discernible effects on sales or employment. By contrast, a market-aligned, fee-based partial credit guarantee that ensures lender and borrower risk-sharing increases employment by 24 percent and sales by 21 percent, with particularly stronger effects among women-led and formally incorporated businesses. These employment gains are substantially larger in regions with higher pre-program unemployment, suggesting that well-designed credit guarantees are more likely to generate net job creation in labor markets with greater slack, rather than merely reallocating workers across firms. Overall, the findings underscore the pivotal role of incentive-compatible program design and local labor market condit ions in determining the effectiveness of financial policies for small and medium-sized enterprises.