What are the Effects of Input Subsidies on Maize Prices? Evidence from Malawi and Zambia

Millions of smallholder farm households in Sub-Saharan Africa (SSA) are net consumers of staple crops, and millions of poor urban households spend a significant share of their income purchasing staple foods. Recent research has underscored the major effects of changes in food prices on poverty, with the weight of the evidence indicating that rising food prices exacerbate poverty and food insecurity. Large-scale input subsidy programs, particularly for maize, have grown in popularity in SSA over the last decade. An important hypothesized but heretofore empirically untested benefit of these programs is that by raising maize production, the subsidies should put downward pressure on retail maize prices to the benefit of urban consumers and the rural poor, who tend to be net buyers of maize. To inform debates related to this rationale for input subsidies, this study estimates the effects of fertilizer subsidies on retail maize prices in Malawi and Zambia using market or districtlevel panel data covering the 2000/01 to 2011/12 maize marketing years. Malawi and Zambia are ideal case studies because both countries have well-known, large-scale fertilizer subsidy programs where the quantities distributed vary spatially and over time. In addition, the scale of the subsidy programs was large enough in both countries to have substantially affected national maize production, and hence have potentially discernible effects on domestic food prices. The effects of fertilizer subsidies on equilibrium retail maize prices in Malawi and Zambia are estimated via country-specific reduced form panel data econometric models of retail maize prices as a function of subsidized fertilizer and other factors. The models are estimated via first-differencing or the Arellano-Bond (AB) dynamic panel data method. Both estimators control for time constant unobserved effects. The major advantage of the AB approach is that it allows for lagged retail maize prices to affect current retail maize prices. The findings from our study are similar between Malawi and Zambia. They indicate that fertilizer subsidies have either no statistically significant effect on retail maize prices or, more commonly, a statistically significant but very small negative effect on those prices. The results suggest that roughly doubling the size of Malawi’s subsidy program (i.e., increasing the amount of subsidized fertilizer distributed to each district by 4,000 metric tons (MT) per year) only reduces real maize prices by 1.2% to 1.6% on average. In Zambia, roughly doubling the scale of the country’s subsidy program (i.e., by increasing the amount of subsidized fertilizer distributed to each district by 1,000 MT per year) only reduces real maize prices by 1.8% and 2.4% on average. The results are statistically significant at the 10% level or lower for most of the models estimated. It should be noted that even small decreases in maize prices would benefit the many poor rural and urban households that are net buyers of maize. However, empirical evidence presented here does not support the often-asserted claim that large public expenditures on input subsidies have major poverty reducing effects because the programs produce large spillover benefits in the form of substantially lower maize prices. The empirical evidence to date suggests that even the large-scale fertilizer subsidy programs in Sub-Saharan Africa may result in very small, if any, reductions in retail food prices in semi-open economies