Degraded and infertile soil, low agricultural productivity, and food and nutrition insecurity are persistent and major challenges facing many countries in sub-Saharan Africa (SSA) up to this day. Agricultural sustainable intensification (SI) has been proposed as a possible solution to simultaneously address these challenges. Yet, there is little empirical evidence on whether SI indeed improves households' incomes, nutrition, and food security. The three essays in this dissertation take various quasi-experimental approaches to investigate child nutrition and household food security effects of SI and examine the role of input subsidies in promoting SI using nationally-representative household panel survey data from Tanzania. In the empirical analysis, I focus on three important soil fertility management (SFM) practices in Tanzanian maize-based production systems: the use of inorganic fertilizer, the use of organic fertilizer, and maize-legume intercropping. I group the eight possible combinations of these technologies into four SI categories: i) "Non-adoption" (use of none of the practices), ii) "Intensification" (use of inorganic fertilizer only), iii) "Sustainable" (use of organic fertilizer, maize-legume intercropping, or both), and iv) "SI" (joint use of inorganic fertilizer with organic fertilizer and/or maize-legume intercropping). This categorization is used in all three essays. In essay 1, results from a multinomial endogenous treatment effects model suggest that the use of practices in the "SI" category is consistently associated with improvements in children's height-for-age z-score and weight-for-age z-score, particularly for children beyond breastfeeding age (i.e., those age 25-59 months). I also find evidence that these effects come through both productivity and income pathways, and that the combined use of inorganic fertilizer and maize-legume intercropping is a key driver of these effects on child nutrition.Essay 2 investigates the extent to which the use of practices in each SI category influences household net crop income (per acre and per adult equivalent) and crop productivity as well as household food access (modified household dietary diversity score (HDDS), food expenditure per adult equivalent, and food consumption score (FCS)). Results from a multinomial endogenous switching regression model suggest that relative to "Non-adoption", use of practices in each of the other SI categories has a positive and significant effect on a household's net crop income-related outcomes and crop productivity. Importantly, for these outcomes, the "SI" category has either larger or similar-in-magnitude effects compared to "Intensification", and consistently larger effects than "Sustainable" practices. The results further suggest that a household's use of packages in the "SI" category is significantly associated with increases in all three food access outcomes, with the size of these effects similar to or greater than those of "Sustainable" practices and consistently larger than the effects of "Intensification". Essay 3 explores whether Tanzania's input subsidy program (ISP) from 2008 to 2014, the National Agricultural Input Voucher Scheme (NAIVS), encouraged or discouraged farmers' use of practices in the various SI categories on their maize plots using a multinomial logit model combined with the control function approach. I find statistically significant positive effects of household receipt of a NAIVS voucher for inorganic fertilizer on maize-growing households' use of inorganic fertilizer only (i.e., "Intensification") and on their combined use of inorganic fertilizer with organic fertilizer and/or maize-legume intercropping (i.e., "SI"). On the other hand, no such effects are found for the "Sustainable" category.
Read
