The impacts of cash transfer programs on rural livelihoods: a study of Caboclos in the Brazilian Amazon estuary region
Rural households that rely on agricultural and natural resources for their livelihoods have been exposed to increasing socio-economic and climatic challenges over the past few decades, which requires urgent scientific exploration to effectively inform policies and other interventions. This dissertation investigates the rural livelihood of smallholders and the impacts of cash transfer programs through the use of empirical analysis and agent-based modelling and simulation (ABM) of the Caboclos in the Brazilian Amazon estuary region. The findings in this dissertation deepen the understanding of the livelihood dynamics of small farming households, provide insight about modelling uncertainty, and evaluate the impacts of policies and other approaches meant to alleviate poverty and enhance resilience. First, the empirical patterns of rural livelihoods, with a focus on the heterogeneous impacts from cash transfer programs, have been captured through statistical analysis of a household survey. Households were classified based on the amount of cash transfer and dependence on cash transfers to demonstrate the heterogeneity in this significant income of rural livelihoods. The results show the high level of heterogeneity among the value of cash transfers that households receive and in the households’ level of reliance on this stipend. Results also illustrate the differences among household characteristics and their significance regarding the degree of household reliance on cash transfers. Second, we constructed an ABM with an ensemble approach to represent the small farming households and simulate their livelihood outcomes with government cash transfer programs under eight experiments that differentiate main livelihood factors. The three ensemble members reflect a range of household behaviors, which include Max Profit (optimizing net economic returns), Max Leisure (pursuing optimal leisure time once subsistence is met), and Subsistence First (a strategy that maintains subsistence requirement first and then pursues market profit). Sensitivity and post-hoc analyses reveal the variability in the outcomes among three decision regimes, where the decision regime proves to be the most significant factor for livelihood outcomes at both the community level and individual level. The mere presence of cash transfers largely increases income and the equality of income distribution, of which the most drastic change occurs in the Max Leisure decision regime. However, household characteristics influence household livelihood outcomes differently within each decision regime. Third, we explored rural household livelihood and poverty dynamics using the ABM through the lens of development resilience. Various external shocks were applied to the household agents and their livelihood dynamics, particularly their resilience attribute, were analyzed. Our results first support the existence of the poverty trap and the relatively better-off zone as the “basin of attraction” that were proposed in resilience theory. Results from the simulation also indicate that external shocks, although similar in duration and magnitude, have significantly different impacts on livelihood resilience, with climate shocks being the most influential. Government cash transfer programs are more likely to be effective with a big initial capital boost, and a Subsistence First strategy, relative to Max Profit and Max Leisure strategies, is most likely to be resilient for vulnerable households, but not in households who are close to being trapped.