Background The economic impact of schistosomiasis and the underlying tradeoffs between water resources development and public health concerns have yet to be quantified. Schistosomiasis exerts large health, social and financial burdens on infected individuals and households. While irrigation schemes are one of the most important policy responses designed to reduce poverty, particularly in sub-Saharan Africa, they facilitate the propagation of schistosomiasis and other diseases. Methods We estimate the economic impact of schistosomiasis in Burkina Faso via its effect on agricultural production. We create an original dataset that combines detailed household and agricultural surveys with high-resolution geo-statistical disease maps. We develop new methods that use the densities of the intermediate host snails of schistosomiasis as instrumental variables together with panel, spatial and machine learning techniques. Results We estimate that the elimination of schistosomiasis in Burkina Faso would increase average crop yields by around 7%, rising to 32% for high infection clusters. Keeping schistosomiasis unchecked, in turn, would correspond to a loss of gross domestic product of approximately 0.8%. We identify the disease burden as a shock to the agricultural productivity of farmers. The poorest households engaged in subsistence agriculture bear a far heavier disease burden than their wealthier counterparts, experiencing an average yield loss due to schistosomiasis of between 32 and 45%. We show that the returns to water resources development are substantially reduced once its health effects are taken into account: villages in proximity of large-scale dams suffer an average yield loss of around 20%, and this burden decreases as distance between dams and villages increases. Conclusions This study provides a rigorous estimation of how schistosomiasis affects agricultural production and how it is both a driver and a consequence of poverty. It further quantifies the tradeoff between the economics of water infrastructures and their impact on public health. Although we focus on Burkina Faso, our approach can be applied to any country in which schistosomiasis is endemic. Graphical Abstract

We extend the celebrated Rothschild and Stiglitz (1970) definition of Mean-Preserving Spreads to a dynamic framework. We adapt the original integral conditions to transition probability densities, and give sufficient conditions for their satisfaction. We then prove that a specific nonlinear scalar diffusion process, superdiffusive ballistic noise, is the unique process that satisfies the integral conditions among a broad class of processes. This process can be generated by a random superposition of linear Markov processes with constant drifts. This exceptionally simple representation enables us to systematically revisit, by means of the properties of Dynamic Mean-Preserving Spreads, four workhorse economic models originally based on White Gaussian Noise.

No abstract

We develop a target zone model with realistic features such as finite exit time, nonstationary dynamics and heavy tails. Our rigorous characterization of risk corresponds to the dynamic counterpart of a mean-preserving spread. We explicitly solve for both stationary and transient exchange rate paths, and show how they are influenced by the distance to both the time horizon and the target zone bands. This enables us to show how central bank intervention is endogenous to both the distance of the fundamental to the band and the underlying risk. We discuss how the credibility of the target zone is shaped by the set horizon and the degree of underlying risk, and we determine a minimum time at which the required parity can be reached. We prove that the interplay of the diffusive component and the destabilizing risk component can yield an endogenous regime shift characterized by a threshold level of risk above which the target zone ceases to exist. All the previous results cannot obtain by means of the standard Gaussian and affine models. We recover by numerical simulations the different exchange rate densities established by the target zone literature.

We present a class of scalar Markovian diffusion processes whose transition probability densities are skewed Gaussian distributions. Their stochastic dynamics involve nonlinear and time-dependent drifts driven by White Gaussian noise sources. The drifts are obtained via generalized h-transforms of a class of local martingales. We can alternatively represent this class of processes as dynamic censoring models with partial observability and time-dependent correlations, which can be used as a skewnessinducing noise source for any diffusion process. We prove the invariance of our class of skew-Normal processes under linear transformations. We extend our results to Ornstein-Uhlenbeck diffusions.