How big is your neighborhood? Spatial implications of market participation by smallholder livestock producers

Identifying ways to increase market participation by smallholder producers requires identifying variables that influence market access. This is usually achieved using probit estimation. An important phenomenon affecting entry decision-making is the entry decision of a \91similar\92 household, where similarity is measured in terms of \91location.\92 When neighborhood influences are significant, it is important to allow for them in discrete decision contexts, such as probit estimation. This paper, therefore, assesses the magnitude of neighborhood influences in smallholder decisions concerning market entry. The empirical model is based on a cross-section of (110) farms situated in northern Philippines, visited (twice) in the 2000-2001 production year (a panel of 220 observations). The vehicle for analysis is a Bayesian formulation of a standard probit model, but one that allows for spatial autoregression in the decision vector. Estimation requires a Metropolis-step addition to a basic Gibbs sampling algorithm and generates useful insights concerning quantities that are important for market-access policy.