Comparison of Hypothesis Testing and Bayesian Model Selection

The main goal of both Bayesian model selection and classical hypotheses testing is to make inferences with respect to the state of affairs in a population of interest. The main differences between both approaches are the explicit use of prior information by Bayesians, and the explicit use of null distributions by the classicists. Formalization of prior information in prior distributions is often difficult. In this paper two practical approaches (encompassing priors and training data) to specify prior distributions will be presented. The computation of null distributions is relatively easy. However, as will be illustrated, a straightforward interpretation of the resulting p-values is not always easy. Bayesian model selection can be used to compute posterior probabilities for each of a number of competing models. This provides an alternative for the currently prevalent testing of hypotheses using p-values. Both approaches will be compared and illustrated using case studies. Each case study fits in the framework of the normal linear model, that is, analysis of variance and multiple regression.     

Improving the speed of adoption of agricultural technologies and farm performance through farmer groups: evidence from the Great Lakes region of Africa

The article examines the effect of membership in farmer groups (MFG) on adoption lag of agricultural technologies and farm performance in Burundi, the Democratic Republic of Congo and Rwanda. We use duration and stochastic production frontier models on farm household data. We find that the longer the duration of MFG, the shorter the adoption lag and much more so if combined with extension service delivery. Farmer groups function as an important mechanism for improving farm productivity through reduced technical inefficiency in input use. We discuss policy implications under which farmer groups are a useful channel to reduce adoption lag, and the means through which improved farm performance can be achieved.     

Raising offtake from cattle grazed on natural rangelands in sub-Saharan Africa: a transaction cost economics approach

Farmers in sub‐Saharan Africa are constrained by large transaction costs associated with marketing of their livestock. However, transaction costs are often not taken into account in the analysis of factors hampering the development of livestock marketing in this region. This article empirically measures the influence of transaction costs on the offtake from cattle grazed on natural rangelands in Uganda. The study is based on the monitoring of 696 cattle transactions in three districts of Uganda from August 2004 to August 2005. The estimated models suggest that proportional transaction costs represented by the state of roads, distance to markets, and time taken to reach the market are important variables constraining market participation. In order to raise offtake from the national herd, it is essential to explicitly address these costs. One potential solution is collective action in marketing of livestock in which proportional and fixed transaction costs are reduced and shared among the group members.     

Optimizing service-level and relevant cost for a stochastic multi-item cyclic production system

A multi-item make-to-order production system in a stochastic environment is analyzed. Assuming a common cycle production approach, the impact of safety stock, cycle time, demand, processing time and setup time on service-level and total relevant cost (holding, setup and backorder cost) is determined. To illustrate this relationship a trajectory for the service-level with respect to the relevant cost (holding and setup) is presented. Furthermore algorithms to calculate the cycle time which leads to maximum service-level at constant safety stock and to calculate the pair cycle time and safety stock which minimize total relevant cost are introduced.     

When do thick venture capital markets foster innovation? An evolutionary analysis

In this paper, we examine the trade off between different effects of the availability of venture capital on the speed of technological progress in an industry. We consider an evolutionary industry simulation model based on Nelson and Winter (1982), in which R&D efforts of an incumbent firm generate technological know-how embodied in key R&D employees, who might use this know-how to found a spinoff of the incumbent. Venture capital is needed to finance a spinoff, so that the expected profits from founding a spinoff depend on how easily venture capital can be acquired. Accordingly, thick venture capital markets might have two opposing effects. First, incentives of firms to invest in R&D might be reduced and, second, if spinoff formation results in technological spillovers between the parent firm and the spinoffs, the generation of spinoff firms might positively influence the future efficiency of the incumbent’s innovation efforts. We study the manner in which this tradeoff influences the effect of venture capital on innovation expenditures, speed of technological change and evolution of industry concentration in several scenarios with different industry characteristics.