Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
Two econometric issues arise in the structural estimation of consumer or producer demand systems in the presence of many binding non-negativity constraints. Firstly, most existing methods entail the evaluation of multivariate probability integrals. Secondly, the issue of statistical coherency must be addressed. We circumvent both of these issues using Gibbs' Sampling, along with data augmentation and rejection sampling. We illustrate our method using several simulated data sets.