Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
This paper makes the point that the choice of solution technique for nonlinear equation systems is a matter of trading the potentially smaller number of steps to convergence of formal Newton methods against the substantially smaller computational burden per step offered by simple first-order iterations such as the Gauss-Seidel method. Experiments with six typical macroeconomic models show that tradeoff to be sharply in favor of the latter. Moreover, reordering algorithms reduce all these models to near-recursive structures with relatively few feedback variables. This property is shown to be a natural consequence of the typical structure of an economic model. Copyright 1990 by Blackwell Publishing Ltd