Gibson and Seventer: A Tale of Two Models

Gibson, B. & Van Seventer, D., 2000. A Tale of Two Models: comparing structuralist and neoclassical computable general equilibrium models for South Africa. International Review of Applied Economics, 14(2), 149-171.

“This paper compares two working models of the South African economy, an orthodox, neoclassical computable general equilibrium model in which savings drive investment, and a more structuralist, eclectic, model for which there is an independent investment function” (149).

“It is seen that the neoclassical model fully supports the principles of the `Washington Consensus’ while the structuralist model requires a far more heterodox set of policies to avoid slow growth or high inflation” (149).

They calibrate both models to the same SAM. The regional focus is South Africa.

They make a distinction between structuralist and neoclassical models. In neoclassical, or orthodox models, government spending is always a problem at the macro level: there is a clear inverse relationship between government spending and economic growth. Government spending can lead to, “…current account deficits, real exchange rate appreciation and an accelerated decline in export performance” (150).

Structuralist models make different assumptions, and do not rely solely on state-based explanations. “It is not typically assumed that resources are fully utilized in structuralist models, thereby opening a range of demand-side, employment generating policy options, options that have little relevance in the orthodox setting. “What makes a structuralist model structuralist is the specific and path-dependent character of the economy under study” (150).

Orthodox modeling conceptions are derived from Dvarajan and Lewis (1990): “There are two sectors, traded and non-traded and three goods, counting imports. The exportable is not consumed at home and the home good is not exported” (151). “A second important assumption is that there is only one race and one class of consumers in the prototype version, even though there are 13 occupational categories, six income classes and four races in the base SAM” (151). “From the orthodox perspective, this model has many desirable properties.

The structuralist model takes time into account. In this way, it is dynamic and more useful than the static neoclassical model.

The authors find that the neoclassical model fits nicely with “Washington Consensus” conceptions of trade openness and taxation. The structuralist model is much more vague in its conclusions.