This study argues that structural factors – principally population per capita income and population density – can explain a significant proportion of the variation in openness defined as (exports + imports)/gross national production. It also provides a conceptual and theoretical basis for the form of the function that explains openness. Spatial theories of order in the location of economic activities – the ideas of Christaller and Losch – which have found much support in studies of geographers, and other empirical findings of spatial order viz., the famous rank size rule of cities' populations, and the equally famous 'gravity-model' of spatial interaction, can all be used fruitfully to understand openness. They explain why openness is inversely related to population sizes, and more generally the functional form of the structural dependence. Biometricians in the early part of this century had used allometric growth models to understand the form and growth of organisms, across not too distant species. Use of the allometric model to the problem of trade openness only calls for recognising the economy as having a structure and therefore as being more than a collection of producers and consumers; and simple assumptions about the economy's principal characteristics. Given this structural determination of openness we are able to explain the long standing puzzle of Tarshis that smaller countries, but not smaller regions within a nation, would tend to show greater economic stability. Out study would also call for re-examination of many of the studies linking growth to trade openness, especially those conducted in a cross-sectional framework: We would contend that the measure of the openness that is induced by policy (and other non-structural factors) would have to be proxied not by revealed openness as such, but by the same adjusted for the structural component.
