growth, but others must accept "regression." "This is not quite a zero-sum game, but it is also inconceivable that all elements in a capitalist world-economy shift their values in a given direction simultaneously." 14
In parallel to Rostow's concept of temporal progress through stages of growth, E. F. Denison has probably been most ambitious in "partitioning" aggregate growth into "factors" by giving weights to various components of growth, although he deals explicitly with only a single aggregate kind of growth. Adjusting labor for differences in education (à la Marshall and Schultz), and ignoring Salter and Schumpeter, he allows for Smithian economies of scale, Ricardian shifts in production structure, specialization, and disembodied technical progress. Denison's "total factor productivity" with its implicit several types of growth, however, still does not explain all of aggregate growth; he is left with an unexplained "residual." 15
Since the emergence of economics as a legitimate intellectual pursuit, and the rise of their interest in economic growth, economists have resisted any notion that growth might not be homogeneous. 16 Classical economics created taxonomies based on such discriminators as physical resources, policy environments, or historical experiences. Radicals relied on social structures, modes of production, or economic functions, and even on race, culture, and climate, to account for disparities in growth, growth rates, and the effects of growth. Like Marshal and Schumpeter, neither Salter nor Schultz, however, nor even Denison nor Rostow, conceived explicitly that they were dealing with a "different" type of growth. Even development economics generally accepted that growth depended on physical resources and how they were distributed. As recently as 1982 (just as Paul Romer was introducing knowledge-based growth) Angus Maddison, in discussing education, could only "acknowledge this factor as one that has obviously facilitated economic growth, but whose precise role cannot be identified, probably because a shortage of educated people has never been a serious drag on growth in advanced capitalist countries." 17
Only a few, perhaps particularly prescient or perceptive, analysts glimpsed the profound differences between resource-based growth and knowledge-based growth, or the dramatic contrast in their social and political consequences.