From all these general comments, however, on the achievements and the limitations of the "mathematical" and the "psychological" aspects of all that was "new" in neo-classical as compared with Ricardian-classi-cal economic theory, I turn now, finally, to a different, over-all comparison of the old and new systems of economic "doctrines."
In trying to discover the general "laws" and results of the "natural" operation or functioning of the free and competitive business economy, Ricardian-classical theory had started with considerations about the national economy's aggregate supplies of land, labor, and capital, including the fixity of the supply of land and the tendencies of the labor and capital supplies to grow as long as the resulting growth of output and the real wages and profit shares of that enabled them to do so. Attention was fixed primarily on the physical production of material goods, the ratios of the outputs to the required inputs of labor and capital, and the tendency (with given techniques of production) of those ratios to fall (in agriculture) as the land became more crowded; and this basic vision yielded, together, a theory of production and a directly connected theory of income-distribution or the wages, profit, and land-rent shares of the national output. The theory of "exchange" and the relative values or prices of the different products then appeared only as the final part of the analysis. Here again, as the magnitudes and changes of the public's demands for the different products were merely assumed, with no effort to explain or analyze them or their "psychological" backgrounds in consumers' wants and subjective evaluations or utility estimates, the stress was all on the adjustments of the supplies of products to the market demands whatever those might be, and—as in the end controlling these adjustments—the physical determinants of production costs ("costs" considered as inverse to physical efficiency in production), i.e., the required inputs of labor-time and real resources represented by the final or last-added units of the equilibrium supplies or outputs of the different products.
Now in contrast with all that, neo-classical theory of the economy's functioning began with consumers' wants for and evaluations of and demands for the final products—the "utility analysis" of the subjective backgrounds of market demands and market-or-demand prices, not attempted in Ricardian-classical theory. And from this new starting point, the neo-classical theory proceeded in the opposite direction, or reversed the field. From the demand prices (for given supplies) of the different products, the new analysis derived the demand prices for the productive services, in the production of those products, of the workers, capital, and land employed, and thus moved from the subject of "value" into that of income-distribution, instead of doing the reverse. This also led to the view that all "costs of production" are in their main aspect "opportunity costs," i.e., the payments which the producers of each product have to make for the productive services, of all kinds, which they hire and use, depend upon or must equal the alternative opportunities or possible earnings of the recipients, in other industries, which in turn depend upon the demand prices, in the markets, for the other products. The competitive equilibrium adjustments of the outputs and supplies of the different products to the demands for them would, it was said, indeed make their selling prices equal to the costs of producing them. But the latter, made up in each case of the payments for productive services, made necessary by the selling prices of the other products, could not be regarded as independent causes or determinants of the selling prices of those particular products. The demands in the economy for all its different products, and the resulting, rival demands for the productive services of its limited supplies of all factors of production, would determine both the prices of the products and those of the productive services, hence also the costs of producing the products. This argument, however, though of course valid in its own field of Reference, and implicitly recognized in Ricardian-classical theory, was no argument against—did not join issue with—the real thesis of the latter's value theory, which in its view of the price-determining role of production cost had reference not to the prices but to the quantities of he productive services "embodied" in a unit of each product. The two theories or views on this general question were really not inconsistent but complementary—each stressed the part of the complete answer which the other "took for granted." But neo-classical theory did lead to a better logical unification of "value and distribution" theory, or theory of the working of the price system in the "pricing" both of products and of the services of the different factors of production. And in the latter field, by generalizing the old "law of diminishing returns" into a law of decline of the "marginal productivity," and demand price for the services, of any factor whenever the supply of it increases relative to the available supplies of the other factors, the neo-classical analysis in a sense ended in the area in which Ricardo's started, but with a more symmetrical and flexible or, in short, general view of the problems in that area.
Yet here again the one-sided stress on productivity and demand, in contrast with Ricardo's stress on cost and supply conditions, led to a different delimitation of the field of inquiry. In Ricardo's theory of come-distribution, the supply of labor was expected to increase to the point of reducing its real wages to the "subsistence" level—the "cost of production" of that labor supply—and the supply of capital, likewise, was expected to increase until the rate of profit or interest would fall so low that no more accumulation would occur; only in the case of land-rent would the changes of "demand" have the decisive influence, since the supply of land is fixed and without any cost of production. But in neo-classical theory the tendency was strong to ignore or rule out the problems of the ultimate "supply prices" for labor and capital, and extend Ricardo's view of land and rent to all the factors and income shares, i.e., to regard the aggregate supply of every factor in the whole economy as simply "given" at each point of time and not to be explained by economic theory, and to consider productivity and demand conditions only, in "explaining" income-distribution. This in a way corresponds to the contrast, at the other "end" of the entire route traversed in opposite directions by the two types of theory, of the Ricardian neglect of, and the neo-classical stress on, the foundations of demands for the final products; as the neo-classicists added what Ricardo omitted at that "end," they omitted what he mainly emphasized at the other "end."