One of the strengths of the NIPAs is the double-entry bookkeeping used in their construction. Independent estimates of total output are developed on the basis of the dollar value of output sales, on one hand, and the dollar value of payments to factors of production, on the other. In principle, these two independently derived sums—the product side and the income side estimates of GDP—should be equal. The difference between the two estimates is the statistical discrepancy, which by construction differs from zero only because of measurement errors. In the conventional accounts, a small statistical discrepancy suggests that the value of output has been well measured, since two independent measurement methods give approximately the same answer; a larger statistical discrepancy signals the existence of measurement problems.
Interpretation of the difference between input costs and output values is somewhat less straightforward in the case of a nonmarket account. In a competitive market context, an inefficient firm—one for which the value of the resources employed exceeds the value of the output produced—will eventually be driven out of business. Competitive pressures do not operate in the same way in the nonmarket context. That households optimize with respect to their allocation of time is a more tenable assumption than the alternatives, but households that fail to optimize are not driven out of business and may continue to exist indefinitely. This introduces the possibility that, depending on how it is measured, the cost of time devoted to home production could exceed or fall short of its productive value.
The conceptual equality of output values and input costs in the market accounts also reflects the convention that is employed for measuring capital costs. Revenues not spent on other costs of production are considered to be a part of the cost of capital; put differently, capital is treated as the residual claimant. An alternative approach to valuing capital services—and one that seems applicable to the nonmarket accounts—would be to use a standard measure of the flow cost of capital. Such a measure would be constructed using information on the value of the capital employed to produce the nonmarket output and an assumed market rate of return to that capital. Using this approach, the cost assigned to capital services could be greater or less than their productive value.
Capital-market constraints, such as those that might arise from lenders’ reluctance to finance the production of assets that cannot be marketed and therefore cannot readily serve as loan collateral, may be particualrly important in the nonmarket context. Absent capital market constraints, larger investments might be made. Because the amount of investment is constrained, however, the return on investments that do occur will be higher than the market rate of return. Valuing nonmarket investments in a fashion that ignores this possibility by, in effect, imposing an assumed normal rate of return—for example, valuing educational