As was indicated in Chapter 1, this monograph had two principle objectives. The first, and most important, of these was to develop a method for determining the ‘goodness of fit’ between an input-output model and data from the real world. To accomplish this goal, it was demonstrated in Chapter 3 that two-stage least squares may be used to estimate the technical coefficients. Specifically, if this technique is employed, consistent estimates of these parameters and asymptotic standard errors for these estimates may be obtained. In addition, Chapter 3 provided an illustration of this estimation technique using cross-sectional data from twenty-nine sectors of the West Virginia economy. Although they are based solely upon ‘columns only’ information, these results at least cast some doubt upon the effectiveness of the input-output approach. That is, once the data were adjusted for heteroskedasticity, only about 36% of the coefficient estimates could pass a 5% significance test.