Carbon dioxide emission data for the eleven-year period 1990-2000 for the U.S. was interpreted in terms of a simple 3x3 spreadsheet which linked the contribution of this greenhouse gas from the three fossil fuels to three consumption sectors for each year: electrical generation; emissions from mobile vehicles; and emissions from all other stationary sources. Two coefficients were derived from this spreadsheet prepared for each year: (1) C_T1 was defined as the ratio of carbon released from oil consumed in mobile applications to the total carbon entering with oil; and (2) C_T2 was the ratio of the carbon released from fuels consumed in the transportation sector to the total carbon entering with the fossil fuels. These coefficients should reflect the structure of the energy economy and in common with those calculated from economic input-output tables, should change only slowly with time. Though it was not possible to prepare the full spreadsheet for Canada and the U.K. due to limited data, the two coefficients could be derived subject to certain assumptions. These assumptions introduced a small but not significant error.

The values of the Coefficients C_T1 and C_T2 were the highest for the U.S. and the least for the U.K. Coefficient C_T1 increased marginally over the eleven years in the case of the U.S., also increased slowly though more erratically in Canada, but rose sharply and consistently in the U.K. Coefficient C_T2 stayed approximately constant in all three countries. The Standard Deviation for C_T1 was about three times higher than for C_T2. The steady increase in C_T1 in the U.K. suggests the energy economy of that country was undergoing a structural change during this period probably due to the steady substitution of natural gas for both coal and the stationary applications of oil.