Trends and changes in total factor productivity (TFP) have been the subject of a considerable amount of research from both theoretical and empirical approaches, at both the microeconomic and macroeconomic levels, as a driver of economic growth and fluctuations. A traditional growth accounting method employed in such research assumes the marginal productivity principleiMPPjfor factors of production. In this paper, we conduct international comparison of TFP growth rates between those measured when assuming the MPP and those measured when not assuming it, to determine the degree of deviation in TFP growth rates under different assumptions and examine its causes from the perspective of business cycles, using data on the five developed nations (Japan, the United States, the United Kingdom, Germany, and France), When not assuming the MPP, TFP measurement requires identification and estimation of the production function. After comprehensive judgment of the results of various estimates conducted under differing assumptions, we decided to employ a Cobb-Douglas production function. Based on this, we compared traditional TFP growth rates assuming the MPP and technological or true TFP growth rates not assuming the MPP, to measure and assess the bias by which the marginal productivity of labor and capital deviate from the rewards of each. For Japan, we conducted similar analysis using panel data at an industry level. The results of this measurement show that, while conditions vary by country and over time, the size of the bias is of a level that cannot be ignored in comparison with the level of the traditional TFP growth rate, and consequently that the degree to which the traditional TFP growth rate deviates from the true technological TFP growth rate is, at some times and in some cases, clearly not inconsequential. Putting another way, the likelihood is high that the marginal productivity bias and deviation between the two TFP growth rate indicators is cyclical and characterizes the historical business cycle in each nation.
