Cross-commodity data rather than time-series data were used to determine the incidence of increased marketing margins. Changes in the farm prices of the several food groups were found to be negatively associated with changes in output per man-hour, the best available indicator of changes in cost of production. On the other hand, changes in marketing margins were not meaningfully related to farm price changes. For the same food groups, the market-basket data on farm and retail prices, plus the USDA indices of changes in farm output, provided a framework for showing the extent of the shifts in retail demand, farm demand, and farm supply functions. The indicated shifts in the farm supply curves were generally consistent with the productivity effects previously noted.