During the academic year, 1979-80, over 2,500 students studying first-year economics in nineteen U.K. universities and polytechnics were involved in a research project, the aim of which was to attempt to assess the efficacy of innovative teaching techniques in basic economics. The new techniques included were TIPS (see Allen C. Kelley, 1968), Cases, Programmed Learning, and Macrosimulations. Course packages were constructed, combining innovative and conventional techniques in different proportions. Each course package was designed within the overall objectives of a research design intended to generate a data matrix with sufficient observations in each cell to test the impact of various teaching techniques on different types of students in different institutional settings. Three conventional courses which did not utilize the innovative techniques were included to provide norming data. The common three-hour final examination, which yielded several measures of output, consisted of 20 multiple choice questions to measure knowledge of concepts and simple to intermediate applications, 1 problem-case to measure complex applications and analysis, 1 micro essay and 1 macro essay to measure synthesis and evaluation. There was no choice in the selection of questions to be answered. To ensure that the case and essay marks were consistent across institutions, each paper was regraded by one experienced university lecturer and a sample regraded for a third time to test his consistency. In the three years prior to the experimental year, a total of 26 pilot courses were run in seven institutions to solve logistics problems. Based on the pilot studies, participating institutions adopted one of three broad strategies: (i) the conventional course was scrapped and innovative techniques were substituted for tutorials and essays-complete substitution; (ii) conventional inputs were reduced and innovative techniques were substituted-partial substitution; (iii) innovative techniques were added to existing conventional inputs-add on. Since each teaching technique has a price tag, the strategies produced widely varying course costs, average total cost per student, and marginal cost. An indication of the scope for cost variation can be gained from the fact that the typical conventional first-year course containing 300 students would require 3 lectures and 30 tutorial hours per week (typically there are 10 students per tutorial). In contrast to the use of graduate students in beginning economics tutorials in the United States, tutorials in the United Kingdom are usually shared among all faculty members. A tutorial hour counts as a full contact hour in calculating teaching loads. Thus the opportunity cost of the conventional tutorial system in the United Kingdom could be as high as 7 or 8 upper level courses. Some innovative courses were implemented which scrapped the labor-intensive weekly tutorial system and substituted TIPS, Cases, and one or two hours of class remedial tutorials which students could attend at their own discretion. A comparative institutional cost model was used to derive average and marginal costs. The model included faculty inputs at average U.K. rates, and assumed a 10 hours per week teaching load, and 30 percent of time devoted to research; items such as course as*The Esmee Fairbairn Research Centre, Heriot-Watt University, Edinburgh. Our research was funded by the Department of Education and Science in the U.K., and The Esmee Fairbairn Charitable Trust.