The field of economics began with theory (Adam Smith and The Wealth of Nations), later adding data and statistics. By the 1940s, people began exploring human behavior with simple games, which led to the important field of game theory. As computers became more widespread they began to be applied to economics. At UC Berkeley in the early 1960s, Frederick Balderston and Austin Hoggatt wrote an elaborate discrete event simulation of the lumber industry. As timesharing advanced, they envisaged computerizing the business simulation experiments up until then being run with written rules, paper forms, etc.
In the 1960s Hoggat and Balderston established the Management Science Laboratory at Berkeley’s Center for Research in Management Science: a 2800 square feet (260 square meters) space with reconfigurable partitions, teletypes, and closed-circuit video cameras and monitors. A combined PDP-5/PDP-8 minicomputer system ran the experimenter’s application system and logged timestamped responses. One of first experiments was a duopoly situation designed to see whether collaboration would take place even though the subjects were in isolated cubicles, interacting only via simulated market interactions.
By 1970, Hoggatt was ready to upgrade the computer system. APL had arrived in 1968 and Hoggatt felt it would combine experiment programming, data analysis, and publication. But APL wasn’t available on a sufficiently powerful, inexpensive computer at that time. Instead of asking NSF for a bigger computer, he proposed spending the money on building a one-off hardware/microcode/software system that could provide general timesharing as well as efficient APL processing.
The money was granted; between 1972 and 1976 the hardware was acquired and/or built, the microcode and software written, and experiments run. The system ran a few more years but was shut down because of high maintenance costs. Nevertheless it helped establish the field of computer-based behavioral research. Today software packages like z-Tree from Universität Zürich allow easy setup using networks of laptops.
“History of CRMS APL” [IEEE; unlocked] (Paul McJones, IEEE Annals, 2026 Jan-Mar) tells the story of CRMS and its APL upgrade from my point of view—as a person who wrote the APL microcode in 1972 and then left the project, only to take up research for this paper in 2024. The paper is based both on project materials I collected plus materials from the Bancroft Library at Berkeley. The materials I collected are available here: crms-apl.computerhistory.org.