The physical scientist, working in his laboratory, conducts carefully planned experiments under strictly controlled conditions. The data thus generated can be measured with remarkable precision. The economist-scientist, on the other hand, neither conducts controlled experiments nor finds economic data readily available or precisely measured. The data problem is particularly acute in the study of economic history, where records are usually incomplete, inaccurate, or nonexistent. In light of these problems, the "discovery" of an experiment in alternative economic systems, conducted under (semi-) controlled conditions and accompanied by a body of high-quality data, is certainly rare, if not unique.