Leathers (1991) shows that while the existence of allocable fixed inputs can cause joint production (as in Shumway, Pope and Nash, 1984), it will not necessarily lead to joint production. The extent to which allocable fixed inputs cause joint production in agriculture is an empirical question. This paper offers an empirical answer. By estimating a short-run joint cost function, it is possible to identify levels of outputs for which joint production may be optimal in the short run but not in the long run. Only in these output regions will there be jointness caused by allocable fixed inputs. For the data in this paper (160 Wisconsin farms), these output regions are very small; thus allocable fixed inputs do not appear to be an important cause of jointness for these farms. Technical causes of jointness appear to be a significant cause of joint production.