Franchising has been and continues to be a very popular way to do business for a number of retailers and service businesses. However, the type of franchising ch ising that has been growing the most, namely business-format franchising, has not grown at the kind of phenomenal rates that the trade press often suggests. Since the Department of Commerce (DOC) canceled its publication Franchising in the Economy, we no longer have access to census-type data on franchising in the U.S. However, looking at the period during which the DOC did publish these darn, one finds that the number of business-format franchisors is highly correlated with the number of units in these chains. Thus, we use data from recent issues of various franchisor directories to assess the number of franchisors in the U.S., and infer from this how business-format franchising has grown in the U.S. We find that business-format franchising has been growing over the last decade at a rate that is, at best, commensurate with the growth of the economy as a whole. We believe that the confusion about the extent of growth in franchising arises, in part, from the fact that many new firms enter into franchising each year, leaning to the notion that this way of doing business is growing tremendously. However we show that many firms also exit from franchising each year, for a net growth rate much below the entry rate. This paper shows that franchising is not a panacea for entrepreneurs, whether franchisor or franchisee. From the franchisor's viewpoint, the high rate of exits suggests that many firms fail despite franchising, and many others choose to stop franchising after trying it for a few years. Clearly, these firms have found that franchising is not right for them. Furthermore, the results show that the characteristics of the chain at the time it becomes involved in franchising, as described in the main franchisor directories-such as the royalty rate, the advertising fee, the franchise fee, the amount of capital required and the sector of operation-have little capacity to explain "survival." The main variable that affects "survival" among those that are typically reported in franchisor listings is the number of years that the franchisor has been in business before starting to franchise. Hence our results suggest this is one dimension in which franchisors can make decisions that affect the probability that they will be successful in franchising. Although we are unable to explain most of the variance in outcome, the results mostly imply that other, less easily observed or quantified characteristics of the chain and the franchisor, such as maybe the "innovativeness" of the product, the amount of support provided to franchisees, the financial backing of the franchisor, etc., likely influence "success" the most, and thus, are worth investigating further. From the perspective of franchisees, the amount of exit found here suggests that in the majority of systems, franchisees cannot expect that their franchisor will be around for the whole duration of their contract-which averages about 15 years according to the Department of Commerce. This does not mean that the majority of franchised businesses will find themselves in an "exiting" system-a small minority of very well-established franchisors accounts for the majority of franchised businesses, and these are likely to remain successful for years to come. Belt entrepreneurs buying franchises from less established systems are likely to face franchisor exit, either failure or departure. This paper confirms that franchisees should thoroughly investigate the franchise system they want to invest in, going beyond the information about royalty rates, advertising rates, rankings, etc., found in franchisor directories, and toward more product, market, and other less easily accessible information about the chain. (C) 1998 Elsevier Science Inc.
