In 1964, Baker and Pound |6j surveyed the state of the art of evaluating and selecting R & D projects. Although their investigation focused solely on R & D projects, their findings, and the subsequent findings of Baker and Freeland (4, 5) lead to some tentative conclusions about the past, present, and future use of project selection methods.
The use of formal, numeric procedures for the evaluation and selection of projects is a recent phenomenon, largely post-World War II. At first, payback period (and the related "average annual rate of return") was widely used. It is still used by those who feel that the uncertainties surrounding project selection are so great that a higher level of sophistication is unwarranted.
The use of formal models slowly increased during the 1950s and 1960s, and a large majority of the models employed were strictly profit/profitability models. As we have noted, the emphasis on profitability models tended to shorten the time horizon of project investment decisions. This effect and the results of several stud-
2.7 THE PAST AND FUTURE OF PROJECT EVALUATION/SELECTION MODELS 79
ies on the use of project selection models are reported in Mansfield 139, App. A]; also see (40, pp. 15—16).
A similar effect on non-R & D projects is easily observed by noting the sharp decline of investment in long-term projects. The increasing interest rates seen during the 1970s forced cutoff ("hurdle") rates of return higher, which cut back investment in projects for which the time gap between investment and return was more than a very few years. For example, neither new steelmaking capacity nor copper-refining capacity was expanded nearly as rapidly as long-run growth in the demand for steel and copper seemed to justify during this period. Producers tended to blame the lack of investment on foreign competition, but given the aging capacity in the United States, it may well be that the level of foreign competition is as much a result of the lack of growth (that is, our failure to invest in newer technology) as it is a cause. Again, the reader is referred to Hayes and Abernathy |25).
A decade later, Baker |4] and Souder |60) reassessed R&D project selection. In this decade there was considerable growth in the use of formal models, again with great emphasis on profitability models. But Baker reported significant growth in the literature on models that use multiple criteria for decision making. He observed a trend away from decision models per se. and toward the use of decision information systems. Among other reasons for this change, he notes [4| that "the decision problem is characterized by multiple criteria, many of which are not easily quantified, and the typical approaches to quantifying subjective preferences are far from satisfactory." He also notes the development of interactive decision systems that allow users to examine the effects of different mixes of possible projects.
More than two decades have passed since Baker's 1974 study. Considerable progress has been made in the development of processes for measuring preferences that yield suitable input data for sophisticated scoring models, models which serve, in turn, as data for goal programming and other resource allocation models. Because it is easy to enter all the parts (data base, decision model, and list of potential projects) in a computer, it is feasible to simulate many solutions to the project selection problem. The decision maker can easily change the criteria being used, as well as the criteria weights. Decision makers can even investigate the sensitivity of their decisions to changes in the estimates of subjective input data, thus directly examining the potential impact of errors in their opinions. In spite of all these capabilities, Liberatore and Titus |35] have found that mathematical programming models are not used for project selection or resource allocation, at least in the firms they interviewed. They did find, however, that scoring models were used for selection—particularly when the firm dealt with outside funding agencies.
We believe that use of these techniques will be extended in the future. As we become more familiar with the construction and use of decision support and expert systems (see (65|), the simulation of project selection decisions will grow in popularity. It seems to us that two concurrent events will support this trend. First is the rapid growth in the ownership and use of microcomputers by organizational executives. The operation of a computer is no longer seen as restricted to computer specialists. Second is the growing realization that profitability alone is not a sufficient test for the quality of an investment.
Almost everyone who has studied project selection in recent years has noted the need for selection processes using multiple criteria. The writings of Michael
Porter [47, 48| and others have emphasized the role of innovation in the maintenance or improvement of a competitive position. Indeed, it is now clear that the firm's portfolio of projects is a key element in its competitive strategy. Suresh and Meredith |62] have added a "strategic approach" to the problem of selecting process technologies for implementation. In sum, the methodology and technology for multiple-criteria project selection not only exist but are widely available. Perhaps more important, we are beginning to understand the necessity for using them.
Was this article helpful?
What you need to know about… Project Management Made Easy! Project management consists of more than just a large building project and can encompass small projects as well. No matter what the size of your project, you need to have some sort of project management. How you manage your project has everything to do with its outcome.