PRESIDENT'S DESK

Beyond analytics

Steve Robinson

Steve RobinsonINFORMS Presidentpresident@informs.org

INFORMS’ core strengths are in operations research (O.R.) and the management sciences (M.S.). We need not only to retain these, but also to continue to strengthen them because they enable everything else we do. We’ve now added data-related capabilities covering descriptive, predictive and prescriptive analytics with particular emphasis on the prescriptive work. One way to visualize the relationship between OR/MS and analytics is to imagine two concentric circles. The inner one contains our core disciplines, O.R. and M.S. The space between the two circles contains analytics, which connects the technical core with the vast array of decision problems. That connection takes place at the outer circle.

An old adage in management science is that the place in a system that requires the most careful attention is its boundary. In the system just described, the boundary was the outer circle, the area outside that circle contained decision problems, and the connections occurred at the circle. This is appropriate for problems that can be resolved by OR/MS techniques augmented by current capabilities for handling and analyzing data. However, there are other problems for whose resolution these are necessary but not sufficient.

One example is global warming. It requires all of the technical disciplines just described, but few people would claim that those suffice. Much of the scientific knowledge and modeling experience required to create and to assess climate models falls outside our disciplines. Further, the use of whatever understanding may come from those models to bring us closer to where we want to be requires skills in negotiation and political management, not only for bringing us closer but even for creating a common understanding of where in fact we do want to be.

A great many problems get far less publicity than does global warming but are still of great importance to many people. For example, enhancing transportation to improve the economic condition of an area will not work if those influential in the local government block the new arrangement to preserve their monopoly on transportation. Similarly, a program of building wells to provide clean drinking water in underdeveloped areas will fail if those in the villages are not both able and motivated to keep the pumps in good repair.

In these examples, successful solutions must use technical principles, but must also employ analysis and understanding of the political and social situations to construct implementable solutions, and negotiating skills to help secure their implementation. A typical O.R. graduate should know how to model the technical aspects of such a situation, but may not even see what the real problem is because key elements are not in the technical dimension at all, but rather in the economic, social and/or political dimensions. The graduate may also be unable to communicate about the problem with stakeholders in terms that they can understand.

Thus, for some of the hardest problems there is another category of knowledge and skill outside the domain described above – a third circle, containing knowledge from disciplines needed to fit solutions to the environment in which they must function. I will call the skills contained in the three circles systems analysis. Moreover, because of the three factors of scale, interdependence and politicization, the proportion of problems requiring systems analysis skills is larger today than ever before. Present education in OR/MS provides insufficient skills to cope with these problems, so practitioners need access to additional skills.

The words “access to” are crucial. Present students learn the technical dimension, which is what they’re good at. They have no time for formal instruction in sociology, politics and mass communication. But they can learn how to work effectively with experts in these other dimensions. An excellent way to do so is through actually doing such work on real problems.

The International Institute for Applied Systems Analysis (IIASA) has been using this method since 1977. Located in Laxenburg, Austria, IIASA (www.iiasa.ac.at) is funded by national member organizations (NMO) in 22 different countries; the United States NMO is the National Academy of Sciences. For 40+ years IIASA has produced excellent basic and applied research along with substantial policy analysis that has been successful in a number of countries. IIASA’s annual three-month Young Scientists Summer Program (YSSP) offers research opportunities to talented young researchers, who then work alongside IIASA scientific staff on current projects.

The 2014 YSSP program had 48 participants. It provides an excellent proof of principle, but nowhere near enough graduates to meet the need for systems analysts.

INFORMS could help to enlarge the supply of systems analysts by making the importance of this field better known, which would increase the pressure on institutions to offer instruction in this area to their students. One initial step could be the formation of a section within INFORMS or the adoption of this area by an existing section. This would provide an environment for interested people to communicate and to initiate programs, while also making systems analysis much more visible to students. Given the orientation of current students toward service in helping to address social problems, many would probably be interested.

It’s worth trying.