Actuarial Education: Fifty Years of Evolution

By Ian Duncan

Expanding Horizons, November 2023

This is the text of a speech delivered during the gala dinner at the 2023 Actuarial Research Conference at Drake University, Des Moines, IA.

Thank you for the introduction and opportunity. I bring greetings from John Robinson, the Society of Actuaries [SOA] president, who is unable to be with us this evening. I am John’s humble understudy—as Winston Churchill said of Clement Attlee, “Attlee is a humble man with much to be humble about.” Nothing humble, of course, about Churchill. I will add that while I will discuss two SOA initiatives in what I have to say, I will occasionally express a personal opinion, not the opinion of the Society of Actuaries. I will try to signal when one is coming.

There is a considerable amount happening at the moment at the SOA. I was on the board 10 years ago, and compared with now, the pace was slow and the issues relatively manageable. I will address two of our current issues that are the purpose of our Section: Education and Research. Those of you who were able to attend Stuart Klugman’s session earlier today will have heard a little of the changes that we are making to the Fellowship pathway. I want to give some of my perspective.

If you will pardon a little excursion into history, I joined the Institute of Actuaries in London in 1974, so next year is my 50th anniversary. There are probably few in this room who have seen the evolution of the profession over this span of time (Professor Shapiro and Professor Klugman may be among the few). I was a reluctant actuary—I actually wanted to be an investment banker (investment banking being the data science of alluring jobs in those days). After failing to get a single interview, I chose the (to me) next best thing: actuary. In my opinion, viewing the evolution of the profession over a long time gives me license to comment on current changes.

I wrote my first exam, which was called Compound Interest. We seem to have promoted this exam with a more sophisticated name—Financial Math. I failed my first exam. I put the blame squarely on the fact that in those days the Institute required written exams using logarithm tables. Today’s students probably don’t know what log tables are, but if you ever have an idle hour, try doing FM exam problems with log tables. Of course, multiple-choice questions were an American invention so no multiple choice either. By the time I rewrote the exam the next spring, I had graduated to a faster method—the slide rule—and I passed. The pocket calculator—Texas Instruments, if I remember correctly—was just coming into use and was prohibitively expensive. I passed all my subsequent exams with the aid of the slide rule.

At the next sitting, I passed three exams; one was a retake of Life Contingencies (slide rule again), and the other two were Fellowship exams. I think that’s some sort of record. I have my son to thank for this; he was a couple of months old and was fed at 4 a.m., so I got up and studied. Alas, my exam career was downhill from there: I failed General Insurance [GI] three years in a row and passed it at the fourth attempt. When I look in the mirror, I see an actuary who is eminently qualified to consult on General Insurance. Mrs. Duncan, who is an FCAS, does not agree and has forbidden me to work in GI.

I mention all of this to show how far we have come in 50 years.

But have we really come that far? I both teach students and hire them for my consulting company. The change to the Associateship exams that requires students to be competent in programming has been a big win (all my employees are able to program). But are we really educating competent actuaries? There are, in my opinion, some shortcomings. I had the benefit of studying economics at Oxford, where you were required to write weekly essays and read them to your tutor (how’s that for intimidation?). This requirement sharpens not just writing and perhaps presentation skills but the ability to reason. Writing isn’t about elegance; it is about taking a position and defending it. I find this ability woefully lacking in our students and young employees. They are certainly better prepared than I ever was in mathematics and statistics, but where it matters for an actuarial career—reasoning, speaking and writing—not so much.

Perhaps the changes that we are introducing to the Fellowship track will help to address this. A major difference between the Society of Actuaries today and the profession of 50 years ago is that we make a point of listening to candidates. What a concept! That whirring noise you hear is the president of the Institute of Actuaries in 1974 rolling in his grave.

Many of the enhancements that we are making are in response to feedback from candidates. To mention them quickly:

1. The Fellowship will now look much more like the type of university course that students are accustomed to, including the ability to choose subjects.
2. The ability to take exams three times per year. I would have become a fellow three years earlier had I not had to wait for annual exams.
3. Course material that is structured toward the learning goals rather than being a random collection of readings.
4. Making much of the regulatory material optional through certificates. Alas, this change happens 40 years too late for me; I had to write parts 9 and 10 of the old SOA exams to qualify for my FCIA. It was a shock after the UK, where regulation was light, to have to study 1,000 pages of New York Insurance law. Students who do not require detailed regulatory knowledge will be able to study a less intensive course.
5. Faster grading. There are many enhancements to grading that I think will speed up the process. I grade the PA exam because I had a lot to do with its introduction 10 or so years ago. I don’t see the need to second-grade half the papers.

All of these changes should speed a candidate’s path to Fellowship and reduce attrition along the way, without reducing the rigor for which actuarial exams are famous. Attrition is a very real problem for employers: I have lost a number of employees to data science or even other industries where actuarial training is valuable. The current exam system is a big detractor for the profession. The one thing that the general public knows about actuaries is that the exams are very difficult. Wouldn’t we rather be known for the quality of our contributions as problem-solvers and risk managers than for the exam process? We employers invest in our employees, and an employee leaving represents a significant loss; we need to rethink how we test candidates to keep them engaged without losing the rigor of our current exams. I do hope, though, that the examiners are able to introduce more reasoning-type questions so that our candidates are able to be true business partners.

I want to turn briefly to the other SOA mission: research. One of the significant enhancements made by the SOA in the last 10 years or so, in addition to significantly increased research budgets, has been to introduce structure to what was previously a pretty haphazard research process. As a result of these changes, the SOA can point to significant numbers of publications, reports, models and other deliverables and thousands of page views. A couple of years ago, the SOA introduced the SOA Research Institute to coordinate research and to allow the SOA to raise external funding. The latter has been developing more slowly than some of us hoped, in part because the SOA is not connected to large grant-giving institutions. I see this in my own field (medical), where the reaction of the NIH and NSF can be summed up as “Actuary? What’s that?” More recently, the SOA Research Institute has turned its focus to partnering with organizations that are awardees of large grants, such as NASA. You may wonder what NASA has to do with actuaries. The answer is translational research. I see this in my own practice, where I am on the advisory board for an NSF grant to develop devices for patients with diabetes. My role is to help the researchers figure out how to get paid for their devices and inventions—and in medicine, that means health insurance companies. This type of collaboration could be an exciting opportunity for both academics and industry practitioners.

Continuing on the theme of translational research, I want to appeal to the academics in the room because I think the gap between primary research and its application in practice is woefully large. Alas, practitioners don’t read the North American Actuarial Journal, however hard we try. A few years ago, we introduced the concept of brief essays distilling the findings of papers in the NAAJ for practitioners. These essays probably help a bit to introduce practitioners to new methods, but much more can be done to bridge the gap between theory and practice. On the practitioner side, I think we are making some progress. Our calls for research proposals are generally oversubscribed, and proposers are most often from industry. Personally, I wish there were a way to train practitioners in research fundamentals. The SOA’s current post-fellowship focus is on certificates; perhaps we need a certificate in research.

And I will raise one more personal plea: the US health care system is 20 percent of the US GDP. Yet Margie Rosenberg and myself are the only two academic actuaries publishing research in this area (I will add Brian Hartman, who publishes in health, although I do not think he would claim this as his focus). I don’t know how to foster more interest among academic actuaries, but I do know that we need to increase our presence in health significantly to be taken seriously.

The good news is how many opportunities we have. As we end, I will say with confidence as the president says at the end of the State of the Union speech that the State of the Society of Actuaries is strong!

Statements of fact and opinions expressed herein are those of the individual authors and are not necessarily those of the Society of Actuaries, the editors, or the respective authors’ employers.

Ian Duncan, Ph.D., FSA, FIA, FCIA, MAAA, FCA, CSPA, is an adjunct professor at the University of California–Santa Barbara. Ian can be reached at duncan@pstat.ucsb.edu.