For colleges/universities the student dropout rate is normally defined as the percentage of students who start a course but don’t complete it with a passing grade. In the UK, university student dropout rates seem to be between 0 and 10%.
For professional bodies requiring exams to qualify (e.g. for accountancy, actuaries, and lawyers), the student dropout rate could be defined as the percentage of students who join but never qualify.
Qualify could have different meanings. The IFoA has been using qualify to mean reaching either Associate or Fellow, so using that definition, for a particular year of entry n we can calculate the qualification rate Q1 = Number of students who become Associates or Fellows / Total number of students who joined in year n, and an associated dropout rate D1 = 1 - Q1. (A similar dropout rate D2 could be calculated using a qualification rate Q2 using only the number of students who become Fellows, where Q2 <= Q1, and hence D2 >= D1). But for comparability with the IFoA I’m interested in looking at D1. For example, if the median time to qualify is n years, then an estimate of D1 would be 1 - Q1 where Q1 = number of new qualifiers today / number of new students n years ago. This figure could be calculated for each of “today” being 2021, 2020, 2019, 2018, or as far back as figures are available.
To estimate this, I’m been trying to find statistics on the numbers of new qualifiers and new students joining in recent years, but I can’t seem to find any of this on the SOA website. A search for annual reports only seems to show the SOA Annual Report from 2009?
For CAS, annual reports do seem to be readily available from the CAS website. E.g. I can see the numbers of new qualifiers (Fellows and Associates) for each of 2021, 2020 and 2019 on page 10 of https://www.casact.org/sites/default/files/2022-06/Annual_Report_2021.pdf. But I can’t see any information on the number of new students - perhaps because (if I understand things correctly) students can’t join the CAS, only Associates or Fellows do (with a very small number of Affiliates)? Instead CAS students seem to join CAS Student Central, however, it does not seem easy to find out how many students join CAS Student Central each year?
Does anyone know whether the societies themselves have published any information about qualification rates or student dropout rates? If not, do you have any suggestions as to how to find out the numbers I am looking for (numbers of new qualifiers each year, numbers of new students each year, median time in years to reach Associate)?
No one really “joins” cas student central, unless things have changed a lot. All the CAS knows about students is how many sign up for exams. I think you can find information on how many people sit for exams each sitting. There may not be public information linking those numbers to who eventually qualifies.
It would be a soft number at best, though, as there is no set time to get through the exams. Some candidates sit for one or two exams every time they are offered. Others aren’t being supported by an employer, and just sit for an exam here or there, as they fit it into their schedule. So when do you judge sometime has dropped out?
The CAS used to publish information on travel time, though, that included how long it took the Nth percentile to reach ACAS. That might be close to what you are looking for, and old travel time reports may still be available.
Not a fan of “travel time” since it only tracks successes.
I suggest taking various time periods, say 10 years to start. 1990-1999 Exam 1 (or whatever “the first exam” has been called), number of sitting for that exam.
Compare it to, say, 2000-2009 FCAS numbers.
Now, maybe 10 years is not good or reasonable. I have no idea what is. For the most part we are looking for that number of years N where the ratio of FCAS in Year S+N years’ hence to starters in Year S is roughly constant (or constantly increasing or decreasing).
Then, the rest of the starters can be considered dropouts.
One might want to adjust the number of starters downward for repeat takers. How much? Well, you’ll need to know the number of passers; then take, say, half of failing candidates and subtract from the total takers for the ten years.
Is half right?
Anywho, 90% “dropout rate” is my guess. I have absolutely no data to back that up. I’m not insterested in working on it. Good news is that you are.
I’m not saying the exams are difficult for a STEM grad. I’m saying a lot of people think they can pass it, don’t study, don’t pass, then decide they don’t want to study to pass, or get some other job that requires less studying on one’s own time.
Did you read my post? The travel time stats weren’t just “median of those got fcas this year”, they also included “how did the 20th percentile of those who started in that year fare?”. They did, in fact, track starting cohorts, to the best of the ability of the volunteers who worked on it to identify a “starting cohort”.
Sadly, that’s among the things the CAS stopped publishing after the SoA attacked it. And they may have removed the old ones from the shiny new (and much less useful) website. The best I can find is this infographic version.
Older ones were much more useful. They do claim on the infographic that the median time from first preliminary exam to FCAS was a little over 8 years – meaning that the median candidate DID achieve FCAS. Also, they say “Almost 2/3 of candidates who register for a CAS exam become members of the CAS, and about 45% become Fellows”. That conflicts with “half who start make it to fellow”, so I don’t know what’s up with that. (And many people register for an SoA prelim, intending to take CAS exams, but don’t make it to register for CAS exams. But the CAS doesn’t know about them at all.)
Anyway, you may be able to get some good estimates from that site for your own purposes, even though there’s a lot of survivorship bias and there may be some issues w/ duplicate names/changed names.
I really don’t think that the SOA or CAS is going to want dropout rates published. I suspect that DTNF’s 90% guess isn’t too far from the truth base upon watching my co-workers success rates over the years.
If you assume that the pass rate for an exam is about 1/3, there is about a 30% probability that a candidate won’t pass the first exam after three tries. At that point, most of the companies I’ve worked at kick you out of the study program which makes it very difficult to pass exams and puts your job in question so many people leave for other opportunities. So if we use that as a rough guide, about 70% of people will advance on an exam. Which implies that there is about a 35% possibility that a person will pass 3 exams. After passing three exams people might have a bit of sunk cost that pushes them on to finish with their Associate designation. So an estimated dropout floor of 65% is not unreasonable. This is entirely a paper napkin kind of analysis so…
I suspect that if the number was inversed, the SOA or CAS would want to show that number as it would be a good recruitment tool to encourage people to enter and stay in the field. From my perspective, the actuarial societies pride themselves on the hard exam process and like to keep out those who are too stupid or lazy as DTNF pointed out. A high dropout rate is an intended byproduct of that philosophy.
We definitely have a certain number of members who leave in ways other than death, if that’s what you mean.
Obviously, some leave due to retirement, but some leave because they do not need membership for anything. I knew a few who let membership drop as it was a detriment for what they were trying to do at the time.
In all these cases, of course, they weren’t actually doing any actuarial work.
“In all these cases, of course, they weren’t actually doing any actuarial work”.
That may the case in the US. But very interestingly, the IFoA has recently said (to the UK High Court in a judicial review case that I’m currently bringing against them) that most actuaries don’t need to be members of the IFoA (and by implication of any other organisation although they didn’t state this explicitly) to practise as an actuary in the UK. See https://www.reddit.com/r/Actuary_news/comments/x4zd0v/save_yourselfyour_employer_between_236_and_715_a/
In the U.S., only some actuarial work legally requires membership in one of the U.S. actuarial orgs, but we did have a few lawsuits back and forth between the competing orgs re: qualification standards.
I have never done any work that requires (by statute) being a credentialed actuary to do it. I would say that’s true of most American actuaries.
At the company I work at, its primary business is asset management for insurers and other institutional investors that have to support a liability (such as endowments or pension funds). We have other businesses that do risk modeling, economic modeling software, and insurance research (the part I’m in).
We do not require anybody to be actuaries for any of this, but we have actuaries in all parts of our company. We just hired another actuary for insurance research.
I’ve mostly done very traditional actuarial work, and I’ve frequently referred to papers on the exams to help me in my work. Nonetheless, I’ve never held a job that had a legal requirement that i be a credentialed actuary.
There’s a tiny handful of US actuarial jobs that has a soft legal requirement of being a credentialed actuary. (“Soft” because even in those jobs it’s possible to get a legal exemption.)
I began doing work that required a credentialed actuary as soon as I got my ASA. Actually was doing work that was relied on by the credentialed actuary not super long after starting my job.
I’ve definitely signed rate filings before that required credentials. My job didn’t require CAS membership, per say, but I wouldn’t have gotten the job without credentials so… implicitly required.