I like to name and shame people who are playing fast and loose with numbers. Usually, this involves taking one “true” data point and then using it to make a point which is unwarranted by the data in context. A couple of examples caught my eye last week.
First up: “Students have at most a 1 in 4 chance that the person at the front of the classroom is a full-time faculty member”.
This is the Canadian Centre for Policy Alternative’s (CCPA) Erica Shaker and Robin Shaban, talking about Ontario Colleges and the number of part-timers in the system (see here for my earlier take on this). Now, let’s be generous and say that their 25% estimate is an innocuous rounding-down of the statement (which seems to be true) that “only 30% of all instructors are full-time”; however, it does NOT follow from this that the chances of having a full-time instructor in any given class is 1 in 4.
Take a hypothetical example. Let’s say you have a college with 1000 instructors, 250 FT and 750 PT (like I say, we’ll grant them the rounding down for the moment). Let’s say further that the average full-time instructor teaches 6 classes a year while the average part-timer teaches 2 (I’m pulling these numbers out of the air because as far as I know there is no public data on this, but if you’re following at home, you can substitute your own numbers here if you think you have better ones). Multiply that out: 250 x 6 = 1500 courses run by full-timers and 750 x 2 = 1500 courses run by part-timers. That’s a 50-50 split, not a 75-25 split. And if you run those numbers without the rounding-down (that is, a 30-70 split, not a 25-75 split), it becomes 1800 vs. 1400, or a 63-37 split.
This should be obvious to anyone who has completed ninth-grade math, so I am not sure what excuse CCPA has for publishing this nonsense. This math, by the way, also applies to all discussions about turning sessionals into full-time staff. Most sessionals teach way less than a full load. Holding the number of courses and class sizes constant (which, absent some magical flood of new money, one pretty much has to), every new full-time hire would eliminate between 2 and 3 part-time/sessional positions. Somehow, this point gets passed over in silence whenever we hear faculty unions talk about “fairness”.
(And, to return to another of my hobbyhorses, this CCPA analysis is exactly why Statscan’s proposed new university staff survey is not a great idea; it will prompt people to base discussions on numbers of personnel and not actual division of total workload. This will make for inane policy discussions).
Next: “Record low youth unemployment”
The second example is perhaps more innocuous than the first, but is important nevertheless. It is a claim made in a backgrounder for the Government of Canada’s recent Fall Economic Statement and you may hear a lot of it the next few months. To wit: that Canada is currently experiencing “the lowest youth unemployment on record”. This is true – the 10.3% rate recorded in September is the lowest since records began in 1976, narrowly beating the previous record achieved in July and September of 1989 – but it doesn’t mean youth employment is booming.
Remember what an unemployment rate actually is: it’s People Without Work Looking for Work (PWWLW) divided by the sum of People Working and PWWLW (the denominator there is sometimes described as (“people in the labour market”). People not looking for work because they are in school full time, for instance, are not included in the calculation. A low unemployment rate can thus either mean “employment is booming” or “there aren’t that many people in the labour market”.
It’s therefore instructive to compare the present day with the summer of 1989, the last time youth unemployment was this low. In July and September of that year, youth unemployment was 10.4%, but youth employment – that is, the fraction of all youth with a job – was over 63%. In comparison, this summer the youth employment rate was hovering between 56 and 57%. That’s not too shabby – it’s about average for the last 40 years and it’s certainly better than it was at any point between 1992 and 2000 </genx get off my lawn speech>. But it is not suggestive of a labour market which is in high gear, either.
The likeliest explanation for the paradoxical combination of low unemployment rate/middling employment rate would be: there’s more 15-24 year olds in school (and hence out of the labour market) than there used to be. That’s probably a good thing overall, but it’s not the image people claiming “youth unemployment at lowest rate EVER” want you to have.
Anyways, stay frosty when evaluating claims made using data. People are trying to fool you: don’t make it easy for them.
Tweet this post
“This CCPA analysis is exactly why Statscan’s proposed new university staff survey is not a great idea; it will prompt people to base discussions on numbers of personnel and not actual division of total workload.”
Agreed… IF the survey ONLY publishes headcounts. The StatCan UCASS consultation process is exploring the creation of some kind of workload measure, and, in my view, this is an absolute must.
I doubt that the problem is so much mathematical as semantic. If by “part-time” Shaker and Shaban mean “non tenure-line,” then that would include lecturers and sessionals with full teaching loads. Around here, that’s 3/3 or even 3/3/1 (with a summer course), whereas the load for tenure-stream faculty is 2/2.
Our 250 notional full-timers would then be teaching 1000 classes, and our 750 notional part-timers (doubly notional — it’s a made-up number and they aren’t really part-time) would be teaching 4500 classes, assuming no summer teaching. So that’s more than three in four.
And that’s not counting the fact that tenure-stream faculty often teach small seminars as part of load, so the odds of any one student facing a tenure-stream instructor in any given class is even lower.
One might object to Shaker and Shaban’s use of terms, but the objection should really be against universities which belie the term “part-time” by using such instructors to deliver so many of their courses.
This is colleges, not universities, so I don’t think tenure enters into it. And while there are certainly non-FT people who teach something like full-loads, a large percentage only teach one class related to their area of professional expertise. Which is why the average load of a non-ft is way below that of an FT.
First, a terminological problem: sessionals, by definition, teach 13–18 hours per week. Partial load faculty teach 7–12 hours, and part-time faculty teach up to 6 hours. So presumably, you meant “most contract faculty teach way less than a full load” rather than most sessionals.
Just to give some perspective, Ontario college full-time faculty can teach up to 18 hours per week. In my experience, typical full-time teaching loads are about 15 hours per week from Sept to April, but that may not be typical across other colleges. Also, there are full-time faculty who don’t teach at all, having been seconded to other duties, so the average teaching load will be lower.
Another point of confusion is that you’re talking about staffing ratios, while the number you cite looks at the experience of a given student. I can’t quantify it, but even if 70% of classes were taught by full-time faculty, that wouldn’t mean that a given student had a 70% chance that the person at the front of the classroom was a full-time faculty member. Full-time faculty teaching loads are governed by a formula that takes into account various factors including the kind of evaluation (e.g., essay vs multiple choice) and the number of students. Contract faculty, on the other hand, have no established limits on the number of students in their classes beyond the limits of the room in which they teach. As a result, it’s quite common for full-time faculty to teach fewer students per hour than their contract counterparts.
Finally, it’s not clear how the 70/30 split is calculated. In one of the programs where I teach, that would be typical for Sept to June, but in July and August, when the program continues to run at full capacity, 100% of classes are taught by contract faculty.
In total, then, I suspect that the 25% number cited is too low, but 67% is much too high, and even 50% seem doubtful.
Right. I missed that. Thanks.
Alex, as Brett describes, the variance is huge from program to program and term to term. I’ve seen 100% contract in some cases. As a contract faculty member, it’s full of surprises. Will I have enough physical seating for the number of students in my class is I suppose a good problem to have. But restricting the course without teacher feedback is common as well.
Meant to say… ‘But restructuring the course….”
Your numbers mystify me, Alex.
As a full-time professor, I could be asked to teach up to six courses per semester, if these were three hour courses = 18 hours in class. If there is curriculum renewal to be done, or coordination duties to take on, then I may get released from one or more sections.
My colleagues who are “sessional” (one version of part-time) can teach a “full” load of 15-18 hours a week, or five to six courses. As noted earlier, partial load are capped at 12 hours or four courses. A part-time part-timer – the semantics get dicey – can only teach a max of six hours or two courses.
While it wouldn’t work in practical terms, let’s say our hypothetical 250 fulltimers are teaching an average of 5.5 courses per semester, accounting for some professors taking some release time for curriculum work and other duties. The ratio of how the parttime teaching is distributed varies from college to college, and from semester to semester. But let’s say of the 750 part-timers, half are truly part time (at two courses), 25% are partial load (at four courses), and 25% are sessional at six courses.
Courses taught by full time professors = 250 x 5.5 = 1375
Courses taught by part-time = 375 x 2 = 750
Courses taught by partial load = 188 x 4 = 752
Courses taught by sessional = 187 x 6 = 1122
= 2624
Total hypothetical courses = 3999
65.6% part time
34.4% full time
Let’s go with my own experience and balance this out a bit with reality. Let’s say of the 750 part-timers, 25% are truly part time (at two courses), 37.5% are partial load (at four courses), and 37.5% are sessional at six courses. Because, in reality, very few people will teach six hours a week for $40-60 an hour, given the amount of prep time and evaluation time that happens outside of the classroom.
Courses taught by full time professors = 250 x 5.5 = 1375
Courses taught by part-time = 187 x 2 = 374
Courses taught by partial load = 282 x 4 = 1128
Courses taught by sessional = 281 x 6 = 1686
= 3188
Total hypothetical courses = 4563
70% part time
30% full time
No question, if you change the assumptions and have sessionals teach more then you will get different results.