How much does it really cost a school to add one more student? I remember a debate about letting Maine National Guard members attend University of Maine Schools for free, and students were not happy about the possibility. I understand most of a college’s expenses are fixed.
A single student can probably squeeze into existing space with negligible added costs. Nobody is going to offer a whole new section of a class because enrollment went from 147 to 148. At minimum, there might be added hourly work for the administrative staff.
But adding several new students would require adding a new sections, which means hiring adjuncts at the very least. If there isn’t spare classroom space, there’s the cost of making more classrooms: converting other unused rooms and upgrading existing classrooms.
If we’re talking hundreds of new students, that will require new buildings to make enough classroom and dorm space for everyone, and hiring more professors, instructors, and administrative staff to handle the expansion.
How many Maine National Guardsmen would choose to go to Maine colleges if the tuition is free? As a first WAG it could easily be hundreds, if not thousands.
This type of analysis applies to pretty much everything.
Most products and services are provided by a long chain of lumpy assets each of which does part of creating & delivering the product. As long as any lumpy asset is operating at less than 100% utilization, you can run one more unit of production through that one asset at (almost) zero cost.
But in any complex delivery chain there are going to be some assets more fully utilized than others. And a lot of business optimization of the last 20-30 years has been about finding those underutilized assets and either working them harder or down-sizing them to be more fully utilized.
So once you start running more than a very few marginal units through your production system you begin to have bottlenecks where you need to upsize the productive capacity of some link in the chain. At which point you end up spending on a lumpy chunk of capital improvement costs or a lumpy hunk of incremental labor costs to process the incremental load. And once you’ve invested in that incremental chunk, odds are this link in your chain is now a bunch underutilized compared to its peers.
The bottom line is that in a modern highly flexible production system marginal cost is not as useful a metric as it was in a 1950s smokestack factory.
The large-scale consequence of all this is that, as **lazybratsche **said, the marginal cost for 100 incremental units is larger, often much larger, than 100 times the marginal cost of 1 incremental unit.
Deciding whether the University of Maine more resembles a modern flexible supply chain or a 1950s smokestack is left as an exercise for the reader.
Usually, but not always: Occasionally, the marginal cost of 1 increment will be nearly as large as the marginal cost of 100 increments. On average, the marginal cost of 100 will be 100 times the marginal cost of 1.
Related to what was said, at my school, we can’t just take extra students. There is a certain amount of students that the infrastructure can handle, and we set the enrollment target at that amount. As capital investments are made, that enrollment target can go up - usually in chunks of a few hundred at a time. If we had a program of admitting tuition-free students, if the group was around 50 it would likely come at the expense of offering admission to other students. If the group was around 150, it would likely require new dorms, study space, labs, network capacity, etc.
(Realistically, the infrastructure can probably handle at least 10% more than the enrollment target, in case we misundersestimate enrollment. So we could take 50 extra students if they just showed up [especially if they were full-pay] but if we were expecting them, we’d need to plan in the budget)
We actually know now that the marginal cost to teach a student computer science, at the graduate level, online, is between $130-$170 per credit hour.
I’m referring to the Georgia Tech online masters which charges this, and according to the information they have said publicly, is able to break even, paying all it’s costs, at this rate. (they recently raised it to $170 which may be profit taking or it may be closer to actual costs)
Notably, on the tuition and fees page for the school, they charge $760 a credit hour to take courses on campus. And that’s in state tuition - additional money, per student, is coming from the state of Georgia.
Basically, what that pays for is the school films these fancy video lectures - using blue screens, with onscreen graphics and a script and so forth - and supplies additional material as well. Some of the money must go to this ongoing effort. They also have to pay TAs, at $20 an hour, to answer questions, grade papers and exams, and to Proctor exams. (there’s some kind of proctoring system where they can view your actual desktop and I think see you through a web cam to discourage cheating during exams)
Anyways, these rates are far more sane than others have been charging, and the content is supposed to be pretty good, and the exams are the same exams they give to the on campus students. Also, the program is considered to be in the “top 10” of all computer science programs.
This makes sense. You can also offer programs in a massive number of other useful fields this way, at cost numbers around this ballpark. Even courses that need hands on stuff - like soldering and work with electronics - you could probably ship to students at home practice kits for a lot of it.
If that school breaks even at that rate (which I’ll assume for the sake of argument), then all we know is that this is their average cost per student. The marginal cost can be very different.
Marginal cost must be lower than average cost, always (I cannot think of any possible exceptions because average cost = total cost/unit quantity. Total cost = fixed costs + marginal costs. My point is that we know that the marginal cost is less than $170/credit hour)
You can have marginal costs higher than average costs, if someone is paying you to do something but doesn’t care how much you do. It’s rare, but I think I’ve heard of tax incentives that end up working this way.
That means the fixed cost is negative. Yes, that’s possible, just rare, and the cost to society is still negative. In the case of a college education, ultimately, the government of the United States/the individual states has a strong incentive to make sure it’s citizenry is education.
Better educated citizenry make more money and pay more taxes, they commit less crimes, and you can have some of them produce more advanced weaponry and technology to make the government stronger. The government benefits and makes a net profit if it pays to educate it’s citizens. Frankly, it would still make a profit if it just paid the whole thing instead of forcing people to come up with tens of thousands of dollars to pay for a subsidized state school education.
Banks don’t make a profit, usually, because loaning money to someone with no credit record and who can wiggle out of paying you back isn’t a good loan. Governments do because they get to tax citizens, guaranteed, from cradle to grave, and also gain any net benefit from an education citizen committing less crimes.
I chuckled.
Sometimes a typo is just a typo. Other times it’s inspired.
No marginal cost above average cost does not mean the fixed cost is negative. That is true if costs are linear a fixed cost and a constant marginal cost. But there are certainly other situations. One case is in electricity production. As more electricity is demanded, utilities must switch production to less efficient or more costly methods. They might switch from hydro to natural gas generators to coal generators (which have large cleanup costs). This is why those who have variable pricing typically pay more for electricity on very hot days in the summer when demand is high.
I think (correct me if I’m wrong) that the difference is that average cost is based on the current level of production/sales/service, while marginal cost looks at the cost of increasing the current level by an additional unit.
So, for example, if you were taking 60 kids on a field trip in a school bus that couldn’t (legally or practically) hold more than 60, the average cost would be the cost of renting/driving the school bus divided by 60, while the marginal cost would be the cost of renting a whole nother bus.
Increased enrollment has not resulted in an increase in the number of professors. So in reality what you are looking at is increased enrollment means professors handling more students. On average, this means that students get less attention the more enrollments are increased. The main way administrators try to handle this is by pushing online courses, which allow the professor to take more students but get away with not paying as much attention to each of them.
Expanding by increasing class size or offering online classes isn’t free. Sure, the school may not hire new faculty, and thus won’t be wasting any money on such things as a living wage for their instructors. Rather, at the very least they’re going to have to hire more adjuncts, teaching assistants, and technical support staff for the online stuff.