Thursday, June 01, 2017


Tech and Costs

Josh Kim, following Jeff Selingo, asked yesterday why higher ed’s adoption of technology hasn’t cut costs.  For present purposes, I’ll focus on cost of provision, rather than price.  (Subsidies affect price but not cost; when the subsidies are frozen or cut, price goes up even if cost is flat.)  And I’ll focus on the community college sector, where the “amenities arms race,” lazy rivers, climbing walls, “administrative bloat,” and the like are mostly irrelevant.

At a really basic level, on my own campus, 80 percent of the operating budget is labor.  (And that’s with carrying a substantial, if not unusual, adjunct percentage.)  When tech doesn’t replace labor, it leaves the major driver of cost in place.  As long as the core enterprise of teaching remains substantially the same, tech will matter only on the margins.

The difference with, say, manufacturing is dramatic.  Part of what has driven “onshoring” of manufacturing has been the much lighter burden of labor costs that new tech has enabled.  If a new machine enables a worker to make twice as many widgets in an hour, and the demand for widgets doesn’t change, the company can reduce its production workforce by half.  We don’t have that option.

In a typical business, the judgment about adopting technology is (supposed to be) based on expected return on investment.  (Sometimes it comes down to the very human fascination with shiny new toys, but the basic argument stands.)  A business will invest in tech if it thinks the tech will pay off.  If it thinks it won’t pay off, it won’t invest.  

That’s not how it works here.  We have to upgrade technology to prepare students for the workplaces they’ll encounter.  That means buying tech not because it saves us money, but because the world for which we’re preparing students is more tech-heavy than it used to be.  If we continued to run our computer labs on Windows XP, our students would be in rough shape when they hit the real world.  

The expense really shows in specialized fields.  We just opened a new allied health building with some pretty incredible simulation tech for nursing.  The gain to the program is in quality; students can make mistakes on simulators that we couldn’t ethically allow them to make on real people.  The mistakes are the key learning moments.  We have rooms that look like hospital rooms, complete with much of the expensive equipment you’d see in hospital rooms.  Even when the equipment is funded by Perkins grants, of which I’m a fan, it still needs to be maintained.  And it doesn’t cut costs at all.  To the trained eye, we’re doing a better job of fulfilling our mission.  To the untrained eye, it looks like we lack “discipline.”

The same is true in Graphic Design, or Photography.  The fields still exist, but the tools of the trade have become more specific, high-tech, expensive, and short-lived.  Drafting tables could last for decades, but software has to be upgraded every couple of years (as does the hardware to keep up with it).  Old-school darkroom equipment could last a decade or more, but digital photo equipment is changing every year.  For professionals in the field, that may pay for itself.  For educators, it’s overhead.

Yes, sometimes we find some labor savings on the margins.  But they don’t come anywhere close to the kind of savings that private businesses can realize, given our different purpose.  

This may sound like special pleading, but it’s easy enough to verify.  The factors I’m pointing out pervade the sector.  Discipline varies from campus to campus and president to president.  Are cost pressures sector-wide, or specific to certain campuses?


Given the right direction, there are cases in which tech can help cut cost.  Open Educational Resources are a clear case, though there the savings accrue to the student rather than to the institution.  Online teaching can cut transportation costs, though again, the savings there accrue to the student, rather than the institution.  From an institutional perspective, given the mission, it’s difficult to realize significant, consistent savings with technology because of our mission.  

Efficiency is great, but we’ve seen what happens when colleges are run like businesses.  Until and unless the basic labor-intensity of the field changes, and as long as the fields for which we prepare students continue to get more technologically advanced, it will continue to be difficult to get real costs down.  The issue isn’t “discipline,” but structure.  The fix will have to be structural, too.

It is true that higher education is still rather labor-intensive, and that Baumol’s cost disease might be a good description as to why increased productivity has not driven down the cost of education. It still requires one instructor to teach 20 kids, just as it did a century ago. Not much productivity increase there. Even though a large fraction of the teaching function has been adjuncted out, the overall cost of education seems to be steadily rising, even faster than the cost of medical care.

I think that a major cost driver in education is the cost of administration. Nowadays, colleges and universities have to show compliance with a long list of government-imposed rules and regulations, most of which are unfunded mandates. This requires the hiring of lots of managers and extra staff, just to show compliance. The requirements imposed by the accrediting agencies mean that a lot of extra effort has to be expended, just to make sure that your accreditation is not yanked. The outcomes assessment fad in particular requires the hiring of a whole bunch of directors, assistant directors, plus a lot of front- and back-office staff. The handling of financial aid also requires additional staff. It is small wonder that college and university administrations have a whole bunch of deans, associate deans, and directors of this and that, all of which are bringing down high salaries, driving up the cost of education. It is not uncommon for the president, the football coach, plus certain high-level administrators as well as faculty superstars who are bringing in tons of grant support money, all to be knocking down salaries in excess of a million dollars a year. The cost savings supposedly offered by the adjuncting-out of the teaching function has not nearly been able to offset the rising cost of administration.

But I suspect that in the next 20 years or so, automation may be able to virtually eliminate the teaching function of the faculty. Most of the courses will be offered online by computer software, and the only faculty members who will still be doing any teaching at all will be superstars teaching thousands of students via MOOCs. About the only places that will be able to offer students a real live teacher in a bricks-and-mortar classroom will be the super-selective liberal arts colleges such as Smith, Swarthmore, and Vassar, as well as the research-intensive universities such as Harvard, Princeton, and the University of Chicago. All of the students forced to attend less-prestigious institutions will be reduced to sitting in front of computer terminals and being watched over by part-timers who have been reduced to the status of teaching assistants.

Not a pleasing prospect. I can't see why anyone would want to seek a career in education nowadays.

Tech savants fail to understand that teaching is one of the areas of human endeavor subject to Moravec's Paradox. Computers can calculate fast, they are good at that where humans are not. Human teachers can teach humans, they are good at that where computers fail utterly.

MOOCs were a fantasy. If they could work to teach, then the books of the Gutenberg revolution would have worked: the radio classrooms of the 30s would have supplanted teachers already: television would have killed the teaching star.
None of that happened, so I don't see how another technology is going to differ.
Hand-waving arguments about AI in MOOCs will not be entertained until some evidence is available.

By way of analogy, these figures from LinkedIn:
Staff Software Engineer - $168,000
Director of Engineering - $164,000
Notice the director gets less than the staff, because the staff are highly-trained professionals doing a job that few people can. This holds true for teaching - administrators are doing a job that lots of people could do well, teachers are highly-trained professionals that need both training and experience to do their jobs well.
Hey, the invention of the microphone made it possible for one PhD in Psychology to teach 600 people at the same time. I was one of the 20 he was teaching (in the 2nd row); the others were just along for the ride or the crossword puzzles. That beat Baumol to death, but it was decided that we should go backwards from there and teach everyone in a more effective fashion. In the next model, the teaching was done by TAs in small breakout sessions. They were cheaper than adjuncts.

But seriously, what bothers me about the tech investments at my college is that I don't think anyone has ever asked if they are a cost effective way of increasing success, or if they are even being used. We have SmartBoards everywhere, but many teachers just use them as screens. I know a few faculty who use them interactively, but that seems the exception apart from some lower-level math classes where they give a scripted set of presentation slides to the adjunct army.

On the plus side, putting that money into class sizes or rewarding top adjuncts is expense rather than capital. We bought all of our classroom tech before the post-recession 30% budget cut. The "people" approach to success would have had to be reversed, but capital just sits there until we can't afford to fix it. It might have paid off, assuming we are getting better grad rates as a result.

I take the opposite view to ArtMathProf on one part of administration. What irks me is that we (all colleges and universities) have done such a bad job of automating "accountability". Everything is a one off. I think it was desperation that led to a recent change toward automating our processes: Our budget has no room for more IR staff or to pay faculty to do some of that work as an extra load, and the "send it all in" mode was a failure. I imagine their offices look like that of the bureaucrat in "Ikiru". Check it out:

The solution was a "one size fits no one" approach where you stuff everything into a pre-set rubric and they suck it out the other end. The only error, and it was a big one, was not telling people that this is what they HAD to do and develop some buy-in and, later, solicit improvements in the process. From my perch, it looks like this could be completely automated, right down to generating the tables and reports. That would be a great investment. We could probably sell it. Of course, no one cares what the faculty think. They would have to hear it from a consultant to take it seriously.

PS: We got a lot of funding for our new nursing simulators from physicians. Those donations generate good will on both sides and are a win for both the nurses and the hospital or physician group that hires them.
It seems that in many cases, the adoption of high technology has actually resulted in higher costs rather than in enhanced productivity and lower costs.

It used to be that all a college classroom needed was chalk, a blackboard, plus a bunch of desks. Now a classroom needs a super-expensive DLP projector, a computer terminal for every student, plus Wi-Fi access for every student.

Just about every major (especially those related to job training) requires access to highly-sophisticated computer software tools. Colleges and universities have to offer their students training in these high tech software tools because the job market requires them and also because other competing schools offer them. This places the college or university at the mercy of monopolistic software companies like Microsoft or Adobe, who charge astronomical fees and who introduce costly upgrades and eliminate support of legacy versions on a regular basis.

Course management software, such as that provided by Pearson, is highly expensive. The cost savings promised by the replacement of faculty with this software are often eaten up by the rising costs of the software, plus the cost of training people in how to use these tools. The cost-savings supposedly promised by the adoption of course-management software can rapidly disappear.

It seems that the adoption of all of this high-tech software does not really increase anybody’s productivity, but simply adds to the overall cost of a college education.

bIt is well known that the cleanliness of houses is one of the most important services necessary in cleaning and upgrading our homes to the modern homes, which have been cleaned modern, which would provide the living conditions of the neighborhood and the quality of a competent body in the field with the provision of guarantees that would ensure the full guarantee of the success of the cleaning service There is no need for more money to be lost through requests for house cleaning services that do not meet their needs and are not in line with their wishes because many home cleaning companies are only seeking to offer their services through promotional methods in order to make money only without taking into consideration the ethical and professional conscience As these companies seek to satisfy customers, nothing more than this is accomplished by mastering house cleaning شركة نقل عفش بالرياض
شركة تسليك مجاري بالرياض
شركة تنظيف منازل بالرياض
شركة تنظيف بالرياض
شركة تنظيف خزانات بالرياض

This comment has been removed by the author.
Post a Comment

<< Home

This page is powered by Blogger. Isn't yours?