What drives increases in University fees? Bennett hypothesis vs Baumol’s cost disease

Over the last 15 years, increases in higher education fees have accelerated and are now rising faster than any other part of the economy.  Outstripping even the rising cost of medicine and health care.  And yet we see no meaningful improvements in productivity or GDP growth over the same time period.  Since 1978, the cost of higher education has increased in the US by 1,120% – more than 11 times.  This graph from Bloomberg clearly illustrates the anomaly of higher education fee inflation.

 Bloomberg tuition (s)Source: Bloomberg, Data: Bloomberg Labor Department

 What’s driving higher education fees higher? 

What we see in the wider economy does not justify an explanation that fees are increasing in line with improved earnings expectations.  There is some improvement in earnings overall but productivity and GDP are not rising anywhere near as fast to justify this optimism.  If we had a matching 1,120% increase in productivity or GDP over the last 30 years the world economy would be a lot more rosier place than it is at the moment. And most of the 1,120% increase has happened in the last 10 years when productivity and GDP growth has been stagnant.

Instead, graduates have better outcomes compared to non-graduates as a result of job-signalling/screening effects.  This is where employers use education credentials as a way to screen for desirable behavioural traits (e.g. self-control) or social background.  You need a university degree to manage a retail shop because a university degree signals that you are reliable not because a university degree actually helps you do your job.  In the past, completing year 12 high school would have been sufficient to signal reliability.  This leads to an arms race of over-education.  Education as a signal is a lot like the Red Queen effect, you have to run to stay in the same place and run even faster to actually go somewhere.

Signalling dynamics are common to a lot of markets and while goods and services linked to prestige can be expensive it is rare to see prices increase to the same extent as we have seen with university /college fees.

Bennett Hypothesis vs. Baumol’s Cost Disease

So why are fees increasing so much?  There are two competing theories.  The first is the Bennett Hypothesis (‘Our Greedy Colleges’, 1987): increases in university fees are directly related to increases in financial aid.  That is, any loosening in individual budget constraints through the availability of financial aid (government or philanthropic) will  feed into higher prices.  The second is the Baumol’s Cost Disease hypothesis: that increases in fees are related to the cost of low growth in teaching productivity compared to the wider economy.  Where over time, the cost of maintaining service levels becomes increasingly more costly as salaries increase in line with the broader economy.   A good explanation of how Baumol’s Cost Disease works is given in this Washington Post article ‘The Tuition is Too Damn High’.

Which hypothesis is correct?  In a recent NBER paper ‘Accounting for the rise in College Tuition’, modelling suggests that the Bennett Hypothesis is the key driver explaining rises in university fees and not Baumol’s Cost Disease.  Baumol’s Cost Disease is shown to actually lead to lower fees not increase fees.

This modelling has two key policy implications:

  1. Any increase in student financial aid will directly feed into higher university and college fees.
  2. Corresponding increases in university costs can not be explained by low productivity growth in delivery of teaching.

Bowen’s Revenue Theory of Cost

If Baumol’s Cost Disease is not a justification for higher fees, why are university costs so high?

The answer maybe found in Bowen’s Revenue Theory of Cost.  When an institution’s dominant goal is prestige (excellence and influence) there is virtually no limit to the amount of money it will spend to achieve this goal.  The institution will raise as much money as it can and spend all that it raises.  Consequently as spending builds prestige and prestige raises more fees to spend, the end result is ever increasing expenditure.

This is in part explains why universities will always be willing to support research even without direct government subsidies. Research excellence has a strong influence on prestige/reputation.  (I’ve blog about this here.)

There has been a lot criticism of universities cross-subsidizing research out of teaching funding – but what are the alternatives?  If not research,  universities will build prestige by spending up big on marketing their ‘brand’, scholarships for the wealthy to build influence, and nice big fancy buildings of little productive value.

All of these alternatives are at the expense of good teaching quality.  But at least with research students get access to the very forefront of knowledge creation.


Meso Soup has a discussion of this blog post here.

%d bloggers like this: