Monday, October 24, 2011

Vendor Bender

The lead article in today's Inside Higher Ed is about the Educause national conference held last week in Philadelphia.  The piece focused on the tension between campus CIOs and the IT industry that services Higher Ed.  I will comment on that in a bit, by doing a stylized economic analysis.  First, let me say a little about the event itself. 

I haven't attended for  a couple of years and having retired last year I don't see myself attending in the near future unless new work makes it practical to do so.   I can't say that I miss going to the conference.   I've never been a good traveler, so there's that.   The venue for the conference is usually a large convention center.  Many of the rooms have no external lighting, so if you're meeting in one of them, after a while you feel as if you're in a cave.  I would usually wear down in the afternoon as a result.  I do enjoy meeting folks from other campuses that I haven't met before.  That's a plus.  But then there is the weirdness that the tone is collegial yet in many cases much of the business would be with vendors, who indirectly would either be trying to sell you something or, if you were already buying from them, then to do things to keep you as a happy customer.  It meant the experience was different and not as much fun if there were only colleagues from other campuses present.  I do miss seeing friends and colleagues.

Let's push on to the analysis.  As an economist who has written a bit and taught in the area of Industrial Organization (a long time ago) I'm going to state some economic "facts" that in themselves should not be controversial at all.  It's in their application that the fun begins.

1)  Horizontal merger typically happens when an industry has reached a mature phase where demand growth is modest.  In the case where demand is actually declining, you then have a "war of attrition," in which case the merger or acquisition is a way for the acquired entity to exit the industry.  Industry behavior therefore is quite different when the industry is young, demand is growing briskly, and new entrants may appear on the scene. 

2) In bargaining, the power a party has in striking a good deal depends first and foremost on their ability to walk away from the deal.  Add to that how patient the party is to wait for a good deal and how risk averse the party is to negotiations breaking down.  If you know the payoff to the parties from walking away (called the threat point in game theory)  you can then predict that the gains in the bargain will be split relative to that.

3) Markets for a broad constituency tend to have several competitors.  Markets for a more narrow constituency tend to have fewer competitors.  The real test is how large a seller must be to fully exhaust economies of scale and scope and the relationship between that size and the size of the market.

Applying the above to Higher Ed and the enterprise systems that are the object of the Inside Higher Ed story, most campuses have been in these markets for 10 to 15 years.  ERP systems got a big boost in demand because of Y2K fears and the LMS as an environment took off in the late 1990s.  I think it fair to saw we're in a mature phase now, with the market fairly saturated.  Further budget issues that the campuses face because of the economic downturn mean that an increased intensity of demand on a particular campus is unlikely to manifest.  So there shouldn't be much demand growth from that source.  Merger or acquisition in the industry, then, should not be surprising. 

I suspect most campuses feel locked into these systems.  Even in the presence of multiple vendors, switching costs are high.  In choosing an enterprise system, most campuses do so for the long haul (in my estimation at least 5 years and perhaps upward of 10 years).  The threat of switching may have some impact on the bargaining power, but in game theory the question would immediately come up - is the threat credible?  Incredible threats don't have impact.  The threat is credible when a campus is in a system that is reaching end of life.  Then switching must occur. 

The article talks a lot about information sharing regarding price (really the full package) across campuses.  Absent that information sharing, there can be idiosyncratic variation in the packages campuses get.  But ask yourself whether information sharing has any impact on bargaining power.  I don't see why it should.  If you're locked in without information sharing, getting the additional information about what other campuses have done doesn't change your degree of lock in.   I can understand why CIOs want the information sharing - they bear that idiosyncratic price risk and they'd like to shed it.  But doing so should not lower the average transaction price.  Indeed, information sharing might have the same effect as "retail price maintenance" where the customers end up being the police to enforce the cartel price.

Here are some observations based on this analysis.

a)  Where Higher Ed might go with systems that are used in other verticals as an alternative to having the systems customized for Higher Ed, some rethinking should be done regarding the costs of the latter.  Those costs shouldn't be measured when the market for that system is immature.  That doesn't measure full cost.  Put another way, demanding customized solutions is like asking to be held up by the vendor(s), not immediately, but down the road.  This needs to be understood. 

b) If Higher Ed IT have become data custodians for their institutions then the cost and quality of protection depends on the extent that information is regulated by the government.   More stringent regulations translate into higher IT costs.  If you think of it this way, then from the student perspective (or the perspective of their families) they may face the choice of higher tuition and higher privacy guarantees or lower tuition but then less privacy guarantees.  The government makes these regulations with a desirable end in mind - guaranteeing student privacy - but without a firm notion of the cost to enforce the regulation.  If as in a particular service area, the sense of what that costs becomes clearer, then the tradeoff perhaps can be articulated with some degree of precision.  At that point, it should make sense to bring the discussion back into the public policy arena.  Can a sensible middle be found?  That might lesson the feeling of lock in that many CIOs seem to be feeling now. 

c)  The article also discussed forming some buyer consortium across campuses, to match the size (and therefore the power) of the vendor.  Thinking this way, one might consider what is happening now with the NBA, where the current season looks to be in jeopardy.  If bargaining power is exercised, it can lead to no transaction for a considerable period of time.  Ask, whether that, even as a veiled threat, is a possibility in this instance.  If not, it is unclear how such a consortium improves the bargaining power.  It is possible that the campuses god get better deals anyway, because dealing with the consortium lowers the vendor costs in an appreciable way - less of a sales force, easier to keep track of contracts, etc.  But, of course, there are real costs in coordinating across campuses to determine something like a consortial preference.  The potential cost saving may not be worth the real pain of coordinating. 

It may not make anyone feel better that what we are observing with the provision of these enterprise IT systems is fairly predictable, given the relevant economic theory.  But perhaps, if the CIOs together thought more about the economics, they'd get a better sense of what their alternatives really are over time.  A while ago I wrote a piece on Dis-Integrating the LMS, where the main argument for disintegration was given along pedagogical lines.  Perhaps disintegration is an approach that should be tried with other other enterprise systems, simply as defense against vendor expropriation.  It's worth thinking that through. 

No comments: