The Institutional Dynamic: Understand First, Act Second—If At All
I have now had a chance to review the excellent posts on the second day, all of which have a common flavor. They expand the universe of relative considerations that need to be taken into account to decide whether imposing caps on interchange fees enhances or reduces overall social welfare. The narrow perspective on this issue, which is difficult enough, is to master the dynamics of two-sided markets to figure out where the fixed costs of running the overall system should be allocated.
That model assumes that the credit card business operates in isolation from all other payment systems present and future. Its effort is to run an efficient allocation of costs in the face of the famous marginal cost controversy that dates back to the 1940s. Unless there is some outside subsidy all relevant players cannot be allowed to pay only marginal cost. Yet to put in the subsidy is to create a tax distortion in some unrelated market whose welfare consequences are virtually impossible to track, given the difficulty that arises in discovery of the incidence of the tax as it works its way through the economy. We are therefore necessarily in the world of second-best even on the simplest possible analysis.
Unfortunately, that simple analysis leaves a lot out of the equation. One key issue is the competition that credit cards have with noncredit card systems, each of which may have built in distortions of their own. We know that the United States has to print and police the use of cash. We also know that it can disappear from company coffers into the hands of dishonest employees. It can be lost. It can be stolen. It can get waterlogged. It can be deposited in the wrong account by accident. Credit cards reduce these costs, and they do so arguably in a more efficient form than the use of checks, which of course clear at par, which means that the cost of interchange is borne by general tax revenues, with the usual set of static distortions. It also creates dynamic distortions because the want of price signals between the players makes it harder to introduce innovation into that space on such critical matters as error control, even if it would result in higher level of reliability in transactions.
In addition, it is also important to consider the other benefits that can arise with the use of credit card payments, one of which is the ability to key in all relevant data from a transaction at once. Quite simply it may well be easier to link in inventory control, for example, with a credit card system than it is with a cash or checking system. And it may speed up the rate of transactions so as to reduce the length of queues that are so important in many retail operations.
The clear upshot is that it is difficult through informed speculation to identify all the collateral consequences of running a credit card system, both positive and negative. The only sure piece of data that we have is that credit transactions have done far better than cash and checks, even if they are losing ground to the next generation of payment systems that rely on cell phones and other technologies that are untied to the now ubiquitous magnetic strip. These dynamic changes could easily force down interchange prices without the need for administrative proceedings.
The hard institutional question therefore is why concentrate major reforms on the interchange fees when all these other components must be added into the mix. On this question, priors really matter. And after reading the assembled posts, my own view is that technological innovation is a far more important driver of improvements than partial fine-tuning of the current system, whatever its flaws.
In one sense, therefore, we, the members of this blog-fest, may well be part of the problem. By putting one part of a complex payment industry under a microscope we divert resources from cost reduction measures that have unambiguously positive effects. How large a cost is this? Frankly, no one knows. But given the risks of error in implementation, the best response still seems to be, play for the next big breakthrough, and in the short run, leave well enough alone.