Editor's note: These remarks were delivered to a conference sponsored by the Federal Reserve Bank of Philadelphia, in Philadelphia, Pa., on Sept. 29, 2005.
It's a distinct pleasure to join you here this evening, and I thank Tony Santomero and Peter Burns for the invitation to participate in this conference on "Recent Developments in Consumer Payments and Credit." Looking at the agenda and papers, I'm struck by the breadth and depth of the payments-related research to be discussed tomorrow. While economists have long been interested in payments-related issues, there seems a renaissance of sorts under way on the topic, a welcome development for the profession and, more selfishly, for the Federal Reserve which, as the agenda for the conference demonstrates, is both a producer and a consumer of such research.
But how precisely might the Federal Reserve make use of
payments-related research? The short answer is that, potentially, this work can have significant implications for the role of the Federal Reserve in the financial services marketplace, and I will focus my remarks on that topic.
Specifically, I will make three points.
- The Federal Reserve, as a general matter, decides which payment services to provide and how to provide them based on economic criteria which emphasize concepts such as competition and
efficiency, and which focus on societal or public benefits; traditional business rationales such as profit maximization play a secondary role.
- Getting the economics "right" is particularly important today as the Federal Reserve makes significant changes in the scale, range of products, pricing and underlying technology of retail payments offerings in particular.
- Economists can make a significant contribution by bringing the developing and related economics of networks and payments to bear on decisions confronting the Federal Reserve.
It is particularly fitting that I focus my talk on this topic at the Federal Reserve Bank of Philadelphia. The Philadelphia Fed is clearly a leader both through the research sponsored, encouraged and distributed by the Payments Card Center and through the leadership of President Santomero on the various Federal Reserve governing bodies that oversee our financial services businesses.
Role of research
As I think you know, the Federal Reserve is an institution which takes economic research seriously, in large part because research has made significant contributions to the policy objectives for which the Federal Reserve has direct responsibility. In at least some cases, considerable value has arisen when this research has challenged the status quo, the conventional wisdom.
Monetary policy provides a number of truly exciting examples in which economic research has had a significant impact on Fed policy over time. While I will not review the material in detail this evening, I will note by way of illustration that, at the Minneapolis bank, for example, we have long supported the research of scholars, perhaps most notably Ed Prescott, advancing theories like rational expectations, time consistency and the real business cycle. Some of this analysis is now part of the mainstream, while clearly other ideas associated with such research have not been accepted widely within the Federal Reserve or the profession. Nevertheless, I know that even in these cases such ideas influence our discussions and are taken seriously by policymakers and staff.
Research of value is certainly not confined to macro-money, but this area should serve as a model for what is possible in terms of rigor and substance
and influence in Federal Reserve policy deliberations. Accordingly, I want to spend some time on the role of economic research in decisions affecting Federal Reserve activities in financial services but, before doing so, let me briefly describe how economics is currently incorporated into the decision framework governing Federal Reserve payments. And let me remind you of the obligatory caveat that I am speaking only for myself and not for others in the Federal Reserve.
In a nutshell, the broad mission of the Federal Reserve in payments, as expressed most succinctly in a 1990 white paper, is to foster the efficiency, accessibility and integrity of the system.1 These characteristics—efficiency, accessibility, integrity—embody significant economic components pertaining to competitive markets, to the network features of payments systems, to the public good aspects of payments. Conceptually, these features provide the Federal Reserve with a wide range of justifications to undertake payments activity, but in practice they have led us to be quite reluctant both to enter new lines of business and, in most cases, to alter existing lines. Indeed, I have long argued that these characteristics imply major emphasis on market failure—lack of competition or underinvestment in security and reliability, for example—as a justification for Fed involvement in payments.
Not surprisingly, we are the only payments provider making the supply of our services contingent on market failure rather than profitability. And that is appropriate, given the distinct role that the Fed plays in payments relative to private-sector providers; if the Fed is "just another provider," we should exit the business.
The characteristics I have pointed to—efficiency, accessibility, integrity—although helpful, are in practice excessively general to provide much guidance to those being asked to run the Fed's payments operations. So when I took on oversight responsibility for the Fed's payments business, I tried to articulate something more specific and something which continues the tradition of looking to economics rather than to profit/loss to guide our direct provision of payments. We settled on the following: The Federal Reserve should ensure that the size, reliability and capabilities of its basic retail infrastructure supporting established services correspond to market demand. Most importantly, where demand for our services is diminishing, we should reduce our provision. This hardly sounds revolutionary, I know, but government has not always been willing to shrink or terminate
long-standing programs. On the other hand, where demand is increasing and others cannot or will not meet it, we should expand our activity.
I should note that this general principle was significantly influenced by thinking on the Fed's role in payments by economists Ed Green and Dick Todd of the Minneapolis Fed (as you may know, Ed is now at Penn State).2 Because I want to return to some of the issues that Green/Todd raised, let me briefly sketch their thinking.
Green/Todd argued that a central bank may well have a comparative advantage in offering bank-to-bank payments processing, a service which in fact the Fed offers through Fedwire funds. And they argued that the Fed should offer retail payments only in cases where markets fail, where economies of scope exist between wholesale and retail payments or where government interventions other than provision of payment services (for example, regulation) are demonstrably inferior to provision. But given that it may not always be clear if these criteria are satisfied, and given the potential disruption that might result when the Fed leaves a payments market, they argued that narrowing the Fed's retail business should occur largely through attrition rather than through formal exit.
Of course, it is not sufficient to articulate principles that have a significant economics component if we do not follow through. So has economics, in fact, played a role in decisions the Fed has made in the retail business? I believe so, and let me now turn to this topic.
The role of economics in
Federal Reserve decisions
During my tenure on the oversight body for the Federal Reserve's payments business, we have focused in part on reducing the size of our check processing infrastructure. By way of context, this is by far our largest business in terms of staff involved and of financial services revenue. Federal Reserve check processing volume recently has been falling on the order of 10 percent or more a year, largely reflecting industry trends. In fact, checks now for the first time make up a minority of noncash retail payments, and their use is expected to continue to decline.
Our response to these circumstances has not been to hold the line on capacity and try to take market share from other service providers who do not have our scale, nor to seek subsidies for this business. Instead we have reduced and rationalized our processing infrastructure. In part, this adjustment stems from our assessment that the check processing market does not appear to suffer from demonstrable market failure, nor are there clear economies of scope between check processing and our wholesale business. Before the end of next year, we will have reduced the number of Federal Reserve processing sites from 45 to 22; the preponderance of this contraction has already occurred.
To ensure that we fully recover our costs—and we are on track to do so in 2005 for the first time in several years—we are not only reducing expenses associated with check processing but also raising prices. Other things equal, we expect these price changes to hasten the decline in check volume and to spur the move to electronic alternatives.
Indeed, in an effort to reduce a barrier to the shift to electronic processing and achieve associated saving of society's resources, the Federal Reserve initiated the so-called Check 21 Act. This legislation, passed by Congress in 2003, allows a bank to electronically process a check and, if asked, print a paper copy at a location closer to the paying institution, thereby reducing transportation costs and presentment times. Equally important, if the banks involved in the transaction agree, no paper has to be presented at all. The Fed is providing services to support this type of check processing, but these services are in fact part of a transitional program. The long-run contribution of Check 21 is likely to be a fundamental reengineering of a major part of retail payments and a much diminished operational role for the Federal Reserve.
While nothing quite so fundamental is under way in Automated Clearing House services, here too developments illustrate the Federal Reserve's adherence to economic principles. The volume of ACH transactions has grown steadily and appreciably in recent years, while the number of processors has fallen. Over the last several years, the ACH market has been reduced to two national processors, the Electronic Payments Network and the Federal Reserve, with the Fed starting with a dominant market share. However, we have not acted like a monopolist. During this period, we have made investments to ensure that we process increasing volume with no diminution in quality. And, at the same time, we have been steadily reducing prices. Nevertheless, Federal Reserve market share is falling and is now less than two-thirds of the market (down from something like 80 percent or so several years ago).
A question I sometimes get is, Why shouldn't the Fed exit this business, given that an effective private-sector competitor exists? What is the market failure that justifies the Fed's role in ACH payments, and how is involvement consistent with the economic principles articulated earlier? In response, it seems clear that the ACH market is not the check market, with its more classic, fully competitive aspects. Thus, it is not obvious that our exit would be welfare-improving. It is conceivable that we could choose to leave the ACH business but, if we did, I would assume that we would face pressure to take on an expanded regulatory role. But would that role lead to a superior outcome compared with remaining a significant service provider? I have thought about this issue but will admit that so far the answer is elusive. Thus, I would benefit from the thoughts of others and, to broaden this out, let me now identify other areas where analytical contributions would be valued.
of economic research
I've tried to note where economics is already playing a role in the Fed's payments operations, both in theory and in practice. But in reality, this discussion raises at least as many questions as it answers, so let me turn to some of them now.
Earlier I referred to the work of Green/Todd. There is room for
follow-up to their analysis in several ways. Specifically, how would the Federal Reserve determine if there are appreciable economies of scope between its wholesale and retail businesses? Given that the same accounts (and accounting relationships across the banking system) that clear Fedwire funds transactions also clear ACH, the answer is not as obvious as it may seem.
Green and Todd put much emphasis on contestability as a means to ensure that competition occurs in markets with few actual providers. What evidence should we rely on to determine if a market is contestable and if contestability is generating a competitive outcome? In this regard, community banks frequently express concern about being at the mercy of large correspondents if the Federal Reserve exits a market although, as my colleague Jeff Lacker has pointed out, anyone with a car can enter the check courier business. How will the shift to electronics influence the issue of contestability, particularly for smaller banks? Are there specific steps the Fed should take to make markets more contestable?
Green and Todd discuss alternative roles that the Fed can choose, besides direct provision of services, which could have welfare-enhancing outcomes, but the "could" here is largely undefined. One option is to play a larger role in standards setting. Right now the Federal Reserve largely restricts involvement in standards to payments areas in which we have a significant presence. Should we expand that role, and in what way? Should the Fed bring together those parties interested in establishing standards? Should we set ourselves up as an alternative to them?
In another arena, some have called on the Fed to take a regulatory role in terms of certain credit card fees—interchange fees, more specifically. I'm aware of some of the literature, supporting both sides in this debate as it happens. At a minimum, I wonder if there are tractable, cost-effective ways of making current models of the market and interchange fees take on more institutional detail and therefore become more directly applicable?
Another area of interest concerns the future of the check processing market. Although check writing is declining, I fully expect many billions of items to be processed annually for some time to come. Yet, it is possible that the vast majority of these items will be converted to an electronic form very soon after being written. How will more extensive electronics affect the market structure of check processing? Will it look like the ACH network? If so, is there still a Fed role, and what would it be?
The Fed's payments business is not identical to that of private-sector competitors; the goal is to run our payments operations to increase social welfare. Economics is critical in helping to determine if and how we can improve welfare. I've tried to provide some context for our current operations and the role that economics has played in decisions concerning them, especially on the retail side of the business. I've also raised some issues and questions that the Fed may have to address as our payments business and responsibilities evolve. Hopefully, there is room for engagement on these matters at this conference and at future venues as well.
See a summary of the white paper "The Federal Reserve in the Payments System
2 See the 2000 Annual Report essay and the interview with Ed Green in this issue.