I first want to thank the Federal Reserve Bank of Chicago for the invitation
to participate in this conference, one of several on the payments system
the bank has held in recent years. Conducting research and analysis of
payments is a valuable and appropriate role for a central bank. I congratulate
the Chicago Fed for vigorously carrying out this charge.
This conference addresses what payment networks may look like in the longer
term, as well as how trends in consumer and corporate demand, technology
and economic conditions affect near-term developments. These issues are
important to the Federal Reserve, given our significant role as a provider
of retail and wholesale payments services. Moreover, and somewhat distinct
from the direct provision of services, the Federal Reserve as a central
bank has a strong interest and role in payments system developments.
While we all need to respond to prospective changes in the environment,
the track record of predictions of the future state of payments systems
is not good, to be generous. In perhaps the most recent example of unexpected
payments developments, many astute observers across the industry were
surprised by the findings of our Retail Payments Research Project; check
volumes in 2000 were appreciably lower than many had predicted and appeared
to have peaked in the mid 1990s.1
Given the difficulty in forecasting the direction that payments might
take and, hence, the particular role the Fed might play, a constructive
alternative is to articulate general principles that can help guide and
set expectations for our future activities. It is particularly important
for a public institution like the Federal Reserve to take such a principles-based
So my objectives today are twofold: first, to review some established
principles that have long helped us decide about our overall role and
about new initiatives and, second, to discuss issues involved in establishing
principles for altering our existing businesses, particularly retail payment
operations. By way of preview, I will suggest that existing principles
guiding the Federal Reserve's role in payments are valuable for delimiting
both new activities and our initiatives in wholesale payments.
But when it comes to the retail side, where the marketplace is in the
midst of substantial change, existing principles may not be fully adequate,
and accordingly we need to think long and hard about how to proceed in
this arena. The Federal Reserve has to strike an appropriate balance between
remaining an effective participant in retail payments while assuring that
the private sector has ample opportunity to innovate and to prosper as
markets evolve. There may be tension here since it is not clear how aggressive
the Federal Reserve should be, and in the second part of these comments
I will explore this tension and several aspects of its resolution. Further,
let me emphasize that my intent today is not to be definitive but rather
to begin a dialogue on this subject.
Established Principles for Federal Reserve Payments System Involvement
The Federal Reserve has long thought conceptually about its role in the
payments system. Perhaps the most well known discussion of such concepts
is the so-called White Paper which articulated roles for the Federal Reserve
in supporting the integrity and efficiency of, and access to, payments
systems.2 A number of other
efforts, including the Rivlin Committee report, also identified principles,
as does the Monetary Control Act.
Based on this work, there are two principles that seem well established
and accepted as guiding Federal Reserve payments policy. The first
is that the Federal Reserve should not provide new payments system services
unless markets are failing and the Federal Reserve has a unique comparative
advantage in providing the service.
Breakdowns in the payments system can be quite costly and some of these
costs arise because of what economists call market failures, whereby too
little or too much of a good or service is produced. In some cases of
market failure in the payments arena, the Federal Reserve may be uniquely
situated to correct the problem. While it is not the only area where market
failures might arise and where we might have a comparative advantage,
I believe Federal Reserve involvement in interbank clearing and settlement
through Fedwire provides a prime example of how we implement this first
principle.3 I think a reasonable
case can be made that our continued commitment to enhancing the functionality
of Fedwire can be justified by this same principle, namely correcting
market failure where we have a comparative advantage. Changes to Fedwire
reflect our intent to assure that a plethora of standards, communication
protocols and networks do not reduce the efficiency of the interbank market.
The Federal Reserve is well positioned to help ensure that wholesale payments
are conducted effectively and with as low a resource cost as possible.
The second principle guiding our activities is that the Federal
Reserve should utilize the full range of its tools to improve payments
system performance. Through its central bank role, including regulatory
authority, the Federal Reserve can identify and correct weaknesses in
payments system infrastructures, provide information to payments system
participants and help overcome coordination problems, among a host of
other tasks. Because many of our tools are less costly than direct service
provision and may not be available to the private sector, we may want
to favor these alternatives, where possible, to achieve broad payments
The check business provides two excellent examples of the application
of this principle: (1) the Check 21 legislation. As you know, Congress
is in the process of considering legislation, originally proposed by the
Federal Reserve in late 2001, that aims to facilitate the transformation
of paper checks into electronic files (truncation) during processing.
(2) Research. Our research on check volumes, already noted, provided
vital information that may not have been produced without our involvement.
Clearly, the Federal Reserve's payments system activities reach beyond
research and legislative changes and beyond wholesale payments. We have
significant market share in the provision of check and Automated Clearinghouse
(ACH) services. In the next few minutes, I will review: (1) some of the
binding and, by the way, appropriate limits on our current retail activities,
(2) initial thoughts that could help guide activities within those broad
limits, (3) examples of activities that are generally consistent with
these initial thoughts and (4) questions suggesting that additional work
on principles is necessary.
One binding limit on our retail operations is our commitment to remain
an active participant for the foreseeable future. It is clear to most
serious observers that mass voluntary retreat from these businesses by
the Federal Reserve is not now a serious option. This is in part because
there are well-developed expectations about banks' ready access to payments
services that would conflict with the Federal Reserve—as a public
entity—simply exiting these businesses. The large market share of
the Federal Reserve implies that leaving these existing businesses could
be highly disruptive.
A second commitment binding our actions in retail payments is that they
adhere to the established principles I earlier articulated, namely those
governing provision of new services and promoting reliance on tools beyond
service provision. These general guidelines, while circumscribing Federal
Reserve actions somewhat, do not provide sufficiently specific guidance
in the face of a dynamic market. Indeed, general limits on our actions
actually highlight the tension we face. If the Federal Reserve does not
respond effectively to changes in customer demand or competitive pressures
in existing markets, we will cease to be viable providers of payments
services and will, de facto, exit the business. On the other hand,
if the Federal Reserve responds so aggressively to competition as to eliminate
viable private sector alternatives, we will risk violating the spirit
of our established principles.
In this vein, while there may on occasion be public policy reasons for
the Federal Reserve to try to lead developments in retail payments, I
think the Rivlin Committee had it right when it said: “Private sector
innovation has been the key driving force behind the evolution of the
U.S. retail payments system and will almost certainly continue to be so
in the future.”
Establishing more precise principles to help focus our response to this
tension—that is, remaining viable without stifling private sector
initiatives—is a priority. To help narrow the range of possibilities,
I will offer a suggestion that, with additional work, might serve as the
basis for one or more well-articulated principles in the future. Specifically,
the Federal Reserve should ensure that the size, reliability and capabilities
of its basic retail infrastructure supporting established services corresponds
to market demand.
This tentative principle offers some guidance on expansion, contraction
and focus of our resource allocation. The basic infrastructure/already-
established services aspect of the principle focuses our resources on
current activities and, by implication, this is a fairly narrow focus.
The matching of size, reliability and capability to market demand implies
that as we maintain our existing services, we can expand them and/or make
investments to keep them operating in a robust and effective fashion.
More specifically, there should be room to alter our existing activities
so that a more desirable technology, for example, can be used to offer
the functional equivalent of an existing service.
The correspondence of size and demand means we also must narrow our operations
if and when demand falls. Our cost-revenue objective requires this response
in any event.
As indicated earlier, there are several ways in which our actions already
reflect these thoughts. One is that, given the realization that the Federal
Reserve's check processing infrastructure 4
is too large and too costly, we are in the process of reducing this infrastructure.
Second, for the last several years, the Federal Reserve has engaged in
a major effort to shift to a common platform to process checks. Standardization
allows us to reduce long-run costs, improve the flexibility and resilience
of our operations and meet demand from customers who operate nationally.
A third illustration is provided by the ACH, which appears to be shifting
to a more flexible system geared to a broad range of retail transactions
such as point-of-sale and lock box conversion of checks, debit card transactions
and greater frequency of one-time payments initiated by telephone and
the Web. The Federal Reserve is in the process of ensuring that our ACH
system will accommodate changes in volume and in the nature of the payments
There are several areas, however, where I recognize weaknesses with the
incipient principle that links the Federal Reserve retail payments infrastructure
to demand. For example, (1) is it tenable to have a clear division between
new products subject to the “established” principles and modifications
to existing products subject to another set? How would such a division
(2) What risks should we assume in estimating and responding to market
demand? If we count only demand as interest expressed in the form of signed
contracts, we will surely be slow in offering desired services. Our tolerance
for risk in this example would be too low. However, if we provide new
services without any real commitment from customers, we will surely get
ahead of markets. Determining the level of risk to assume and the state
of demand, therefore, are key to our viability, as they presumably are
to other providers. However, because the Federal Reserve is a public institution
with unique responsibilities and objectives, selecting an acceptable level
of risk is a complex matter.
(3) How do we balance our cost-recovery mandate with assuring ease of
access? This is an issue because covering a wide geographic region in
order to facilitate access can be quite costly, thus making our financial
targets difficult to meet. Pricing our products based on the cost of providing
the service to a single customer would facilitate, by definition, cost
recovery but could raise questions about our commitment to access.
(4) How do we respond to new private sector payments services that might
substitute for our own? The nature of our response has implications for
our ability to remain an effective direct participant in payments systems
and possibly for our commitment to address market failure as well.
Many are confident that the future payments system will differ in important
ways from that prevailing today. Knowing what the differences will be
with any confidence, however, is a different story. In the face of this
uncertainty, perhaps the best we can do is to articulate principles to
help us determine the response to future developments, a step that is
particularly important for a public institution like the Federal Reserve.
In this spirit, I have articulated two principles that are reasonably
well established and have helped the Fed determine what sorts of new payment
activity it should enter.
I have also discussed a principle to help us determine how we might modify
our existing payments operations, particularly on the retail side. Although
the principle offered seems consistent with our current actions, I recognize
that it needs additional attention. In any event, though, I am convinced
that principles such as these will be essential as the Federal Reserve
proceeds in payments.