| The
Case for Downsizing the Fed
By George J Benston and David
B Humphrey
Historical conditions that spurred
Fed involvement in retail payments processing no longer
exist, prompting another look at arguments for privatization
Federal Reserve Board chairman
Alan Greenspan in October appointed a committee of senior
Fed officials to conduct "a fundamental review" of the
Fed's role in the payment system. The committee, headed
by vice chair Alice M. Rivlin,will presumably examine
whether it still makes sense for the central bank of the
United States to be directly involved in payments processing,
both paper- and electronic-based. The Fed inquiry comes
on the heels of a report by the General Accounting Office
suggesting it is time to consider privatizing some of
the Fed's payments processing operations.
Such a review is appropriate--and long
overdue. The Federal Reserve payments processing operation
was developed to facilitate the banking transactions of
a bygone era, when federal law prevented both nationwide
banking and regional interstate branching. Decades later,
the two main problems that spurred Fed involvement in
this area no longer exist: "non par checking," or the
practice by individual banks of clearing checks at a discount
from face value; and the lack of a safe and speedy way
to settle national and inter-regional payments.
Moreover, rapidly advancing electronic
technology is rendering paper-based payments processing
an anachronism--and a costly one at that. Private interests
developed the credit card and have implemented new electronic
technology better than the Federal Reserve, which essentially
developed the automated clearing house as an electronic
substitute for checks. Tellingly, the Fed has done little
to promote the adoption of electronic check collection
or rein in the remote disbursement of checks, despite
its avowed mission of improving efficiency in check collection.
An argument can still be made for preserving
the Fed's role in handling large-value wire transfer payments,
known as Fedwire, where the average payment exceeds $4
million and the integrity of the financial markets is
at stake. It would be difficult, if not impossible, for
private interests to eliminate settlement risk and provide
the same degree of safety and liquidity for the transfer
of these large payments as is now accomplished by Fedwire.
But the central bank's responsibilities
and facilities for more mundane, labor-intensive check
processing and ACH activities could be transferred to
private hands.
Fed representatives have raised various
objections to privatization. They say it would disrupt
the payments system, unreasonably impose higher prices
for check services on small- and mid-size banks, and compromise
the Fed's ability to conduct monetary policy.
These arguments are not strong.
First, the Fed currently controls only
22% of retail payment processing, so privatization need
not be disruptive. Second, while payment service prices
would likely rise for small, mostly rural banks, this
would merely reflect the introduction of market pricing,
as those banks currently do not pay the full costs attending
the servicing of remote locations and small check volumes.
Heretofore, these small bank subsidies have been recaptured
through overcharges on large banks. The fairness of this
is questionable, given all the other advantages of rural
living. As for the effect on monetary policy, the Fed
could make up the loss of input from the regional banks
with advisory committees composed of outside bankers,
business people and academics. In fact, relieving the
Fed of its retail payment processing responsibilities
would allow the central bank to concentrate more fully
on its most vital public functions.
When announcing the appointment of
the review committee in October, chairman Greenspan stated,
"Given the significant changes occurring in payment processing,
this is the opportune time to assess the Fed's role in
the payments systems of the 21st century." One might interpret
this statement as an acknowledgment by the Fed that historical
conditions that gave rise to the Fed's payment systems
operations no longer exist. It is increasingly hard to
understand why the U.S. central bank, virtually alone
among those of developed countries, provides check processing
services to its banks.
Non-Par
Checking
The Federal Reserve was devised in
1913 as a bankers' bank devoted to regional and nationwide
check clearance. In support of this original mission,
the 12 regional Federal Reserve banks were dispersed among
the nation's major transportation and communication hubs
of that time--the crossroads upon which trade and payment
flows historically centered. An additional 36 Federal
Reserve bank branches were established along transportation
routes used to collect checks.
At the time, the United States did
not permit nationwide banking or even regional interstate
branching. Individual states prohibited banks chartered
in other states from operating banking offices within
their territory. Most states also prohibited or severely
restricted branching within their own borders. The U.S.
predictably ended up with thousands of small banks, many
possessing only a single office. Meanwhile, large, nationwide
branch networks became the norm in most other countries.
So long as check transactions were
few and confined to a bank's local domain, these restrictions
on bank geographic location and branching had little effect
on commerce. Local banks banded together and formed clearing
houses to clear and settle checks in their own areas.
Correspondent bank relationships and settlement balances
proved reasonably effective, at least initially, for handling
the small number of regional and national checks.
However, many individual banks charged
check-clearing fees by discounting them from face value,
a practice known as "non-par checking." Correspondent
banks located along major transportation routes also capitalized
on their site convenience by exacting higher discounts
than those correspondents less well situated. Anxious
to minimize these charges, banks circuitously routed many
checks away from the main thoroughfares, delaying collections
and hindering the smooth flow of commerce.
By contrast, the Canadian system, developing
about the same time, operated more efficiently. Since
Canadian banks could operate nationwide, they did not
rely as heavily on other banks to collect checks for them,
promoting speed in the payments system. Although Canada
also experienced non-par checking, circuitous routing
never became a problem because the discount from par was
uniform.
The
Cure Outlasts the Cold
In the U.S., the Federal Reserve entered
payment processing to eliminate non-par checking and provide
a speedy and secure way to settle inter-regional payments,
which obviated the need for the physical movement of gold
and the use of cumbersome, multiple, correspondent balances.
The first objective was achieved by having the Fed and
member banks clear checks at their face value. The second
was taken care of by building a nationwide wire transfer
network, which banks could not establish themselves on
account of the prohibition on crossing state lines.
Conditions have now changed dramatically.
Non-par checking disappeared long ago. Nationwide branching
will finally be possible this year. In addition, legislation
passed in the early 1980s requires the Fed to charge for
its payment services and provide them to any bank that
wants to buy them. Previously, the Fed had supplied these
services free to member banks, somewhat offsetting members'
forfeitures of income arising from requirements that they
hold non-interest bearing reserves with the central bank.
The Fed is now in the unique position
of regulating the same large banks with which it competes
in providing payment services to other banks--a potential
conflict of interest. It is reasonable to ask, as the
General Accounting Office report did, whether conditions
have improved or changed sufficiently for the Federal
Reserve to reduce its direct involvement in certain aspects
of the payment system.
The issue is not as simple as it appears,
since the U.S. payment system is composed of two distinct
markets--wholesale and retail.
The wholesale market accounts for less
than 1% of the 79 billion yearly payment items processed,
but it constitutes 86% of the $587 trillion in total dollar
value. Wholesale payments consist of large-value business
payments and financial market transactions made by banks,
both of which rely on instant and highly secure electronic
wire transfers.
Fedwire connects all banks in the U.S.
for domestic interbank payments only. The Clearing House
Interbank Payments System, or CHIPs, which is owned by
a group of large New York banks, makes both domestic and
international interbank payments, with an emphasis on
international, and directly serves only very large banks.
The Federal Reserve's share of total domestic and international
wire transfer transaction volume is 61%, leaving CHIPS
with the remainder.
The retail market is composed of smaller-value
check, credit card, debit card, and ACH payments, accounting
for 99.9% of the 79 billion payment items but only 14%
of the dollar value. The Federal Reserve's check and ACH
processing operations service about 22% of this market.
Banks and private processors handle the remaining 78%,
and they effectively process all credit and debit card
transactions.
Status
Quo Player
Overall, the case for privatizing Federal
Reserve payment operations (not including settlement)
is strongest for check and ACH processing and weakest
for large-value wire transfer. The debate regarding the
former revolves around five issues: efficiency; disruption
to the nation's payments system; the effect on small and
remote banks; the Federal Reserve's ability to conduct
monetary policy; and expectations about future demands
for payments services.
Although the Fed's current payment
operations are run efficiently, the Fed enjoys no comparative
advantage in processing checks and ACH payments. Banks
and private processors hire similarly skilled workers
and use identical processing technology. The Fed may even
be at a slight disadvantage owing to the location of its
Reserve banks, which improbably tend to be situated some
distance from airports even though air freight is the
major transportation system used to collect non-local
items. In part, this reflects political pressure to maintain
a "downtown" presence and preserve urban jobs. But clearly,
private sector providers do not have to bear similar costs.
It is ironic that the Fed entered the
payment system to promote speed and efficiency in check
collection but has done little to spur the adoption of
electronic check collection. Although technological advances
would permit the more timely electronic delivery of check
payment information, the Fed has not sought the legislative
changes necessary to implement the technology on a widespread
basis.
This advocacy is needed, for example,
to eradicate certain outdated provisions in the Uniform
Commercial Code that foster delay-producing gamesmanship
on check float.
Thanks to UCC rules written before
electronics became cost-effective, paying banks can demand
to physically inspect checks before payment. This rarely
is necessary--only for the few very large-value checks,
or where fraud is suspected. But banks often invoke these
rules solely to lengthen collection times and increase
check float--benefiting themselves and corporate customers
at the expense of efficiency. The same problem exists
with the remote disbursement of checks, a practice that
inhibits the adoption of more efficient business payments
via electronic business data interchange.
Since the Fed's current market share
of retail payment processing is relatively small at 22%,
privatization need not be overly disruptive. Indeed, it
could largely be accomplished through a change in ownership.
In addition, checks and ACH payments do not have high
average values, keeping the risks associated with processing
these payments relatively low and manageable. Such is
not the case with wire transfers, where more than $1 trillion
is transferred via Fedwire daily.
Effect
on Small and Rural Banks
It has been argued that many small
and remotely-located banks would likely face higher prices
for check services under a privatization scenario. This
is true, but not because the private sector would "take
advantage" of small banks. Rather, it is because post-privatization
pricing would more accurately reflect that fact that it
is more costly to serve banks remotely located and having
small check volumes.
Currently, the Fed partially subsidizes
payment services to smaller banks, covering the cost by
charging fees to larger, higher-volume banks. But people
who live in remote areas enjoy compensating advantages
such as less traffic, cleaner air, and lower land and
housing prices. There is no need to subsidize their use
of the payment system by underpricing their higher per-check
sorting and transportation costs.
A related concern is that a fully privatized
check payment system might operate as a cartel, excluding
some banks or treating other banks unfairly. Antitrust
laws, however, make such exclusion illegal, as is the
case for other enterprises. Additionally, a fully competitive
payments system is the best defense against unfair treatment.
Monetary
Policy
A final argument against privatization
of payment functions is that it might weaken the Fed's
ability to conduct monetary policy. The concern is that
the central bank might lose an important part of its constituency,
as well as the benefit of obtaining regional advice from
Reserve bank presidents who are members of the Open Market
Committee. To compensate for that loss, the Fed could
rely more on advisory committees of bankers, business
persons, and academics.
It is often said that the Fed's independence
from the administration and Congress, and hence its ability
to take anti-inflationary measures, is enhanced by the
inclusion of the Federal Reserve bank presidents on the
Open Market Committee. It is not clear that the Fed's
track record supports this assertion. After all, this
structure has been in place since 1913. The ensuing years
witnessed alternating periods of inflation, recession,
depression, and price stability. Regional representation
on the Open Market Committee appears to have had little
effect on the central bank's varied record in coping with
these challenges.
Indeed, one could make the reverse
case: that the Fed's ability to conduct monetary policy
and bank regulation has been compromised by the need to
manage its nationwide payment operations. Removing responsibility
for processing retail payments from the Fed would permit
it to concentrate fully on performing its vital public
functions.
Market forces are working against continued
Fed involvement in the payments system. Over time, checks
and check processing are expected to play a declining
role in retail payments. The number of checks written
per person is projected to reach a peak and then begin
falling within a few years, just as per-person use of
checks has already dropped in all other developed countries.
Electronic payments, primarily credit card transactions,
currently comprise 22% of all non-cash payments and are
forecasted to rise to between 50% and 60% in 15 years.
In addition, as banks continue to merge both regionally
and nationwide, more checks will become "on-us" items,
which do not require outside processing. Thus, even without
privatization, the Fed's role in retail payments will
naturally decline over time.
Privatizing
Reserve Banks
Changes in technology and public tastes
are notoriously difficult to predict. But experience teaches
us that privately-owned firms are likely to meet such
challenges more effectively than even well-run government
agencies. When confronted by change, the instinct of established
providers, both private and government, is often to protect
their positions and improve what they do, rather than
develop new methods. But private firms have stronger incentives
to develop better, cost-effective products--and usually
do not possess the power to restrain competitors.
We suggest that the individual Federal
Reserve banks return to their original purpose--as bankers'
banks dedicated to processing payments.
Each bank could become a separate corporation
with publicly traded stock. These corporations would own
the physical facilities of each bank and its branches
and employ the people now working there in check and ACH
payment processing. Proceeds from the sale of stock in
the individual former Federal Reserve banks would be transferred
to the taxpayers via the U.S. Treasury. The price of the
stock for each bank would be based on an appraisal of
the properties and an estimate of the income and expenses
of the banks, much as is done for initial public offerings.
The stock owned by member banks would be valued similarly.
Once privatized, some of the former
Federal Reserve banks probably would merge into more efficient
units. Some branches probably would be closed, especially
those where check volume is relatively low already.
The Federal Reserve Board, under this
plan, would continue to own its headquarters in Washington,
D.C. and be responsible for monetary policy, maintaining
banks' reserves, and the settlement of payments among
banks. It would also be responsible for meeting currency
demands and supervising the banking system as mandated
by law. Bank examiners and vault cash personnel could
rent space in the buildings of the now privately-owned
Fed banks. The Federal Reserve Board could continue to
collect and process regional economic information and
make economic analyses available to banks and others,
either at offices rented from the former Federal Reserve
banks or elsewhere. Fedwire would continue within the
Fed's purview, reflecting the importance of this network
for business payments and financial markets, as well as
the risks involved in the very large payment values transferred
daily.
The U.S. financial system stands at
a turning point. Within a year, the United States will
join the rest of the world in permitting banks to operate
on a nationwide basis. Technology is rapidly changing
the way payments are made. The old system of having retail
payments processed by the central bank is no longer necessary
to meet the needs of today's dynamic financial services
industry. Now may well be the time to return the retail
payments system to private hands.
Mr. Benston
is John H. Harland professor of finance, accounting and
economics at the Goizueta Business School of Emory University.
Mr. Humphrey is a professor of finance and F.W. Smith
Eminent Scholar in Banking at Florida State University.
Copyright © 2003 by Banking
Strategies, published by BAI.
back
to top |
