The Federal Reserve System was designed so that each of the Reserve
district banks would play a significant role in regional economic stability.
Subsequent changes in legislation have centralized control of monetary
policy in the Board of Governors and Federal Open Market Committee.
Public interest in the System has naturally followed the policy power
shift to the Board in Washington, D.C.
Interest today in the Reserve district banks is often limited
to questions such as: Why are there two Federal Reserve banks in the state
of Missouri? And why are there so many Reserve bank cities in the East
vs. the West? Answers to these location questions are usually attributed
The charge that politics played a role in locating the Reserve
district banks was made by many, including one of the architects of the
original Federal Reserve Act, Henry Parker Willis, who commented in a
1923 work: "In none of the ... surveys of the situation was ... a bank
at Richmond ... ever seriously considered, although that city, like others,
had early engaged special counsel of various kinds and Comptroller Williams
was well known to be doing his utmost to forward its claims." Willis adds,
"... only a casual survey was necessary to make it clear that there was
a serious error in the attempt to insert a Richmond district. ... There
was very considerable ground for question as to the wisdom shown in the
selection of Cleveland, the home of Secretary of War Baker. ... Atlanta
might have been omitted ... but the influence of Senator Hoke Smith was
generally regarded as having turned the scale in favor of ... Atlanta;
indeed, there was a prevailing impression that the designation of Atlanta
was part of a ... political understanding." 1
But are these claims disingenuous? Other than Willis' commentary,
very little exists on the processes and procedures used in making the
location decisions by those involved at the time. The goal here is to
explain the process leading to the location of the Federal Reserve district
banks, including the very central role played by Henry Parker Willis and
focusing, in large part, on review of an unpublished 1914 document written
by Willis. This document, we will see, not only provides insight into
Willis' thinking at the time key decisions were made about the nascent
Federal Reserve, but also somewhat contradicts statements Willis would
make in the 1923 work cited in the previous paragraph.
The Federal Reserve Act of 1913
The Federal Reserve Act of 1913 created the Federal Reserve System. Distrust
of private banking interests and the strength of American populismembodied
in the political power of William Jennings Bryan, President Woodrow Wilson's
Secretary of State during the Act's drafting and passagewas counterpoised
against the fear, expressed by private financial interests, of the federal
government becoming involved in monetary and financial markets. The Act
was a compromise between these interests, reflecting the attempt to balance
private interests with federal government assistance, protection and oversight. 2
Attempting to provide economic stability, the Act called
for "not less than eight nor more than twelve" Federal Reserve district
banks. This system would provide an elastic currency, regulate interest
rates, supervise and regulate, and act as a clearinghouse for member commercial
banks within their district, wrote economist Richard Timberlake. The Reserve
banks were to be spread throughout the country, with populists proposing
one Reserve bank for each state, plus the District of Columbia and perhaps
Alaska, according to Timberlake.
The House version of the Act reflected the populist leaning
and provided for "not less than twelve" Reserve cities. The Senate's version
of the Act called for between eight and 12 district banks. The final range
was arrived at in legislative conference with the Senate's language prevailing.
Coordination and management of the System was provided
for by a Federal Reserve Board, consisting of the Secretary of the Treasury,
the Comptroller of the Currency, members ex officio and five members appointed
by the president and confirmed by the Senate, sited in Washington, D.C. 3
The Reserve Bank Organization Committee
The Federal Reserve Act delegated the actual districting and designation
of Federal Reserve cities to the Reserve Bank Organization Committee,
consisting of the Secretary of the Treasury, the Secretary of Agriculture
and the Comptroller of the Currency. The Reserve Bank Organization Committee
was also charged with setting in place an operational plan for the Federal
Locating the Reserve cities and districts was a contentious
The members of the committeeWilliam McAdoo (Treasury), David F.
Houston (Agriculture) and John Skelton Williams (Comptroller of the Currency)met
Dec. 26, 1913, to plan their schedule.5 They
set an ambitious itinerary of cross-country travel, starting in New York
City Jan. 5, 1914, and ending in Cleveland Feb. 17, 1914, to take testimony
on the location of district cities and district boundaries. This resulted
in upwards of 5,000 pages of testimony taken in 18 cities from over 300
individuals. A copy of this testimony is kept in the Federal Reserve Board's
library in Washington, D.C.
Simultaneously, the Treasury Department sent ballots to
the 7,471 nationally chartered commercial banks to assess their preferences
as to which city they wanted as their Reserve district headquarters. The
ballot allowed for the naming of first, second and third choice. The ballots
were gathered prior to designation of any district boundaries, so banks
were unconstrained in their choice of cities. Comments by committee members
in the various cities indicate that they had access to the results of
the balloting prior to both the completion of their tour and their deliberations,
which occurred between Feb. 17, 1914, and April 2, 1914.
Other than specifying the number of Reserve districts, the
Act gave little instruction to the committee on districting. The Act stated
that: "... the districts shall be apportioned with due regard to the convenience
and customary course of business and shall not necessarily be coterminous
with any State or States." 6
Henry Parker Willis and the Preliminary Committee on Organization
Early in January 1914, Secretary McAdoo, on behalf of the committee,
appointed Henry Parker Willis as chairman of the Preliminary Committee
on Organization.7 Willis' role was more significant
than just chairing the preliminary committee. As Willis writes: "Secretary
McAdoo assigned to the author individually the duty of preparing a districting
plan ... which should afford the basis for the [Reserve Bank Organization]
Committee's final work. At the various hearings ... reports were taken
... and these with the various briefs and documents filed by witnesses
were reviewed by him and ... used as the basis for a report in which plans
were laid down for drawing the district lines." 8
The Willis report suggests business flow, transportation,
district capital and factors other than political opportunism for the
number and location of Reserve districts and district cities.
Willis and the preliminary committee submitted their report,
which included their thoughts on districting in early 1914. However, the
section on districting in this report is general in scope and limited
In its report, the preliminary committee provided the general
principles that they felt should guide districting. These principles followed
the broad guidelines in the Act. According to Willis in 1914, the fundamental
principles of districting should be:
(a) Eight to 12 districts, the choice of the number within these
limits entirely open and to be decided without prejudice.
(b) The Act intended to establish a number of different and independent
institutions, each sufficiently strong to care for itself.
(c) The institutions should be reasonably similar to one another
in size, located as will most nearly convenience the business of
(d) The creation of any one large bank should be avoided. No bank
should be so preponderating in importance as to make it ipso facto
the most conspicuous and by far the strongest element in the System.
Creation of two distinct classes of banks, one consisting of large,
powerful institutions, the other consisting of smaller and weaker
institutions should be avoided.
(e) While the law requires that a minimum capital of $4 million
shall be present in each and every Reserve district, there is no
harm in approaching closely to it or even in going below this limit
if in the near future growth will provide the necessary capital.
(f) Special study should be given both in establishing the districts
and in establishing the point in each district where the headquarters
bank is to be situated. The facilities and speed of transportation
both between such point and those at which other headquarters banks
are located, and between such headquarters point and the outlying
portions of the district itself, should be considered.
A constraint identified by Willis in 1914 was "the fact
that to a certain extent the sites of reserve banks must be regarded as
practically predetermined, as in the case of New York, Chicago, and St.
Louis, the present central reserve cities." Willis noted that assigning
these cities also suggested regions around them and this affected districting
in the rest of the country.
These are the specifics on districting from Willis and the
preliminary committee in 1914. Intriguingly, Willis refers to a "detailed
statistical and geographical analysis" giving the "sites for the Reserve
Banks and the limits of the districts pertaining thereto." According to
Willis, this analysis includes versions of a System with eight, nine,
10, 11 and 12 Reserve bank cities and districts, which Willis handed to
McAdoo and the Reserve Bank Organization Committee. Further, according
to Willis in 1923, "It was an analysis designed to apportion the banks
of the country upon commercial and financial lines from the point of view
of convenience and financial efficiency." However in the preliminary organization
committee's report of 1914: "The Committee having since that date established
the districts and indicated the headquarters of the banks, the detailed
analyses are here omitted."
Willis' writing in 1923 on the Federal Reserve System extends
and develops the points made above but fails to provide more information
on the district banks and districting that he may have suggested to the
Upon public announcement of its decision (see Figure
1) on April 2, 1914, critics charged that politics influenced the
Reserve Bank Organization Committee in the choice of the number of districts
and the location of as many as four of the Reserve bank cities. Especially
upset with this plan were people and politicians in the cities of Baltimore,
Denver, Pittsburgh and New Orleans, who believed that St. Louis, Kansas
City, Cleveland, Atlanta and/or Richmond were indefensible choices as
Reserve bank cities.
Private bankers in large cities (for example, New York and
Chicago) thought that too many Reserve districts were chosen; these critics
felt that the committee should have started with eight or even fewer districts
and Reserve banks.
As quoted above, Willis' 1923 account of the choice of number
of districts and Federal Reserve bank cities agrees with the critics of
the time that political activity explains many of the choices made by
the Reserve Bank Organization Committee. Whether the charges were true,
the committee fully expected that its decision would arouse strong feelings.
For example, Secretary of Agriculture David Houston wrote in his memoirs
that the committee would be "in for a great deal of roasting no matter
what we decided."
Willis and the "Report on Districting"
If politics indeed played a significant role in the location decisions,
Willis, with an academic background and with fewer constraints than the
members of the committee, would be expected to provide districting schemes
different from that chosen by the committee. (For example, Secretary McAdooWilson's
son-in-lawwas politically ambitious, seeking the Democratic Party's
nomination for president in 1924.) Evidence of significant differences
between Willis' recommendations and the final outcome could provide evidence
for the "political influence" hypothesis. On the other hand, if Willis
suggested virtually the same districting scheme that was announced by
the committee, this would either weaken the political influence hypothesis,
or it suggests that Willis gave the committee what it desired (thus playing
politics himself). This would allow the committee to hide its political
decisions behind a veil created by a supposedly neutral and unbiased expert.
In this case, one expects that the committee would justify its location
decisions by publicly referencing Willis' report.
In his 1914 "Report on Districting" to the Reserve Bank
Organization Committee, Willis provided plans for a System featuring eight,
nine, 10, 11 and 12 Federal Reserve districts along with suggestions for
Reserve bank cities for each district.9
Willis' analyses leading to his various plans show that
he sifted through the 5,000 pages of testimony carefully in an attempt
to satisfy the Act and the organizing principles. However, it does not
appear that Willis had access to the ballots submitted by the national
banks when he was preparing his report.
Most individuals testifying before the Reserve Bank Organization
Committee, unsurprisingly, provided arguments as to why their particular
city should be the headquarters city for a region that they conveniently
had surrounding their city. As Secretary Houston put it in Kansas City
on Jan. 23: "Each place we go is the center." This boosterism was sometimes
Willis' report starts with "Part I: A theoretical treatment
of the general principle. ..." The section runs for the first 13 pages
and is the source for the "fundamental principles" paraphrased above from
the Report of the Preliminary Committee and also the source for the lengthy
treatment on the same topics that is reproduced in Willis' 1923 volume.
The Willis report is sharper in focus, detail and attitude than either
the report of the Preliminary Committee or his volume of 1923. For example,
whereas in the Report of the Preliminary Committee, Willis writes that
the choice of New York, Chicago and St. Louis is "practically predetermined,"
in his 1914 report Willis says "the choice of these cities may be regarded
Parts II and III present his proposed eight, nine, 10, 11
and 12 Federal Reserve district plans. Part IV is "A defence of the grouping
suggested." Part V is "A discussion of the location of the headquarters
banks and of their branches."
Concentration of banking capital in New York and Chicago
meant these cities had to serve either very small, but capital-dense,
districts (New York, where there was even the suggestion by one witness
that New York City be split into two different reserve districts) or very
large and capital-sparse (outside the headquarters city) districts (for
example, Chicago). This is especially true when the plans for a smaller
number of districts (eight or nine) are considered. Willis' eight- and
nine-district plans appear as Figure 2. The
districting is identical except for the division of the Chicago district
into two, with the new district headquartered at Minneapolis spreading
north and westward toward the Pacific Northwest.10
In all of Willis' plans, determining boundaries for the
Northeast and Atlantic seaboard were the most vexing. Starting at New
"The argument in favor of ... a bank at New York is so strong
as to need no further development," Willis wrote in 1914. Moving to the
Northeast, testimony taken in Boston made it clear that while bankers
and businessmen considered it their financial hub, transport time did
not dictate a district bank in Boston. And witnesses in Boston said that
they could deal with a district bank in New York. However, in order to
keep the New York district bank from being disproportionately large in
capitalization, Willis placed the New England states, the eastern third
of New York and northern Connecticut into a separate Boston district.
Then Willis moved on to the Philadelphia district. (With
New York City and Boston being self-evident locations for Reserve bank
cities, Willis referred to the districts surrounding them by their city
names. However, as his work progressed, he followed a schema of districting
first and then determining and recommending headquarters cities. This
did not stop him from referring to some districts by a city name, for
example, the "Philadelphia District," but justification for this came
later.) Again, balancing the flows of business and banking with transportation
considerations and the requirements of capital size in districting, Willis
created a district south and east of New York. Virginia was included in
this district because, as Willis wrote in 1914, "testimony ... seems to
show that the banks of Virginia are considerably closer in their connections
with those at New York, Philadelphia and Baltimore than they are to the
institutions south of them."
Willis next determined the eastern extension of the Chicago
district and then his Great Lakes district, which went east into New York
and south to include parts of Kentucky. Here Willis in 1914 was very deliberate
in not identifying the district with a city: "It will have been noted
that nothing has been said in this discussion with reference to the headquarters
[city]. The cities ... which might ... be ... headquarters ... include
Cincinnati, Pittsburgh, Cleveland, and possibly Detroit, Buffalo, and
Louisville. From the standpoint of accessibility, convenience and general
advantage it is believed that Cleveland is probably to be preferred. ...
The question of proper headquarters point is merely touched upon here
and will be dealt with more fully in the section devoted to that subject."
Willis then jumped to the "Problem of the West ... A large
region without the excessive concentration of banking capital ... is to
be subdivided. The problem is that of establishing a division corresponding
... with the commercial habits of the people, placing together those cities
which are most nearly allied to one another and joining in the same district,
territories which are easily accessible one to another by available railway
transportation." In 1914 Willis refers to this as the Chicago district
and adds: "One further point ... needs to be considered. This is the relation
of the Chicago district to the other districts. For somewhat the same
reasons which were controlling in the case of the New York district it
is desirable that the Chicago district should not be too large."
At this point Willis outlines the St. Paul-Minneapolis district
and a large Pacific Coast district with San Francisco as its headquarters.
The St. Paul-Minneapolis district is suggested, but Willis says that if
the committee wants an eight-bank plan, then that district should be subsumed
into the Chicago district.
The St. Louis district comes next. Willis writes in 1914:
"The real problem is ... the drawing of a line from the southwestern corner
of the Great Lakes district ... to include in the St. Louis district the
territory naturally and properly belonging thereto as distinct from the
southeastern district while making a proper subdivision of capitalization
... and at the same time making the best possible provision for transportation
requirements in the two districts."
In this plan Texas is trisected. "Such a division of Texas
is sustained by a study of the transportation situation. ... Analysis
of the testimony taken at New Orleans, Austin, Tex., and elsewhere will
... sustain this division ... it being the present opinion that to whatever
city Dallas and Ft. Worth may be assigned as headquarters, their business
tends northward rather than to New Orleans. Thus in the absence of
a headquarters bank in Texas itself the middle and western region
of Texas is thrown to the north of a line dividing it from the southeastern
section of the country in somewhat the way indicated." (Emphasis added.)
Finally, the Southern district, a "district ... established
entirely by a residual process." Willis also suggests on the same page
in his 1914 report "a headquarters located either at New Orleans or some
point further east."
These are Willis' eight- and nine-district plans. Summarizing,
he writes in 1914: "... the plan just mapped out provides ... a reasonably
practicable division of the territory. ... this division is in certain
sections ... far more satisfactory than in others ... but this condition
is unavoidable, in view of the irregular distribution of population ...
banking capital, as well as the differing character of transportation
facilities, if it be desired to adhere as closely as possible to the smaller
number of institutions indicated." 11 See
Capitalization of the District
(In Thousands of Dollars)
Boston or New England
Philadelphia or Allegheny
Chicago or Central
St. Paul-Minneapolis or Northwestern
St. Louis or Western
Kansas City or Southwestern
St. Louis or Mississippi Valley
Source: Willis, 1914, pages 108-109.
Addressing the weaknesses in the eight- and nine-bank plans outlined,
Willis creates 10, 11 and 12 bank plans as shown in Figure
3. 12 In his
10-district plan, Willis creates a district from portions of the St. Louis
and Chicago districts, also adjusting the St. Louis district's eastern
boundary. In this section in his 1914 report, Willis addresses the issue
of the location of the headquarters bank explicitly. The candidates were
Omaha, Lincoln, Denver and Kansas City. Willis finds that Kansas City
has the best claim based on flows of business and Colorado's lack of bank
capitalization. (See "Willis' Choice of Headquarters
Cities and Selection Criteria.")
Following his method of balancing capitalization, flow of
business, district size and transportation, Willis also proposes 11- and
12-district plans, as indicated in Figure 3. Addressing the issue of geographical
size, he splits the Pacific Coast district into North and South Pacific
districts with headquarters cities at Portland and San Francisco, creating
Finally, Willis addresses the issue of the disproportionate
capital in the Great Lakes district by splitting it into Eastern and Western
districts and breaks out capitalization in the eastern Great Lakes at
$11,690,000 and in the western Great Lakes at $7,295,000, while capitalization
in the other districts remains unchanged.
Of these plans, Willis strongly recommended the 12-district
plan, stating that it is "more convenient and workable from the standpoint
of transportation and hence of bank clearing, [and it is] also obvious
that this division measurably satisfies the demands of the country for
the establishment of reserve banks, and affords recognition in a substantial
way to practically every distinct section of territory included
it." 13 It is surprising, then, to read Willis in 1923 claiming that "either
nine or ten was the first recommendation made to the [Reserve Bank Organization]
Committee." This is certainly inconsistent with his opinion in his confidential
report of 1914.
Willis was also concerned in 1914 with the legality of rearranging
districts that would be necessary if a smaller number of districts were
to be created initially, with the number increased later. He even held
out hope that if 12 proved to be too few "... that, when the need arises
for more, Congress will make due provision for ... others." 14
Reserve bank cities
The choice of Reserve bank cities was the most hotly contested issue
before the committee. Almost without fail, witnesses in every location
made the best case for their city as a Reserve city. Given the newness
of the Act, the uncertainty of the power and scope of the institutions
it was creating and the speed with which it would be implemented, this
type of testimony was to be expected.
Witnesses knew only that eight to 12 Reserve district banks
would be created. The nature and responsibilities of these banks were
unknown. When the subject of branches (of the Reserve banks) came up,
more confusion arose. At the time of the committee's tour it was not at
all clear what role the branch banks would play and what power they would
have. Consequently, witnesses were not at all mollified to hear that their
city might receive a branch bank if it did not get a Reserve bank.
Willis addresses the choices of Reserve bank cities in Part
V of the 1914 report: "Wherever possible, the preference should be given
to a city which has ... a distinct leadership in matters of business,
and with which the banks within the district are in the habit of dealing.
Where the choice between two cities would otherwise be difficult to make
on account of practical equality in the extent and importance of their
commercial relationships, the city to be chosen should be that which has
the advantage in ease of communication. This is fundamentally desirable
because of the necessities of the clearing process and the advantage to
be gained from having all parts of the district within easy reach of headquarters." 15 (See table for headquarters cities recommended
by Willis and his rationale for selecting some of them.)
Willis and the Reserve Bank Organization Committee
The committee's districting plan, Figure 1, at first glance looks very
different than Willis' 12-district plan in Figure 3, especially in the
East. There is little explanation for the differences.16 Willis submitted his report prior to the committee's final deliberations,
but the committee does not refer specifically to Willis' report in either
the release of April 2, 1914, or that of April 10, 1914. I have found
no meeting notes covering the committee's deliberations; further, Willis'
letter to Hamlin (see endnote 13) suggests that there
was no recorded vote.
In 1923 Willis strongly suggests that politics played a
major role in the placing of district banks in Richmond, Atlanta and Cleveland
(and the absence of one in New Orleans). Yet, his own confidential report
of 1914, presumably untainted by politics, recommends Cleveland and Atlanta,
while not recommending New Orleans. Willis makes no mention of these similarities
In an effort to stem growing criticism of its plan, the
committee did put out supplementary material April 10, 1914, including
publishing the results of the national commercial bank balloting in support
of their plan. Yet they did not cite Willis' report.
If politics played as large a role as Willis claimed in
1923, his strongly suggested 12-district plan and that of the committee
should be significantly different. At the very least, the choices of the
Reserve bank cities should be different. However, the two schemes are
similar in many respects. They are both 12-district plans; the districting
in the western region of the country is quite similar; and, while the
committee made a district out of Texas (and parts of Arizona, New Mexico,
Oklahoma and Louisiana), this is in the spirit of Willis' comments (see endnote 16).
The two biggest differences are the committee's expansion
of the New York district to include the entire state of New York (whereas
Willis split the state between greater New York City, Boston and the Eastern
Great Lakes districts) and the creation of a district including West Virginia,
Virginia, North and South Carolina with Richmond, Va., as a Reserve city.
Willis, though aware of the political pressure for Richmond,
does not recommend a district with Richmond as headquarters city. The
potential for political pressure was seen to come from the facts that
Carter Glass, the U.S. House of Representatives member who wrote the Act
with Willis and saw it through the intense political trip through the
U.S. Congress, represented Virginia; and that the Comptroller of the Currency,
John Skelton Williams, and the president, were both from Virginia. Willis
himself had taught in Virginia at Washington and Lee University for nine
years prior to joining the National Monetary Commission and then the House
Banking and Currency Committee as an expert in 1912.
Willis did not recommend Richmond as a Reserve bank city.
However, in comments upon the outcry against the committee's decision
and the small likelihood of relocating the Reserve banks away from the
largest cities once the decision had been announced, Willis writes in
1923: "With these [New York City, Chicago and St. Louis] was usually ranked
the Richmond bank, not because of any fondness for that institution, but
because of the belief that it had special protection which would enable
it to resist all attack."
However, Willis does give the following advice to the committee:
"To sum ... up, the division should be made by assigning each individual
bank to a given district so far as possible in accordance with preference."
Now, it was these preferences (of the banks) that Willis
did not have or, at least, never refers to. Meanwhile, the committee does
refer to these preferences when justifying its districting plan. The committee
also specifies that following state boundaries where possible was an objective
and states its decision of April 2, 1914, in the U.S. Senate document.
This self-imposed constraint (the language not being in the Act) may explain
the committee's squaring off of several of Willis' suggested districts.
For example, fewer states are bisected or trisected under the committee's
scheme. Further, appeal to the national bank ballots does support the
Richmond decision, according to the U.S. Senate document.
Willis and the committee agree on 10 of the 12 headquarters
Reserve bank cities, differing only on Richmond and Dallas (committee),
vs. Cincinnati and Portland, Ore. (Willis).
Willis' 12-district plan presents the committee with basically the same
districts and cities as they announced on April 2, 1914, excepting the
Dallas and Richmond districts. It cannot be denied that politics may have
played a role in the decision to designate Richmond as a Reserve bank
city, as charged by critics. However, its selection is also supported
by the votes of the nationally chartered banks in that district. The Dallas
district was mentioned as a possibility, albeit briefly, in Willis' report
and is also supported by the ballots.
Given this, the Willis report (echoed by the committee)
suggests business flow, transportation, district capital and factors other
than political opportunism for the number and location of Reserve districts
and district cities.
The political wrangling may have been played out in both
the House and Senate during the debates over the Act. It is also possible
that those individuals disagreeing with the Reserve Bank Organization
Committee's decision thought that the soon-to-be appointed Federal Reserve
Board would significantly alter the committee's districting scheme in
both number and shape.
Of course, this is exactly what happened, but attempts
at large-scale change were cut short by the decisions of the attorney
general and the failure of the dissident group within the Federal Reserve
Board to carry their challenge to the Supreme Court.
The advent of World War I may have played a role in the
failure of the proponents of a more centralized Federal Reserve System
to take this dispute to a higher court. The Act created an entirely new,
national and quasi-governmental financial organization in the United States
at almost the same time as the start of the war. Perhaps not wishing to
add additional uncertainty to an international financial scene that was
about to become very unsettled, proponents of a smaller, more centralized
Federal Reserve may have withdrawn their criticisms in support of presenting
a united financial front to the rest of the world.
Further, given the nature and content of Willis' confidential
report, it is surprising that the committee does not use it in defense
against the charges of politics. A tentative reason may be that the committee
was also receiving strong criticism from the powerful private commercial
banking communities that they had chosen too many districts. Perhaps the
committee did not wish to provide those critics with viable, worked-out
plans for eight or nine districts, subject to the concerns Willis expressed
earlier (see endnotes 11 and 12),
produced by its own expert that critics could have then used against the
What is also surprising is Willis' failure to publish his
report of 1914. The material in his book of 1923 is at odds in important
respects from the confidential report that he provided to the committee.
The recommendations in his report are much closer in most respects to
the committee's final decision than one would guess from his 1923 volume.
Why this is remains a mystery requiring further study.17 At this point we know that from 1914 to 1918 Willis was the first Secretary
to the Federal Reserve Board, two members of which were also on the Reserve
Bank Organization Committee (McAdoo and Williams). Following that, he
was the Federal Reserve Board's director of analysis and research in New
York (1914-1918) while also teaching at Columbia University (1916 until
his death). A cynical, but unsubstantiated, suggestion is that Willis
saved his charges of politics until 1923 when, presumably, they would
be harder to refute and also when his charge would have fewer repercussions
on his career.
This article is a somewhat abbreviated version of
the author's original research
paper. [PDF format]
Bothwell, Wilbur Clarence, The Federal Reserve Bank of
St. Louis, Ph.D. dissertation, Washington University, St. Louis, Mo.,
Federal Reserve Bank of Minneapolis, The
Region, Vol. 2, No. 2, August 1988.
Gunn, John A., A Century of Liberal Education for Business
and Selected Professions, Washington and Lee University.
Johnson, Roger T., Historical Beginnings ... The Federal
Reserve, Federal Reserve Bank of Boston. Revised, December 1995.
Mishkin, Frederic S., Money, Banking and Financial Markets,
6th ed., Boston: Addison Wesley, 2000.
Morgan, Perl W. (ed.), History of Wyandotte County Kansas
and Its People, Chicago: The Lewis Publishing Co., 1911.
Nelson, Clarence W., Reflections From History: First
Half-Century of the Minneapolis Federal Reserve Bank, Federal Reserve
Bank of Minneapolis, November 1964.
Odell, Kerry A., and Weiman, David F., "Metropolitan Development,
Regional Financial Centers, and the Founding of the Fed in the Lower South," The Journal of Economic History, Vol. 58, No.1, March 1998, pps.
Preliminary Committee on Organization, Report to the
Reserve Bank Organization Committee, Confidential No. 45, Library
of the Federal Reserve System, Washington, DC, 1914.
Primm, James Neal, A Foregone Conclusion: The Founding
of the Federal Reserve Bank of St. Louis, Federal Reserve Bank of
St. Louis, 1989.
Reserve Bank Organization Committee, Stenographic Reports,
unpublished, held at the Library of the Federal Reserve System, Washington,
Timberlake, Richard H. Jr., Monetary Policy in the United
States: An Intellectual and Institutional History, Chicago and London:
The University of Chicago Press, 1993.
United States Senate, Location of Reserve Districts in
the United States, 63rd Congress, 2nd Sess., Document No. 485, Washington,
DC: Government Printing Office, 1914.
Warburg, Paul G., The Federal Reserve System: Its Origin
and Growth: Reflections and Recollections, New York: Macmillan Co.,
West, Robert Craig, Banking Reform and the Federal Reserve,
1863-1923, Ithaca: Cornell University Press, 1977.
White, Eugene Nelson, The Regulation and Reform of the
American Banking System, 1900-1929, Princeton, N.J.: The Princeton
University Press, 1983.
Willis, H. Parker, The Federal Reserve System, Legislation, Organization
and Operation, New York: The Ronald Press Co., 1923.
Willis, H. Parker, Report on Districting, unpublished
manuscript, National Archives, College Park, Md., 1914.
Willis, H. Parker, Letter to Governor Hamlin, unpublished,
National Archives, College Park, Md., Dec. 22, 1915.
Willis, H. Parker, Collected Papers, Rare Book and Manuscript
Library, Columbia University, New York.
1 Willis' account is repeated in Bothwell (1941) and recounted in Primm
(1989). The charge of political influence is hard to discount. James Beauchamp
Clark, a Democrat and Speaker of the U.S. House of Representatives during
this period was from Missouri. The Missouri congressional delegation featured
two Democratic senators and 15 (of a possible 16) Democratic House members.
Kansas, meanwhile, had one Democratic Senator and five (of eight) Democratic
House members. This political power is offered as the reason why there
are two Federal Reserve district banks in Missouri.
It is undoubtedly this type of information
that prompts questions like: "What political realities might explain why
the Federal Reserve Act of 1913 placed two Federal Reserve banks in Missouri?"
(See Mishkin, 2000, p. 389, emphasis added.)
Prominent Virginians, including President
Wilson, were involved both in drafting and passing the Act, often this
is presented as evidence for the location of the Reserve district bank
in Richmond, Va. Other influential members of the Cabinet were thought
to be inveigling for a city or cities in their home state.
Odell and Weiman (1998) investigate the
economic characteristics of major Southern cities in the early 1900s and
suggest that economic factors made Atlanta and Dallas likely candidates
for a Reserve district bank. They report (1998, pages 104, 119) Willis'
(1923) contention that politics played a decisive role in the choices
of Atlanta and Dallas as Federal Reserve district headquarters. They note,
however, that both Willis (1923) and the Reserve Bank Organization Committee
(1914) cite economic analyses and justifications for the choices of these
Given this, it appears as if a simple
charge of politics as the answer to the location questions would be harder
to sustain. Odell and Weiman (1998) make clear that in the cases of Atlanta
and Dallas economic factors justify their selection. This supports the
claims by the Organization Committee (1914).
2 The Federal
Reserve Act was signed into law Dec. 23, 1913. There are echoes in the
Act of the longstanding issues confronting the nation since its founding.
Issues of states rights vs. federal rights, urban vs. agrarian interests
and the distrust of private power vs. the distrust of centralizedusually
federalgovernment power were hotly debated. See West (1977), White
(1983) and Timberlake (1993, especially chapter 15).
3 Prior to
the Act, the Aldrich Plan and the Aldrich-Vreeland Act proposed to reform
the U.S. monetary system. The Aldrich-Vreeland Act was a "temporary palliative" (Johnson, 1995, p. 17) and the Aldrich Plan had no hope of passing in
a Congress hostile to a Republican president (Johnson, 1995, p. 19). It
was the election of Woodrow Wilson and Democratic majorities in both the
House and Senate that allowed such large-scale monetary reform as proposed
in the Federal Reserve Act to take place (West, 1977, pps. 90-91; Timberlake,
1993, chapter 15).
4 See U.S. Senate (1914), Willis (1923, chapters
24 and 34), Nelson (1964), Minneapolis Federal Reserve Bank (1988), Primm
(1989) and Johnson (1995).
5 McAdoo was
raised in Georgia, becoming prominent in New York before being named to
the Cabinet. Houston, a New England native, was president of Texas A&M
University and chancellor of Washington University in St. Louis prior
to Cabinet service. Williams, of Richmond, Va., awaited confirmation by
the Senate to the office of Comptroller of the Currency as the committee
began its work. Consequently, he was not present on the tour of the country,
joining the committee for its final deliberations and decision-making.
See Primm (1989, p. 38) and Johnson (1995, pps. 35 and 36).
6 The Act also
specified that the member commercial banks in each Reserve district should
be able to supply at least $4 million total in paid-in capital into the
district's Reserve bank. This would be generated by each member commercial
bank buying stock in the Reserve bank equal to 6 percent of their paid-in
capital. This made the Reserve banks privately owned by member commercial
banks. It also provided a constraint on districting, as the district would
have to have at least $66.67 million in nationally chartered and state
member bank capital in order to satisfy the requirement.
7 Other members
chosen by Willis were Edmund D. Fisher, Andrew A. Benton, O. Howard Wolfe,
Joseph A. Broderick, Ralph Dawson and Stephen H. Farnham. These members
worked on matters regarding the organization and operation of the Federal
Reserve System. From Willis' comments quoted in the text of this article,
none of these members of the Preliminary Committee assisted Willis in
his drafting of the report on districting for the Reserve Bank Organization
8 The exact
date of this assignment is unknown. Willis testified before the committee
in New York City Jan. 5, 1914. There was no mention by either McAdoo or
Willis, at that time, of Willis being engaged upon work for the committee.
At the time of his testimony, Willis was editor of the New York Journal
of Commerce. Previously he had been a professor of economics at Washington
and Lee University, a professor and dean at George Washington University
(1907-1912), the executive director of the National Monetary Commission
and, through 1913, had served as the expert adviser to the House Committee
on Banking and Finance, chaired by Carter Glass of Virginia (see Gunn,
9 This report
is in the National Archives, College Park, Md. It is typescript in the
form of a letter from Willis to the Reserve Bank Organization Committee.
It is approximately 140 pages long, signed in ink by the author on the
last page and covered throughout with pencil markings, notations and instructions
for printers, appearing as if it were being readied for typesetting and
printing. On the first page Willis has written in pencil across the top: "Appendix II. Report on Districting. The following is the report submitted
by the author to the Reserve Bank Organization Committee on the question
of dividing the country into federal reserve districts."
either corrections or mapsare in the Willis Collection at Columbia
10 Given that
discounting of bills was to be done by banks presenting the bills to be
discounted to their Reserve bank, much care and consideration was given
to travel times and railway schedules. When debating the Act in the Senate,
Sen. John Schafroth of Colorado argued that there should be a "one night's
train ride" constraint on distance from the farthest flung bank and its
district Reserve bank (Timberlake, 1993, p. 220). The Reserve Bank Organization
Committee sought information on these transportation and timing matters
in the testimony presented, and Willis also considers this constraint
important in his work.
11 Willis (1914,
p. 67) finds the Chicago district in the eight-district plan awkward: "This district would be clumsy and unwieldy with long stretches of transportation
for checks passing to the headquarters at Chicago from the "far-flung" western and northern borders of the territory, but it must be admitted
that the general drift of transportation and trade would permit such a
consolidation, aided by the establishment of a suitable branch or branches,
were it desired to keep the number of banks down to the minimum of eight
established by law."
12 Most of
these weaknesses involve transportation. Willis (1914, p. 69): The southern
district "is very large ... transportation between its eastern and western
borders is decidedly unsatisfactory. In the Pacific Coast district the
area is extremely large ... time of transportation ... very great. The
districts in the area between the Great Lakes ... and the Rocky Mountains
... are likewise large and ... somewhat unsatisfactory in transportation
... the Great lakes region ... is widely extended ... while its transportation
is good it ... includes an amount of bank capitalization which could ...
13 The recommendation
of the 12-district plan is underlined in Willis' original report on page
110. It is the only such underlined and emphasized segment in the entire
report. Odd then, that on Dec. 22, 1915, when there was a roiling internal
debate (see endnote 14, following) about whether the
Federal Reserve Board could alter, rearrange and/or reduce the number
of Reserve districts, Williswho by this time had been hired as the
Secretary to the Federal Reserve Boardshould write to then-Governor
of the Federal Reserve Board Charles S. Hamlin on the issue:
"My dear Governor Hamlin: I have your letter
of December 22 as to redistricting.
I am not sure what is meant by the expression
"ballot of the hearings," but if this refers to a tabulation that was
prepared of the testimony before the Organization Committee, I may say
that there was no "vote" there. ... With reference to the quotation you
make as to my supposed views, I may say that my views are the same that
I expressed at a meeting of the Board some two weeks ago ... viz.: (1)
I hold substantially the same views today that I held in my original report
submitted to the Organization Committee. (2) That report contained a plan
for eight, one for nine, one for ten, one for eleven, and one for twelve
reserve banks. I also added that in conversation with the Secretary of
the Treasury at the time I expressed a feeling that the establishment
of probably nine or ten banks at the start would be wisest, leaving two
or possibly three for subsequent establishment by the Federal Reserve
Board. ... Trusting that this covers the ground, I am Sincerely, H. Parker
Willis, Secretary." (Emphases in the original.)
This letter is also in the National Archives, College
Apparently, the Willis who strongly recommended 12 districts
in his confidential 1914 report, "forgot" this when he was corresponding
with Governor Hamlin less than two years later when he claims to have
supported a nine- or 10-district scheme. It may be this contradiction
that caused Willis to refrain from publishing his report.
Given the suspicions that surrounded the committee's
choice of Richmond, and a district surrounding this city, it is unlikely
that the committee would push to publish Willis' report. In fact, the
committee does not name Willis' report in its listing of material sent
to the U.S. Senate (see U.S. Senate, 1914, pps. 374-384 for a listing
of the materials given to the Senate that aided the committee in its selection
of Reserve bank cities and districts).
It is hard to underestimate the vehemence and passions
surrounding the issue of whether the Federal Reserve Board had the legal
authority to change the decisions of the Reserve Bank Organization Committee.
In two rulings, Nov. 22, 1915, and April 16, 1922, the attorney general
of the United States ruled that the Federal Reserve Board could not substantially
alter the committee's decisions. Regarding the first decision, the Federal
Reserve Board, via letters from Hamlin to both the attorney general and
the president, attempted to keep the attorney general's opinion from the
public. President Wilson acceded to this request. However, someone leaked
the decision to the press and Hamlin withdrew his requests of both the
attorney general and the president (this correspondence is at the National
Archives, College Park, Md.).
14 These fears
and concerns were well founded. In the year following the announcement
of the Reserve Bank Organization Committee decision, several challengessome
from within the Federal Reserve Board itselfwere made. Some dealt
with trying to reduce the number of districts, some dealt with challenging
the choices of Reserve bank cities. See Willis (1923, chapter 34), Warburg
(1930), Nelson (1964, pps. 18-27), Primm (1989, pps. 46-52), Minneapolis
Fed (1988, p. 8).
15 Here Willis is trying
to solve a multi-objective programming problem. He was, in effect, the
Reserve Bank Organization Committee's "computer" (I am indebted to Professor
Richard Timberlake for this piquant characterization).
Here Willis digresses into a discussion
on "Location of Branches" advising the committee (p. 124): "Care and judgment
should be exercised in ... establishing branches on the first-named basis
... they should not be created unless they are actually needed for immediate
purposes. Even in the latter event, there should be no undue haste in
creating them ... the headquarters banks should be allowed to get a satisfactory
start before the complicating elements involved in branch organization
are allowed to enter the problem ... permitting the question of branch
organization to remain in the background until such time as the actual
establishment of the branches is resolved upon, when the type of such
organization to be adopted may be indicated."
16 Willis does
state (p. 118) that "the state of Texas has sufficient capital of its
own and is organized upon a sufficiently independent economic basis to
warrant the creation of a district running east and west, instead of dividing
Texas between two longitudinal districts running north and south. This
question was sufficiently considered, it would seem, in connection with
the discussion of the southwestern and Mississippi Valley districts. The
point to which attention is called just here is that probably no-where
else in the districting, could this plan be employed to advantage."
17 This and other
Willis-related work is currently in progress with Professor Doug Wills
of Sweet Briar College, Virginia.