The interstate banking debate: a historical perspective.
Hendrickson, Jill M.
INTRODUCTION
Today the commercial banking landscape in the United States is
rather diverse as it is comprised of extremely large bank holding
companies with established branches throughout the county, regional and
mid-size banks also with established branch systems, and small community
bankers who primarily service local markets. This has not always been
the case. Rather, the historical banking landscape is one of thousands
of small, unit banks. Students of U.S. bank history recognize that part
of the reason for the proliferation of these small banks was the legal
inability of most bankers to open both interstate and intrastate
branches. Indeed, branch restrictions emerged soon after the birth of
banking, in the late eighteenth century. Contrary to these restrictions,
many economists, regulators, and bankers recognized very early in U.S.
banking history that branching could add stability to the enterprise.
Ultimately, it took well over two hundred years for the United States to
allow banks to branch freely. During that two hundred plus year process,
there were many attempts to get regulators and legislators to change
branching laws. Who or what was behind efforts to reform branching
restrictions? Why did so many efforts fail? What case was made to
preserve the branching ban and what case was made to eliminate it? Is
there evidence that one side had a better case than the other?
While these questions, and their answers, are perhaps of interest
to the economic historian, they should also be of interest to
contemporary policymakers. In light of the 2007-2008 financial crisis
and the government's role in its resolution, there is intense
interest on how policy and regulation towards financial institutions,
including commercial banks, will be altered in an attempt to avoid a
similar crisis in the future. It is not known if these discussions will
include interstate banking specifically but they will certainly include
proposals to limit acceptable bank activity. In this way, the
contemporary policy discussions are sure to return to discussions had
hundreds of years ago when regulation was enacted to keep banks from
competing with one another and with other financial institutions.
Another certainty is that bankers of all sorts and sizes will have an
incentive to influence the outcome of the policy debates.
From this perspective, this research may be helpful to policymakers
and economic historians alike because, to answer the questions above, it
identifies and then empirically examines the commercial bankers'
position on the branch banking issue throughout history. Though there
are interested parties outside of banking, for example, non-financial
businesses, the public, and regulators, to establish all positions is
beyond the scope of this paper (see Kroszner and Strahan (1999) who
consider the economic and political interest groups in relation to the
deregulation of branch banking). Following this introduction, there are
two main sections of this paper. First, this paper breaks the history of
branch banking into four eras and attempts to establish evidence of the
small and large banker position on the branching issue. Evidence of
these positions comes from reading primary sources such as the New York
Times and the Wall Street Journal, trade publications such as
Bankers' Magazine and the American Bankers Association Journal, and
government documents such as Congressional Quarterly. At the same time,
this section of the paper also carefully documents market developments
that certainly impacted the interest group positions in question. While
this paper is not the first to address the history of branching in the
United States, it is the first to document primary sources over the
history of branching to establish interest group position. Almost all
existing work takes as given the position of the small and large banker.
This paper, in contrast, attempts to carefully document the positions
from the beginning of the debate.
The second primary section of this paper attempts to empirically
test the prominent interest group positions over the three eras for
which data is available. Of interest is whether there is statistical
evidence to support the soundness of one position or another and to
determine if that evidence varied over the history of the debate. The
empirical test finds evidence that both the small and large banker had
statistically valid positions. More specifically, the analysis finds
evidence that legitimizes the small bankers' fear of increased
competition, while the large bankers' prediction that branching
would enhance stability was also valid. To the author's knowledge,
this is the first paper to examine these issues across the history of
branching in the United States.
Since the beginning of commercial banking, the regulation of
intrastate and interstate banking has attracted both critics and
defenders. Interestingly, the arguments on both sides of the issue
remained relatively consistent in the two hundred year debate. With
time, some of the smaller points changed as the nation and banking
changed. However, as this paper clearly illustrates, the earliest
arguments for and against branch banking were still being voiced in the
final debate in 1994.
Supporters of a branch system argue four fundamental points. First,
they often argue that branch restrictions harm banks by limiting their
ability to diversify (see, for example, Newfang (1901), Sprague (1903),
Collins (1926), Wernette (1932), Jay (1933), White (1982), Wheelock
(1992, 1995)). A unit bank is forced to extend loans and make
investments in its immediate community that is frequently not too
diverse. This means that an adverse development to the community could
wipe out most of the bank's loans since its loan portfolio is the
homogenous reflection of the community. The assumption behind this
argument both in theory and in practice is that banking markets are
highly segmented. Bodenhorn (2003) provides several examples of banks
during the antebellum era diversifying their portfolios by lending or
investing beyond local markets. While this particular argument was
leveled against branching restrictions up until the 1994 repeal, the
homogeneity of community life was particularly acute during the
antebellum and national bank eras. Branching restrictions also make a
bank vulnerable on the liability side of its balance sheet. Like the
assets, the liabilities of a unit bank cannot be too diversified which,
in turn, magnifies the possibility of runs and gives banks a more narrow
base from which to draw deposits.
Second, branch proponents argue that the branch restrictions, in
our early history, kept banks from meeting the public's demand for
depository services and constrained banks' ability to meet the
financing needs of larger non-financial firms (see, for example, Newfang
(1901), Sprague (1903), White (1982), Giedeman (2005), Preston (1924)).
As the economy grew westward many communities were too small to support
a national bank or even their own state bank but they needed banking
services nonetheless. At the same time, growth in the non-financial
sector resulted in larger and larger firms, most of whom required
external financing. The contention is that commercial banks were too
small, because of branch restrictions, to adequately provide large firm
financing.
Third, those favoring a system of branches argue it would increase
the competitive environment in banking and lead to lower priced loans
and higher priced deposits to the benefit of the consumer. Branching
would have the immediate effect of increasing competition as branch
units compete with existing banks (Carlson and Mitchener (2006)). In
theory, the increased competition would force weaker banks to exit
leaving behind a stronger network of branches.
The fourth point often made by advocates of a branch system is that
it may improve bank profitability and stability. Bank profit could
increase through the more diverse balance sheets mentioned previously.
Some scholars have also indicated that branching reduces administrative
costs which could increase profitability (Southworth (1928)). Other
scholars argue that branch systems afford banks access to each others
deposits so fewer reserves are required which frees funds to be used for
profitable purposes (Carlson (2004)). Improved profitability certainly
enhances stability, but branch networks are seen to stabilize banking
through another mechanism. If bank panics and runs are the result of
asymmetric information problems so that depositors are unable to
differentiate healthy from unhealthy banks, runs and panics ensue. A
network of branches provides resources to banks caught up in the panic
or run thereby reassuring depositors and restoring confidence (Bodenhorn
(2003)).
On the other side of the issue are those determined to maintain a
system of unit banks. Perhaps the most frequently stated reason for
retaining the prohibition on interstate branching is the fear that it
would lead to monopoly banking. As mentioned above, proponents of
branching argue it increases the competitive environment to have
branches competing against single office banks. However, opponents argue
that if branching forces smaller banks to close their doors because they
could not operate profitably, branching would have the effect of
increasing industry concentration and reducing competition. A reduction
in competition may mean a reduction in interest rates paid on deposits
and an increase in interest rates charged for loans (see, for example,
Sprague (1903), Collins (1926), Wheelock (1992)). In essence, this
position is concerned with the possibility of branching leading to a
monopoly control of commercial banking.
Proponents of unit banking also argue that branching will hurt the
small local borrower as the bigger, and more distant, bankers will
perhaps be less willing to extend credit (Jay (1933)). Sprague (1903)
identified early in the debate the asymmetric information problem that
may arise with branching. Specifically, Sprague recognized the personal
nature of many unit bank loans and the close relationship between the
lender and borrower. This relationship provides the lender with more
information about borrower risk characteristics. It is argued that some
of this information may be lost, and important loans not made, under a
branching system. Collins (1926) identifies this as the fundamental
problem with a branch system.
THE EVOLUTION OF THE INTEREST GROUP POSITIONS
This paper chronologically analyzes the evolution of the branch
banking debate in the United States. The analysis begins by attempting
to identify the interest group positions for four periods in U.S.
banking history: the antebellum era which considers the period from the
inception of banking to the civil war; the national bank era which runs
from 1864 to 1913; the Great Depression era which includes the years
1914 through 1935; and the modern era which picks up the analysis in the
early 1960s through the passage of the Riegle-Neal Interstate Banking
and Branching Efficiency Act of 1994 that allows for interstate banking
in the US effective June 1997. In addition to analyzing the interest
group positions for these historical periods, market developments are
also identified and analyzed. The empirical testing of the positions
follows in the next section of the paper.
Antebellum Era
The state of commercial banking varied tremendously throughout the
country in antebellum America. This variation was a reflection of the
diverse degrees of economic development. Not surprisingly, branch
banking regulation also varied from state to state. Chapman and
Westerfield (1942) provide the most thorough consideration of branching
during the antebellum era and find that most states actually allowed for
branching, either intrastate or intracounty.
The antebellum period did see two cases of interstate branching as
both the First and Second Banks of the United States were permitted to
branch. That is, while state banks were prohibited from interstate
branching, both federal banks established networks of branches across
the states affording them all the advantages of branching. The First
Bank of the United States (1791-1811) operated eight branches and the
Second Bank of the United States (1816-1836) operated twenty eight
branches (Boeckel (1937)). Thus, banks chartered by state authority were
prohibited from interstate branching while the two banks with federal
government charters enjoyed the privilege. Not surprisingly, this
discrepancy, particularly with the Second Bank, did not sit well with
many and thus the branch bank controversy, on a national scale, was born
in the United States.
The Congressional Globe documents congressional activity between
1833 and 1873. As such, it is the appropriate source for government
discussion of branch banking during the antebellum era. There is no
discussion directed specifically to branching. Rather, the vast majority
of banking debate surrounded three issues: 1) the constitutionality and
charter of the Second Bank of the United States; 2) currency activity of
state banks; and 3) the payment of interest on deposits at state banks.
The only specific reference found to branching was in 1836 when Senator
Benton spoke against a proposed bill that would give authority to the
Treasury to establish branches of state banks in states without banking
(Congressional Globe, June 18, 1836, 500). Seven states--Arkansas,
California, Florida, Iowa, Oregon, Texas, and Wisconsin--prohibited
banking at some point in antebellum America (Hammond (1957)). Like many
issues of the day, this branching idea was criticized as an intrusion of
state rights. In the end, it appears Congress did not consider the
branch issue during the antebellum era.
The New York Times historical database begins in 1851 and there is
no evidence from this source that branch banking was a part of the
national dialogue. Several other national sources also fail to identify
branching as an issue during this era in banking history. Consequently,
direct evidence of interest group positions in antebellum America on the
branching issue has not been found.
Tables 1 and 2 contain a sample of the interest group positions on
the branch banking issue for the final three banking eras. Historical
research indicates that the small banker was consistently opposed to
branching on the grounds that it would create a monopolistic banking
system at the expense of the local bank and community (see table 1). In
the early history more emphasis was placed on the monopoly perspective
and later more emphasis was placed on protecting communities and their
bankers. The large banker was also consistent, through all three eras,
advocating that branching was stabilizing and would benefit all bank
clients (see table 2). In the earlier eras, the large banker also
emphasized the position that limits on branching made banks vulnerable
to local economic downturns. Tables 1 and 2 paint a broader picture of
the nature of the interest group positions throughout history. In what
follows, more details of this evolution are provided for the final three
banking eras.
National Banking Era
Financial historians refer to the national banking era as the
period between the creation of nationally chartered banks in 1864 and
the 1913 creation of the Federal Reserve. The opportunity to obtain
national bank charters came with the passage of the 1863 National
Currency Act and its subsequent revision as the National Bank Act of
1864. The National Bank Act was interpreted to prohibit the branching of
national banks. That is, the act did not specifically prohibit
branching, but the act required that all business take place at the
location designated on the certificate of operation. The second
Comptroller of the Currency, Freeman Clark, interpreted this to mean
that national banks could not open branches (White (1983)).
Nationally chartered banks were not allowed to establish branches
either within a state or between states. At the same time, many states
prohibited branching for state banks and even in those states that
afforded branching opportunities, few banks actually participated.
Throughout the entire national banking era, there were fewer than 500
branches in the entire country (Federal Reserve Bulletin, various
years).
The national banking era also witnessed five banking crises. While
four of these were largely contained to New York and the surrounding
area, the 1893 crisis was larger in terms of the number of bank failures
(over 500) and in its geographic spread throughout the country. All of
these experiences were certainly on the mind of those interested in the
branching issue.
At the same time, U.S. banking witnessed a clear tendency of
increased competition from foreign banks. In 1902, a writer for the New
York Times described the growing trend for foreign banks to set up
branches in cities other than the city of the parent bank. This happened
because foreign banks were not subject to the branch restrictions of
U.S. banks. He adds:
The fact appears to be that these changes are only another
manifestation of the force impelling the financial business of the
country as it has already impelled industrial business in the direction
of larger organization and co-operation. Americans are as far behind the
rest of the active business nations in their banking system as they are
ahead of the rest of the world in industrial organization. (New York
Times, May 19, 1902)
The domestic crises and growth of foreign banks both added fuel to
the branching debate during the national banking era. Yet, in the end,
branching made little headway during this period, largely because of the
political strength of the small banker.
Great Depression Era
Branch banking, as an important public issue, fell off the
nation's collective radar following the national bank era. It was
not until the early 1920s that branch banking came to the forefront of
public attention. By the middle of the 1920s scholars were calling the
branch issue the most important domestic bank issue facing the nation
(Collins (1926)). Where did this interest come from? Much of it came
from developments in the banking sector. For example, between 1920 and
1926, over two thousand commercial banks failed and most of these were
small unit banks (Collins (1926)). One response to these failures was
interest in a branch system over a unit system. Another development was
the changing composition of commercial banks. Bradford (1930) indicates
that beginning in about 1915, the growth in state banks began
threatening the position of national banks. Moreover, Bradford (1930)
points out that much of the state bank industry gains were the result of
branch banking powers in certain states. Preston (1924) makes a similar
claim to Bradford but is even more precise by arguing that four
developments generated this renewed interest in branching.
First, the Comptroller of the Currency, D.R. Crissinger, observed
that in some states many of the state banks were withdrawing from the
national bank system so that they could branch. The Comptroller feared
that the branching in some states represented an unfair advantage to
those state bankers. In 1921, Mr. Crissinger asked Congress to grant
nationally chartered banks the ability to branch within the city or
county that the parent bank was located. When Congress failed to
respond, the Comptroller took matters into his own hands. In early 1922
he let national bankers know that he would permit them the intra-city
branch rights in those states that allowed branching. National banks
responded by opening branches.
The second development was a market response to Mr.
Crissinger's declaration. The First National Bank of St. Louis
immediately established a branch within the city even though the state
of Missouri prohibited branching. After a court battle, the bank was
required to close its branch. Nonetheless, the controversy facilitated
more national discussion on branching.
The third event was regulatory change by the Federal Reserve. The
Federal Reserve Act allowed a state member bank the right to retain and
perhaps even expand its branch system. However, regulators, much like
the Comptroller, were becoming increasingly concerned that the state
banks in branching states had a competitive advantage over nationally
chartered banks. The 1924 revision to the act gave the Federal Reserve
more authority in determining whether or not the state member bank could
expand its branch system.
The fourth development that elevated branch banking to the national
spotlight was a 1924 bill authored by Chairman of the House Committee on
Banking and Currency, Mr. McFadden. The primary emphasis of this bill
was the intrastate branching rights of state and national banks. This
bill, in a revised form, would later become the 1927 McFadden Act whose
details are outlined below.
Though not mentioned explicitly by Preston (1924), another
development during this era that certainly contributed to the branch
banking discussion was the tremendous growth in commercial banks and,
paradoxically, the increased industry concentration as the number of
mergers increased throughout the 1920s. In 1920 there were over 22,000
state banks and over 8,000 national banks. At the end of the decade,
there were 7,530 nationally chartered and 17,440 state chartered banks.
Some of the declining numbers may be explained by the significant wave
of mergers during this period. Between 1920 and 1930 there were 4,101
bank mergers in the United States (White, (1985)).
In addition to the growth and consolidation during the 1920s, there
were three important regulatory developments during this decade. As
mentioned above, in 1918 the federal government passed a law that
allowed national banks to merge with one another. However, the act did
not allow for the easy merging of national banks with other banking
institutions. Indeed, if a national bank wanted to merge with a state
bank it had to first convert to a state charter, merge, and then be a
merged state bank. Since state banks had more intrastate flexibility in
terms of branching (in some states) and fewer regulations, the number of
national charters fell (in some states) as they converted to state
charters.
The second regulatory development in the 1920s was the 1927 passage
of the McFadden Act. Six years later, in 1933, Congress adapted further
significant bank legislation as the banking sector was devastated by
thousands of failures. Both of the regulatory developments are important
to the branch banking history of the United States and are discussed
next.
Alarmed by the number of conversions, in 1927 Congress passed the
McFadden Act which allowed national banks to merge with state banks. The
McFadden Act of 1927 also addressed issues of intrastate and interstate
branching. The act allowed national banks to branch within the cities in
which they were located, if state law permitted. However, the McFadden
act effectively prohibited intrastate branching because a national bank
could not open branches throughout the state as could state charted
banks. In 1927, and for many years thereafter, states prohibited
interstate branching for state chartered banks and the 1927 McFadden Act
extended that prohibition to nationally chartered banks. Thus, the bank
regulation of the 1920s on the one hand, allowed for consolidation
while, on the other hand, confined banks geographically by prohibiting
intrastate and interstate branching for national banks.
In the 1920s the branch issue was brought to the forefront of
national bank policy by two market developments. One was the rapid
increase in branch banks coupled with the conversion of nationally
chartered banks to state chartered banks in order to take advantage of
branch opportunities. Between 1918 and 1926, 206 nationally chartered
banks were converted to state chartered and between 1901 and 1926 the
number of branch units increased from 60 to 2,233 (New York Times,
February 6, 1926, 14). The second market development was the alarming
rate of failure among the smaller unit banks. In 1924, 613 small banks
failed and the following year there were 464 failures (New York Times,
February 6, 1926, 14). These market developments certainly gave the pro
branching group a bit of an advantage in the policy discussion. However,
the political power of the small banker was too much to move policy
beyond intracity branching.
Small bankers continued to oppose extending branch rights. The
evidence suggests that the primary position of the small banker was that
it would be too much competition and that the small banker would be
eliminated leaving behind a monopoly banking system (see table 1).
Because of the market developments mentioned above, by October of 1922
the New York Times indicated that the biggest issue facing the American
Bankers' Association was that of branch banking (October 3, 1922,
6). At their annual meeting, the members of the Association decisively
voted against branching based generally on the fear that branching would
lead to the elimination of the small banker and the monopolization of
banking by a few.
At the 1922 American Bankers' Association meetings, many
bankers representing large bank interests were present to try and
convince the association that branching could actually help all sizes of
commercial banks. While the large bankers argued that branching improves
efficiency and allows banks to better meet the needs of the
non-financial sector, primarily agricultural business (see table 2).
Many of the small banks were already at their limit on agricultural
loans and larger banks, with more capital, could extend more to each
industry. Further, the large bankers were eager to establish branching
rights and argued it would enhance the safety and stability of banking.
In general the large banker was eager to establish branch networks.
There were two significant changes within commercial banking that
certainly contributed to the widespread branching discussions and
legislative developments. First was the rapid pace of conversion of
national to state charters in part because of the branch privileges of
many of the state banks. The second development in the banking sector
was the relative stability of branching over unit systems. Calomiris
(1990) indicates that between 1921 and 1929 only 37 branching banks
failed. He also points out that those who witnessed banking in the 1920s
were struck by the stability of branch systems and that many states
responded by changing their branching laws in response to these
observations. More specifically, between 1924 and 1939, the number of
full or partial branching states increased from 18 to 36; a clear
majority (Calomiris (1990)).
Another market development that cannot be ignored was the invention
and use of the automobile. Collins (1926) discusses how traffic
congestion made it increasingly difficult for customers to reach their
downtown banks. Customers began demanding easier access to the banks
outside of the busy downtown. Bankers wanted to meet this demand by
establishing intra-city branches and, in many cases, intrastate
branches. Further, improvements in telephone networks during this period
enabled branch managers to communicate more effectively and at a reduced
cost. In this way, branching became an increasingly attractive way for
banks to expand.
Despite both the industry consolidation and expansion of national
bank branching rights during the 1920s, the 1930s were a terrible
disaster for U.S. commercial banking. Over 10,000 banks failed between
1921 and 1931. The majority of these took place after the passage of the
McFadden Act and over 80 percent of all failures were small, unit banks
(Jay, 1933). Given the banking crisis, it is not surprising that
regulatory reform was called upon. The Banking Act of 1933 is a landmark
piece of commercial bank legislation in terms of altering the
operational parameters of banks. Among other things, the 1933 Act
expanded branching rights for nationally chartered banks from the city
level in the McFadden Act to the state level. The Banking Act of 1933,
to a certain extent, liberalized branch banking laws for nationally
chartered banks by allowing them the same branching rights, within
state, as state banks.
During the late 1920s and early 1930s, the vast majority of
commercial banks were single-office unit banks. As was the case in
earlier years, their numeric strength and dominate control over most
banker associations put the unit bankers in a position to pressure
regulators and legislators. The unit banker feared that if national
banks were allowed to establish branch offices in their communities,
they would be unable to compete with the larger banks (see table 1). The
argument was that the branch system of larger banks would distance the
banker from the needs of the local communities thereby leaving the
public in a disadvantaged and insecure position.
Using their strength in numbers and their voice in influential
industry institutions, such as the American Bankers' Association,
the small commercial banker fought to keep the larger banks from
entering most markets though out the country. They were motivated, in
large part, by the fear that they would be unable to compete in a branch
system.
Most large bankers were in favor of a branch system as a means of
providing stability and prosperity to commercial banking.
Representatives of large commercial banks argued that the smaller banks
lacked necessary capital and were unable to properly diversify in order
to be profitable (see table 2). Further, many large bankers argued that
the process of intrastate and interstate branching would offer the
diversification levels necessary for prosperity. Large bank advocates
fought for liberalized branching rights based on the argument that the
financial community needed to adapt to the changing conditions of the
business community. During the late 1920s, the United States witnessed
impressive growth in technology, transportation, and energy services.
Additionally, industrial enterprises were changing in structure, due to
competitive pressure, by becoming highly integrated and hence larger.
In the end, the coalition of small bankers was able to successfully
fend off any further erosion in interstate banking. While the large
banker enjoyed some small victories in this era with the intra-city
banking in 1927 and then intrastate banking in 1933, interstate banking
continued to be off limits.
Modern Era
The post depression years were rather quiet for commercial banking
and while the interstate banking issue was kept alive it was primarily
in academic circles. No serious attempts at changing legislative
constraints were considered after the depression legislation until the
1990s. There were pockets of national interest over this time frame but
interstate banking was not the primary issue in legislative and
regulatory talks. More specifically, as market developments revealed
weakness and fragility in the 1960s banking sector, the branch issue
once again captured the interest of policymakers but nothing came of it.
Major banking legislation was implemented in the 1980s but much of this
was not primarily related to interstate banking. It was not until the
early 1990s that the focus was directly on interstate banking. The
following provides some detail of the evolution of the issue in the
modern era.
The branch discussion resurfaced in the early 1960s when James
Saxon, Comptroller of the Currency, called for greater branching
opportunities for nationally chartered banks. Saxon was concerned that
the limits on branching would hinder macroeconomic growth. As population
growth shifted to suburban areas the Comptroller was concerned that the
1933 legislation was often an obstacle for meeting the banking needs of
the changing demographics. This was particularly true in states that
either prohibited branching or limited branching.
As word of Saxon's proposal spread, branch advocates and
opponents spoke in all too familiar terms. Small bankers resumed the
argument of their predecessors as they were concerned that more
branching by national banks would lead to an increase in bank
concentration by eliminating many of the small banks. Large bankers
favored expanding branch opportunities by arguing branching was more
efficient, lowered costs, improved profits and consequently improved the
stability of banking.
In the 1980s, regulators and legislators again began discussing
branching but not directly. Rather, the concern was over the growth of
"non-bank banks". In January of 1986, the Supreme Court
limited the Federal Reserve's regulatory authority over non-bank
banks. The Bank Holding Company Act of 1956 defined a bank as an
institution offering both demand deposits and commercial loans. Non-bank
banks could offer one, but not both, services. In this way, they did not
fit the legal definition of a bank and so could cross state lines.
According to financial experts at the time, the court ruling essentially
opened the door to interstate banking. Thus while Congress was, in the
1980s, actively engaged in attempts to make important changes to
commercial banking, the focus was not directly on branch banking laws
outside the impact that non-bank banks would have on the interstate
banking ban.
By the early 1990s, regulators and legislators could no longer
ignore the branching issue because of important developments in the
market that effectively forced the hands of these policymakers. Faced
with increasing bank failures, a depleted deposit insurance fund, a
rising number of non-bank banks, and the increasing number and size of
regional banking compacts, Congress began looking at the interstate
banking ban once again. In 1991 the issue was moved closer to real legal
change than the efforts of the 1980s but still fell short.
Unlike the 1980s when branch banking was discussed primarily
through indirect channels, in 1991 legislators and regulators pushed
hard for interstate branching rights. However, the branching proposal
was tied to a proposal to reduce insurance authority previously granted
to commercial banks. In the end, this kept legislative leaders from
lifting the branching ban. By 1994, market developments progressed to
the point where they could no longer be ignored. Further, a bill passed
by Congress required large banks to buy, at least for a time, a bank in
another state before opening branches. This acquisition requirement
appealed to the small bankers because it added value to their franchise.
In many cases, this provision was enough to get the small banker to
retreat somewhat in their opposition to the bill (New York Times,
September 14, 1994, D1).
The debate in the modern era largely saw the small banker make many
familiar arguments against interstate banking. For example, many small
bankers continued to fight on the grounds that they would be driven out
by big banks entering the market (New York Times, September 30, 1991,
D5) (see also table 1). However, one development does set the modern era
apart from earlier eras in that some small bankers (although certainly
the minority) recognized publically that interstate banking may actually
have benefits. Evidence of this change was found both in the 1991 debate
and again in 1994. For the first time in the interstate banking
dialogue, some small bankers actually spoke out in what may be described
as lukewarm acceptance of this idea. A columnist for the American
Bankers Association Banking Journal wrote:
Legislation easing restrictions on interstate banking will promote
continued consolidation in the banking industry, but the impact on
community banks will not necessarily be negative. The number of US banks
grew steadily as the country expanded westward and has been shrinking
just as steadily since then. Community banks will not necessarily
disappear in this consolidation, but they will have to become more
competitive. (Smith (September 1994, 17)).
A small banker from Missouri was also optimistic as early as 1991
when he commented "When this state went to statewide bank holding
companies, people were saying we might as well hang up 'For
Sale' signs. That was 20 years ago and we're doing better now
than ever." (Cocheo (April 1991)).
The fact that some in the small bank community were speaking in
relatively positive terms about the possibility of interstate branching
was a significant turning point. Since the chartering of the Second Bank
of the United States in 1816, the small banker strongly and unwaveringly
opposed interstate banking. It is likely that their experience with
interstate branching through bank holding companies, which did not prove
to destroy them, gave them the confidence to recognize that there was a
place for both small and large institutions in commercial banking.
Further, the large number of bank failures in the early nineties led
many bankers, both big and small, to realize that a more diverse banking
system could have withstood the downturn in both real estate and energy
that led many banks into trouble (Bacon, 1994).
The large bankers once again pushed for the right to establish
branches across the country. They argued that this would greatly reduce
their costs by eliminating duplicative officers and functions (New York
Times, February 4, 1994, D14). In 1991 Bank of America indicated that
branching would allow the bank to save millions of dollars by
eliminating duplication in labor, capital, and functions (Cocheo (April
1991)). In 1993, the CEO of BankAmerica Corp. estimated the costs
savings more specifically stating that interstate branching could save
the bank $50 million annually (Cocheo (January 1993)). In addition to
cost savings, large bankers advocated that interstate branching would
strengthen the banking system (see table 2).
Perhaps the most important factor behind the legislative lifting of
the interstate banking ban was developments in the market that took
place outside the halls of Congress. By the 1990s, commercial banks had
successfully pressed into what was essentially interstate banking.
Consider, for example, that at the end of 1991, all but two states
permitted out-of-state companies to own banks within their borders (CQ
Weekly Online (January 4, 1992)). By the middle of 1993, only Hawaii
prohibited bank ownership by out of state bank holding companies (LaWare
(1993)). Thus states had made it clear that they accepted interstate
branching by creating these regional banking compacts throughout the
country. Indeed, by the fall of 1994, there were 242 multibank holding
companies in the United States (Kane (1996)).
Another market development in the modern era that allowed banks to
effectively by-pass the interstate branching ban was the birth of
electronic banking. Initially, this took the form of establishing a
large network of automatic teller machines that allowed customers to
engage in limited banking activity across state lines. Further, some
banks were beginning to offer computerized banking which could also
cross state lines. Thus, by the time the debate came to an end and
intrastate branching was legally recognized, commercial bankers had
essentially moved beyond the issue by practicing branching despite the
national limitations.
EMPIRICAL ANALYSIS OF INTEREST GROUP POSITIONS
As the previous section of this paper indicates, the interstate
branching debate was a long one and the banking environment evolved in
significant ways throughout the life of the debate. One relative
constant, however, was the position of the interest groups and the
rationale behind their position. This section of the paper attempts to
empirically test the validity of these positions.
Much has been written about the impact of bank regulation on bank
performance (see, for example, Barth et al. (2004), Calomiris (2000),
Levin et al. (1999), Kaufman (1996)), on how interest groups influence
the regulation and policy process (see, for example, Becker (1983),
Stigler (1971), McChesney (1997), Heclo (1978), Kroszner and Strahan
(1999)), and more specifically on how branch banking impacts bank
performance (see, for example, Amos (1992), Jayaratne and Strahan
(1999), Carlson and Mitchener (2006), Ramirez (2003)). This research
draws, to some degree, from all these types of existing literature to
formulate the models presented here. At the same time, this research
extends the body of literature in several ways. First, this paper
considers how branch banking impacts bank performance across several
historical eras. In contrast, most previous work utilizes either Great
Depression data or data from the 1980s forward. This is probably due
largely to data availability as it is difficult to create a
comprehensive, pre-depression data sample. Second, this research
carefully establishes the interest group positions and then empirically
tests these over time. This research finds support for the small bank
contention that branching may produce less competitive outcomes. At the
same time, there is evidence to support the large bankers' position
that branching enhances stability.
This analysis draws on previous empirical work that considers how
branching impacts performance or market structures, something that both
the small and large banker tried to use to influence branch policy. Amos
(1992) was interested in learning more about the relationship between
branching and bank closures in different states. His ordinary least
squares (OLS) specification finds limited evidence to suggest that
states which completely restrict branching and states with intrastate
branching had more bank failures in the 1980s. Cebula (1994) extended
the analysis of Amos by controlling for the financial condition of
banks, by creating a single dummy branch variable, and by extending the
analysis from the 1980s into the early 1990s. Cebula also finds that
states with limited branching experience fewer bank failures which is
consistent with comments in Amos (1992). Loucks (1994) takes the Amos
relationships but estimates them with the Tobit estimator rather than
OLS because some of the dependent variables in Amos are truncated at
zero. Louck finds a smaller percentage of bank closures in states with
less branch restrictions. Ramirez (2003) takes a more micro approach to
the analysis of branching and bank failures by comparing the experience
of a branching state with a non-branching state in the 1920s. He finds
evidence to indicate more bank failures associated with the unit banking
environment. Further, Ramirez suggests that this is because branching
reduces costs allowing for larger banks who enjoy more diversified
balance sheets.
Other literature controls for the economic diversity of the area in
which banks operate. Throughout history, the large banker consistently
made the case that branching enhances the diversity of a bank's
portfolio and, in doing so, makes the banking system more stable and
profitable. Shiers (2002) creates a measure of economic diversity and
empirically tests whether this diversity impacts bank risk and profits.
For the period 1966 to 1981, Shiers finds that economic diversity
reduces bank risk and he also finds similar findings for the time period
1982-1996. More recently, Carlson and Mitchener (2006) test how the
growth in branching during the Great Depression influenced bank
competition and how the competitive changes affect bank failures. Their
hypothesis is that as branching expands, competition increases which
leads to exit by banks less able to compete so the banking sector
becomes more stable. The empirical findings of Carlson and Mitchener
(2006) suggest that, for nationally chartered banks, the competition
hypothesis contributed more to stability than did an increase in
portfolio diversification. This suggests that the arguments for
diversification were perhaps not as valid during the Great Depression
era as were the fears of the small banker who anticipated they would be
unable to compete with the bigger banks.
This study uses annual aggregate state data for an eleven year span
within each banking era. For each banking era, the data is pooled across
the states and all bank data includes both state and nationally
chartered banks. Ideally this empirical investigation would break the
commercial bank data (during the first two eras) into two groups: one
for nationally chartered banks and another for state chartered banks.
This would be extremely helpful since in these earlier years different
chartered banks operated in different branching environments. For
example, during the national banking era, some state banks were allowed
to branch within state while national banks were not. However, neither
failure nor merger data at the state level for both charter types has
been found for each year of the sample periods.
In terms of failure data, the Annual Report of the Comptroller of
the Currency records the number of national bank failures by state. The
number of state bank failures in each state, in contrast, is not
available for each year of the national banking era. There are several
sources in addition to the publications of the offices of the
Comptroller that provide historical bank failure data. These include
Goldenweiser (1933), Banking and Monetary Statistics, All-Bank
Statistics, and historical issues of the Federal Reserve Bulletin. In
the end, however, there is not currently available bank failure data by
state and charter type for both the national banking era and Great
Depression era.
In terms of bank merger data, like the failure data, there exists
several data sources including Chapman (1934) and the Federal Reserve
Bulletin (1937). While some of this historical merger data is at the
state level, it does not break down into both charter types for both the
earlier bank eras. Because of failure and merger data limitations, this
paper uses state level banking data without distinguishing between
charter types (see Appendix A for all data sources and definitions).
The national banking era sample covers 1900 to 1910, the Great
Depression era sample runs from 1924 to 1934, and the modern era sample
is from 1984 to 1994. These dates were chosen, in part, due to data
constraints but also because of the timing of events such as regulatory
changes or banking crises. For example, both the national banking era
and Great Depression era experienced financial crises. The national
banking era sample years were chosen because there was only one crisis
during this time frame and no significant changes in bank regulation. It
is nearly impossible to avoid bank crisis or regulatory change in the
Great Depression era, unless the analysis ended in or before the 1927
passage of the McFadden Act. In this analysis, the 1933 crisis is
controlled for using a dummy specification for the crisis year and the
branching regulatory environment is also controlled for. In the modern
era, the sample begins after significant bank regulatory change and ends
with the passage of the Riegle-Neal Interstate Banking and Branching
Efficiency Act of 1994.
For each of the banking eras with identified interest group
positions, three interest group positions are specified. First, the
small banker unfailingly argued that an extension of branching rights
would lead to increased market concentration and monopoly power. The
first model attempts to capture the impact of branching on competition.
Second, the large banker insisted that allowing for more liberalized
branching would make the banking system more stable. Thus, model two
captures the impact of branching on bank failures. Finally, the large
banker consistently argued that branching would allow them to diversify
their portfolios and, consequently, be able to withstand macroeconomic
downturns. The failure model contains an explanatory variable meant to
capture diversity. Consequently, inferences will be made regarding the
relationship between the ability of banks to diversify and its impact on
bank failures.
In constructing the models, it is necessary to control for the
financial health of the banks (in the failure model), the market
structure the banks are operating in, the regulatory environment
influencing banker behavior, and the economic health of the state. Model
specifications and variable definitions follow.
Competition Model
The small banker spoke loudly and often throughout the branching
debate and their message was consistent: the extension of bank branching
will lead to less competition and even monopoly banking. To test the
validity of this point of view, a competition model is constructed in
which the dependent variable captures the extent to which each state
banking market was competitive (see equation 1 where i represents the
state and t represents the year):
(1) [COMPETITION.sub.it] = [[beta].sub.0] + [[beta].sub.1]
[BRANCHING(-1).sub.it] + [[beta].sub.2] [NUMBER(-1).sub.it] +
[[beta].sub.3][YEAR.sub.t] + [[beta].sub.4][CRISIS.sub.t] +
[[beta].sub.5][STATEECONOMY(-1).sub.it] + error
Following Carlson and Mitchener (2006) COMPETITION is defined as
the number of bank mergers. Ideally, a measure of competition could be
created that more directly measures competition. For example, a
Herfindahl-Hirschman Index would be ideal. However, data constraints
prohibit this. In terms of the independent variables, the model tests
two specifications to control for the regulatory environment within each
state (labeled BRANCHING(-1) in equation 1). One specification, BRANCH,
utilizes the one period lag of the number of banks operating branches to
control for the level of branching in each state. During the national
banking era, if two national banks wanted to merge, one bank was
liquidated while the other purchased its assets and liabilities. State
banks could often merge, but could not merge with a national bank unless
the liquidation process described between two national banks was
employed. Alternatively, a state bank and national bank could merge by
taking out a state charter. Because of these difficulties, there was not
a significant amount of merger activity in this early banking era. In
terms of the impact of branching, state banks did use merging as a means
of establishing branches so we may expect a positive relationship
between branching and merging. During the Great Depression era, national
banks were allowed to merge and consolidate without liquidation provided
the Comptroller approved. The hurdles to state and national bank merging
were in place until the passage of the McFadden Act of 1927 which
allowed national banks and state banks to merge under the same rules. As
in the earlier era, many banks used mergers to establish branch
networks. Consequently, it is expected that this measure of branch
activity, BRANCH, to positively impact the degree of concentration in
the banking market.
For the alternative branch specification, Shiers (2002) suggests
UNIT, the lag of the ratio of the total number of unit banks to the
total number of banks, as a measure of the extent of non-branching in
each state. This specification captures the opposite of BRANCH and so is
expected to have a negative impact on this measure of competition. These
two alternative specifications, BRANCH and UNIT, used to control for
branching are utilized in the all three eras. However, in the modern
era, more specific data indicating the date branch laws changed in each
state are available (Stiroh and Strahan, 2003). Consequently, in the
modern era, an additional regression is analyzed using controls for the
legal changes in branch laws (results in table 6).
The remaining independent variables control for the state economy,
the market structure, and time-specific effects. %BUSFAILURE and GSP
control for the performance of the state economy (labeled
STATEECONOMY(-1) in equation 1). During the two earliest eras, the state
economy is proxied as the lag of the percent of business failures in
each state. During the modern era, this is measured as the lag of the
gross state product. Certainly, the healthier the state economy, the
more likely the banking sector is also healthy. A strong economy may
signal more opportunities for mergers as banks expand and have the
financial ability to purchase other banks. Another possible scenario is
one in which there will be more mergers when the state economy is
struggling as banks find themselves in weaker positions. From this
perspective, bank mergers may be a way for weaker banks to avoid
failure. Consequently, the impact of the state control on merger
activity is uncertain. To control for the market structure that the
banks are operating within, this paper relies on the lag of the number
of banks, NUMBER, operating in each state. An increase in the number of
banks would lead one to expect fewer bank mergers since merging reduces
the number of bank entities. Indeed, Berger, Kashyap, and Scalise (1995)
find an increased reduction in the number of banks following the
introduction of branch opportunities. Since more branching tends to
come, in part, from merging, NUMBER is expected to negatively impact
COMPETITION.
YEAR controls for time-specific effects. Given that, with the
passage of time, merging became easier in the Great Depression and
modern era, it is expected that YEAR will positively impact the number
of mergers, particularly in the later two eras. CRISIS controls for any
financial crises during the sample period.
Failure Model
The large banker long insisted that more opportunities for
branching would improve competition and leave the more efficient and
stronger performers standing. From this perspective, branching makes
banking more stable, at least after the weaker banks have exited. To
test the impact of branching on stability, this paper relies on a
failure model since there should be fewer failures, after a period, with
opportunities to branch:
(2) [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]
The dependent variable, FAILURE, is defined as the number of bank
failures in each state. BRANCHING, NUMBER, STATEECONOMY, CRISIS, and
YEAR are defined as in the COMPETITION model.
To control for the financial health of the banks, three ratios are
utilized. The lag of the capital to asset ratio, CAPASSET, indicates the
extent to which a bank is able to withstand unexpected adverse
conditions (Ramirez (2003), Cebula, (1994)). A lower capital to asset
ratio indicates more risk and consequently more failures. LOANASSET, the
lag of the ratio of loans to assets, is an indicator of bank risk since
loans are typically the most risky asset on a bank balance sheet. As
suggested in Ramirez (2003), the higher the ratio of loans to assets,
the greater the likelihood of failure. Deposits are usually the cheapest
source of funds for a bank and, according to Ramirez (2003), banks with
large deposit bases are seen to be financially stronger given the cost
advantage from those deposits. Consequently, it is expected that the
larger the lag of the ratio of deposits to assets, DEPASSET, the fewer
failures. Finally, DIVERSITY is defined as the lag of the ratio of the
total number of branches to total number of banks in each state. Carlson
and Mitchener (2006) use a similar measure and point out that more
opportunities to branch should improve balance sheet diversification and
therefore reduce failures.
Given that the data in this research contains both cross-sectional
(state) observations and time-series (year) observations, the estimation
procedure must be appropriate to this panel data. There are several
types of models for panel data. These include the constant coefficient
models, fixed effects models, and random effects models. For this
research, significant state differences are expected because of
different branch laws so the constant coefficient model is not
appropriate. To determine whether the fixed or random effects model
should be used, it is common to use the Hausman specification test
(Wooldridge (2002)). Doing so indicates that testing the impact of
branching on competition and bank failures should be done utilizing the
fixed effects estimation strategy. The correlation matrix detects no
significant multicollinearity in the specifications. Further, each
specification is tested for the presence of heteroskedasticity using the
Breusch-Pagan test. However, the DW statistic indicates the presence of
autocorrelation. This is corrected for utilizing an AR estimator.
RESULTS
Tables 3 through 6 contain the regression results for the fixed
effects COMPETITION model for the national banking era, Great Depression
era, and modern era respectively. These findings are consistent with
existing literature that finds branching leads to more bank mergers.
Specifically, Carlson and Mitchener (2006) find that states with branch
systems witnessed more mergers during the Great Depression and White
(1985) analyzed bank mergers in the 1920s and suggests that there is a
positive relationship between branching opportunities and merger
activity. The results in table 4 are consistent with this existing Great
Depression era research and the results in table 3 indicate that
branching in the pre-depression era also increased mergers, and
consequently, consolidation in a state's banking system. The
estimation results in table 5 are consistent: even in the modern era,
the estimated coefficients for both branch specifications support the
hypothesis that more branching leads to more banks mergers. These
results validate the small bankers' position which was to
consistently fight attempts to expand branching on the grounds that it
would lead to highly concentrated markets.
The results also suggest that the performance of each state economy
in the two early eras has a different influence on COMPETITION than the
modern era. More specifically, in both the national banking era and
Great Depression era the estimated coefficients on %BUSFAILURE indicate
a reduction in bank mergers, i.e. more competition, during more unstable
economic times. In contrast, the results in the modern era suggest that
slower economic growth corresponds to more mergers, i.e., less
competition. Yet, the results are only statistically significant in the
modern era and in one regression in the national banking era. Some of
these differences may reflect different measures (the lag of the percent
of commercial failures in the two early eras and the lag of gross state
product in the modern era). Since gross state product is a more
comprehensive measure of a state's economic health, the results of
the modern era may be more indicative of the relationship between the
state economy and level of bank merger activity.
These results suggest that the structure of the market also matters
in terms of bank merger activity. In each regression, the control for
market structure, NUMBER, is of the expected sign and statistically
significant. This is consistent with existing research that finds merger
activity to be related to the number of banks in the market.
Finally, the control for time-specific effects is statistically
significant across the three eras. Further, the estimated coefficients
indicate a positive relationship between time and merger activity, as
expected. The control for bank crisis in the time series is
statistically significant in the national banking era suggesting the
crisis impacted merger activity but the crisis control was not
significant in the Great Depression era.
In each era, the COMPETITION model was regressed first without the
state control and then with it (regression 1 and regression 2 in tables
3 through 6). While the state control was sometimes statistically
significant, it does not appear to contribute much to the overall
explanatory power of the regression models.
Table 6 contains the regression results utilizing a different
branch specification for the modern era. In this most recent era, it is
known with more precision when both intrastate and interstate branching
laws were altered. This information allows for the creation of dummy
variables that reflect the legal changes in branch laws. The results
indicate that allowing for both types of branching increase merger
activity but only the interstate control is statistically significant.
It is possible that by the 1980s and 1990s, banks had exhausted much of
branching through mergers within state but were still merging across
state lines since interstate banking typically came later than
intrastate banking.
Tables 7, 8, and 9 contain the regression results for the FAILURE
model for the Great Depression era and modern era respectively. Because
the author does not have access to state level failure data during the
national banking era, this model cannot be estimated for the earliest
period. While the results of the COMPETITION model provide support for
the small bankers' position against the expansion of branching, the
FAILURE model provides confirmation of the large banker position that
branching stabilizes banking. More specifically, all specifications of
branching are of the predicted sign and indicate that more branching
opportunities decrease bank failures. It is interesting, however, that
the branch specifications are not consistently significant in the Great
Depression era. It may be that during the depression, the impact of the
wider economy mattered more than the branching structure in explaining
bank failures. Given the severity of the economic downturn in this era,
this would not be too surprising. Regardless, the results of the FAILURE
model clearly indicate that the large banker was correct on the branch
issue: branching reduces bank failures.
Four other observations from tables 7 through 9 are worth noting.
First is that DIVERSITY is not statistically significant in either era
or in any model specification. This suggests that, at least for this
particular measure, diversification of bank balance sheets may have been
less of a contributor to bank stability than perhaps the increased
competition from branching. That is, the branching may have weeded out
the weaker banks leaving a stronger system in place. This finding is
consistent with the research of Carlson and Mitchener (2006) who find
that, during the Great Depression, branching was more important to bank
stability than diversification. Second, in all FAILURE regressions, the
control for the health of the state economy was an important factor in
explaining bank failures. The health of the economy clearly contributes
to the health of the commercial banking sector. Third, the financial
health of the bank affects bank failures. However, in the Great
Depression era, the failures are explained more by the cost of deposits
and the reliance on loans whereas in the modern era, the level of bank
capital is statistically more valuable in explaining the failures. The
fourth and final observation is that the ratio of total deposits to
total assets, DEPASSET, is of the expected sign in the Great Depression
but not in the Modern Era. Recall that the argument was that deposits
are the cheapest source of funds and so could provide a cost advantage
to banks with a higher ratio. The results in tables 8 and 9 indicate the
opposite during the modern era. It may be that liability competition in
the most recent era, from both commercial banks and other financial
institutions, has eroded the cost advantage of deposits.
CONCLUSIONS
This paper carefully identifies the small and large banker position
on the issue of branch banking throughout U.S. history. Once identified,
this research attempts to empirically test the positions in three
distinct historical eras. The findings indicate that both the small
banker and large banker had valid arguments. Specifically, the
estimation results confirm the small bankers' position that
branching leads to market concentration and reduced competition. While
the small banker often spoke of monopolies, this paper does not attempt
to establish the degree of market concentration but, rather, the change
in the level of competition as measured by the number of bank mergers.
Certainly, the threat of a pure monopoly outcome has not been validated.
Rather, across all indentified bank eras, the results indicate that
branching increases merger activity. Despite these results, it is
important to note that merging need not increase monopoly power,
especially if combined with an expansion of branches. Indeed, it is
possible that the combination of more mergers and more branches could
result in a more competitive environment.
There is also evidence that the large banker was right. The large
banker consistently argued that branching would allow banks to become
more stable and hence less prone to failure. The estimation results in
this paper validate this position. However, the results within this
paper also indicate that for a particular measure of diversity it seems
that bank failures are not explained by a banks' ability to
diversify. Previous research by Carlson and Mitchener (2006) analyze
this question using data from the Great Depression and they find
evidence that branching contributed quantitatively to the stability of
banking rather than geographic diversity. In the end, it appears that
the large banker was correct that branching would enhance stability but
perhaps because it would weed out the smaller and weaker banks and not
because it would improve the diversification of the balance sheet.
These findings paint a picture of a banking system that has fewer
banks but more stability. A glimpse at the current market structure is
consistent with this picture. Since the 1994 regulatory change that
allowed for interstate banking across the country, the number of banks
has steadily fallen. Between 1994 and 2007, the number of commercial
banks has decreased just over 30% (fdic.gov). However, the number of
branches during the same period has increased just over 43%. At the same
time, the banking sector has enjoyed a period of stability as the number
of bank failures has averaged 4.25 failures per year in the post
interstate branching era (through 2007). However, it appears that the
commercial banking sector, and wider financial sector, is sure to
undergo significant change from this point forward given the turmoil of
2007 and 2008 and ongoing instability in 2009.
REFERENCES
All-Bank Statistics: U.S. 1896-1955 (1959).
Annual Report (various years). Office of the Comptroller of
Currency, Washington, D.C.
Amos, Orley M. (1992). The Regional Distribution of Bank Closings
in the United States from 1982 to 1988. Southern Economic Journal,
58(3), 805-815.
Bacon, Kenneth H. (1994, February 25). Nationwide-Bank Bill Picks
Up Steam As Even Opponents See Measure Passing, Wall Street Journal, A3.
Banking and Monetary Statistics, 1914-1941 (1976). Washington,
D.C.: Board of Governors of the Federal Reserve.
Barth, James R., Gerard Caprio, Ross Levine (2004). Bank Regulation
and Supervision: What Works Best? Journal of Financial Intermediation,
13(2), 205-248.
Becker, Gary S. (1983). A Theory of Competition Among Pressure
Groups for Political Influence. The Quarterly Journal of Economics,
XCVIII(3), 371-400.
Berger, Allen, Anil Kashyup, and Joeseph Scalise (1995). The
Transformation of the U.S. Banking Industry: What a Long, Strange Trip
It's Been. Brookings Papers on Economic Activity, 55-218.
Bodenhorn, Howard (2003). State Banking in Early America: A New
Economic History. New York: Oxford University Press.
Boeckle, R.M. (1937). Expansion of Branch Banking. Editorial
Research Reports, Washington, DC: CQ Press.
Bradford, Frederick, A. (1930). Proposed National Banking
Legislation. The American Economic Review, 20(3), 400425.
Calomiris, Charles W. (1990). Is Deposit Insurance Necessary? A
Historical Perspective. The Journal of Economic History, 50(2), 283-295.
Calomiris, Charles W. (2000). U.S. Bank Deregulation in Historical
Perspective. Cambridge: Cambridge University Press.
Carlson, Mark (2004). Are Branch Banks Better Survivors? Evidence
From The Depression Era. Economic Inquiry, 42(1), 111-126.
Carlson, Mark, and Kris James Mitchener (2006). Branch Banking,
Bank Competition, and Financial Stability. Journal of Money, Credit, and
Banking, 38(5), 1293-1328.
Cebula, Richard J. (1994). The Regional Distribution of Bank
Closings in the United States: An Extension of the Amos Analysis.
Southern Economics Journal, 61(1), 202-208.
Chapman, John M. (1934). Concentration in Banking. New York:
Columbia University Press.
Chapman John M. and Ray B. Westerfield (1942). Branch Banking: Its
Historical and Theoretical Position in America and Abroad. New York:
Harper & Brothers Publishers.
Cochero, Steve (1991, April). Banking Reviews the Blueprints. ABA
Banking Journal, 83(4), 33-37.
Cochero, Steve (1993, January). What Do you Get For $50 Million?
Zip. ABA Banking Journal, 85(1), 33-36.
Collins, Charles Wallace (1926). The Branch Banking Question. New
York: The Macmillan Company.
Congressional Globe. (1836). The Deposit Bill. Senate, 24th
Congress, 1st Session, 500.
Congressional Quarterly Weekly Online. (various years), from
http://library.cqpress.com/cqweekly.
Congressional Record. (various years).
Cranford, John R. (1992, January). Marketplace Forces Hold Key To
Future Bank Legislation. CQ Weekly Online, 2, 9-14. Retrieved from
http://library.cqpress.com/cqweekly.
Federal Reserve Bulletin. (various years).
Federal Reserve (1943). Banking and Monetary Statistics, 1914-1941,
Washington, DC: Board of Governors of the Federal Reserve System.
Giedeman, Daniel C. (2005, March). Branch Banking Restrictions and
Finance Constraints in Early-Twentieth-Century America. The Journal of
Economic History, 65(1), 129-151.
Goldenweiser, E.A. (1931). Federal Reserve Committee on Branch,
Group and Chain Banking: Statistical Data. Unpublished Manuscript,
Federal Reserve System.
Goldenweiser, E.A. (1933) Bank Suspensions in the United States,
1892-1931. Unpublished Manuscript, Federal Reserve System.
Hammond, Bray (1957). Banks and Politics in America, from the
Revolution to the Civil War. Princeton: Princeton University Press.
Heclo, Hugh (1978). Issue Networks and the Executive Establishment.
In Anthony King (Ed.), The New American Political System, Washington DC:
American Enterprise Institute for Public Policy Research.
Jay, Pierre (1933, January). The Structure of the Banking System.
Proceedings of the Academy of Political Science, 15(2), 11-21.
Jayaratne, Jith and Philip E. Strahan (1999). The Benefits of
Branching Deregulation. Regulation, 22(1), 8-16. Kane, Edward J. (1996).
De Jure Interstate Banking: Why Only Now? Journal of Money, Credit and
Banking, 28(2), 141-161.
Kaufman, George S. (1996). Bank Failures, Systemic Risk, and Bank
Regulation. Cato Journal, 16(1), 17-45.
Kroszner, Randall S., and Philip E. Strahan (1999, November). What
Drives Deregulation? Economics and Politics of the Relaxation of Bank
Branching Restrictions. The Quarterly Journal of Economics, 114(4),
1437-1467.
LaWare, John P. (1993). Statements to the Congress. Federal Reserve
Bulletin, 772-780.
Levin, Ross, Gerard Caprio, James R. Barth (1999). Banking Systems
Around the Globe: Do Regulation and Ownership Affect Performance and
Stability? Research Working Paper Series, The World Bank.
Loucks, Christine (1994). The Regional Distribution of Bank
Closings in the United States from 1982 to 1988: A Brief Note. Southern
Economic Journal, 61(1), 191-194.
McChesney, Fred S. (1997). Money for Nothing: Politicians, Rent
Extraction, and Political Extortion. Cambridge: Harvard University
Press.
Newfang, Oscar (1901, December 30). Feasibility of Branch Banks.
New York Times, D9.
New York Times. (various years).
Preston, H.H. (1924, September). Recent Developments in Branch
Banking. American Economic Review, 443-462.
Ramirez, Carlos D. (2003). Did Branch Banking Restrictions Increase
Bank Failures? Evidence from Virginia and West Virginia in the Late
1920s. Journal of Economics and Business, 55(4), 331-352.
Shiers, Alden F. (2002). Branch Banking, Economic Diversity and
Bank Risk. The Quarterly Review of Economics and Finance, 42, 587-598.
Smith, Daniel R. (1994). After Interstate: 8 banks or 8,000? ABA
Banking Journal, 86(9), 17. Southworth, S. (1928). Branch Banking in the
United States, New York: McGraw Hill.
Sprague, O.M.W. (1903, February). Branch Banking in the United
States. Quarterly Journal of Economics, 17(2), 242-260.
Stigler, George J. (1971). The Theory of Economic Regulation. Bell
Journal of Economics and Management Science, 2, 3-21.
Stiroh, Kevin J., and Philip E. Strahan (2003, October).
Competitive Dynamics of Deregulation: Evidence from U.S. Banking.
Journal of Money, Credit, and Banking, 35(5), 801-828.
U.S. Comptroller of the Currency (1876). Annual Report of the
Comptroller of the Currency. Government Printing Office, Washington,
D.C.
United States Department of Commerce (1976). Historical Statistics
of the United States Colonial Times to 1970. Part 2.
Wall Street Journal. (various years).
Wernette, John (1932). Branch Banking in California and Bank
Failures. Quarterly Journal of Economics, 46(2), 362-65.
Wheelock, David C. (1992, December). Regulation and Bank Failures:
New Evidence From the Agricultural Collapse of the 1920s. The Journal of
Economic History, 52(4), 806-825.
Wheelock, David C. (1995, March-April). Regulation, Market
Structure, and the Bank Failures of the Great Depression. Review,
Federal Reserve Bank of St. Louis, 27-38.
White, Eugene Nelson (1982, March). The Political Economy of
Banking Regulation, 1864-1933. The Journal of Economic History, 42(1),
33-40.
White, Eugene Nelson (1983). The Regulation and Reform of the
American Banking System, 1900-1929. Princeton, New Jersey: Princeton
University Press.
White, Eugene Nelson (1985). The Merger Movement in Banking,
1919-1933. Journal of Economic History, 45(2), 285-291.
Wooldridge, Jeffrey M. (2002). Econometric Analysis of Cross
Section and Panel Data. Cambridge, MA: The MIT Press.
Jill M. Hendrickson, University of St. Thomas
Table 1 Evidence of the Small Banker Position
Argument Evidence Bank Era
1. Branching 1. Small bankers from Kansas National
leads to proclaimed "we hereby affirm our Banking Era
monopoly unanswering allegiance to that
banking. view of the proposition which
condemns it (branching) in all
its forms as being unpatriotic,
un-American, un-businesslike,
and as tending to establish a
monopoly of the great and
honored business of banking in
the hands of a few millionaires,
to the exclusion of the men of
the West, old and young, who
have labored so faithfully and
well to make our banking system
what it is to-day, the best in
the known world." (New York
Times, May 15, 1902)
2. A group of small bankers from
Missouri condemned branching as
" ... it to be unwise, unsafe,
unsound, and detrimental to the
banking interests of the United
States." (New York Times, May
15, 1902)
3. Declaration: "Resolved by the Great
American Bankers' Association, Depression Era
That we view with alarm the
establishment of branch banking
in the United State and the
attempt to permit and legalize
branch banking; that we hereby
express our disapproval of and
opposition to branch-banking in
any form by State or national
banks in our nation."
"Resolved that we regard branch
banking or the establishment of
additional offices by banks as
detrimental to the best
interests of the people of the
United States. Branch banking is
contrary to public policy,
violates the basic principles of
our Government and concentrates
the credits of the nation and
the power of money in the hands
of a few." (New York Times,
October 5, 1922, 8)
4. A small bank representative:
"Mr. Frame, speaking first, said
he would speak on "Monopoly vs.
Democracy in Banking." Asserting
that the 30,000 independent
banks of the United States had
done more to build the country
than "all the cream- skimming
monopolistic banks have done for
other nations," he denied the
contention of pro-branchers that
branches served the people
better or that failures and
losses to depositors are
lessened under branch banking."
(New York Times, October 5,
1922, 8)
5. A representative of the
American Bankers' Association
warned that: "Branch banking
will drive the small,
independent community banker
from America ... and in their
stead would spring up a system
of hundreds of branch banks
controlled by one or two leading
national groups." (New York
Times, November 7, 1923)
6. Excerpt from a small bank
president: "The passage of the
bill as it now stands,
containing the branch bank
feature, means the ultimate
elimination of State and unit
banks; it means the destruction
of individual initiate and
development, which is the thing
that every American cherishes
... The backbone of our country
is the small, independent
business and banking
institution." (Congressional
Record, 72nd Congress, 2nd
Session, Vol. 76, Pt 2, 1468)
2. Branching 1. "Although the American Great
hurts the Bankers' Association, composed Depression Era
small borrower of a majority of small
and local institutions, has taken a strong
community. opposition to nationwide branch
banking on the grounds that
small communities are best
served by local
institutions."(New York Times,
October 28, 1927, 37).
2. "The fact that the local
banker is the executive of final
responsibility, that his
greatest interest is in the
community which the bank serves,
that he is not restrained by the
lifeless rule of bank
bureaucracy, that he can,
because of his intimate
knowledge of clients, beam upon
the spirit of petty local
enterprise, all these are
marshaled in support of the
independent bank." (Bankers'
Magazine, September 1930, 301)
3. A small community banker: Modern Era
"You'd have some guy in Atlanta
deciding on whether to make a
$50,000 mortgage in Mount
Pleasant. He may decide that at
that moment he'd rather put the
money in Mexico for a better
return." (Cocheo, April 1991).
4. "Interstate branching is bad
public policy; it provides no
benefit for community bankers;
and it is little more that
special interest legislation to
allow our nation's biggest banks
to consolidate their empires,"
according to an Independent
Bankers Association of America
statement. Consolidation, they
claim, would cost communities
jobs." (Dahl et al., 1995).
Table 2 Evidence of the Large Banker Position
Argument Evidence Bank Era
1. Restrictions 1. Thomas Preston, the President Great Depression
limit balance of the Hamilton National Bank of Era
sheet diversity. Chattanooga, Tennessee, argued
that branch banking "gives a
diversification of loans and
investments that is not possible
in a unit bank, and in many
instances a large bank with
branches has been known to lend
more in some small communities
than the entire resources of
that particular branch." (New
York Times, August 1932, 15)
2. Restrictions 1. William A. Nash, President of National Banking
limit banks' the Corn Exchange National Bank, Era
ability to meet spoke before the New York
client needs. Bankers' Association and he
claimed that the greatest
benefit of a branch system is
the ability to move funds from
one branch to another to meet
demand for deposits and loans
(New York Times, October 10,
1902, 16).
2. A President of a large bank Great Depression
spoke to Congress: "Banking, Era
(too), has felt the pressure
towards larger units and
interconnections of units, the
better to serve growing
industries and communities. The
quickened tempo of today's
business has emphasized the
interdependence of communities
within the same natural
industrial area, and had
indicated the need for a more
comprehensive and more closely
knit banking service that has
been available heretofore
through the isolated unit
banks." (Congress, House,
Committee on Banking and
Currency, Branch, Chain, and
Group Banking, 71st Cong., 2nd
Sess., April 1930, 1039)
3. Restrictions 1. A banker representing the National Banking
limit large bank interest commented: Era
competition "In respect to branch banking
resulting in our system is unique. The laws
higher prices of every other important nation
for consumers. encourages branch banking, and
the results of it have never
tended to enslave the people, to
build up dangerous monopolies,
nor to increase the interest
rate. The result, in fact, has
been quite the reverse. Rates
are kept uniform over a large
territory, the tendency toward
violent fluctuations is reduced,
and the privileges and benefits
of safe banking widely
disseminated." (New York Times,
December 14, 1907, 6)
2. A large bank representative Great Depression
observed: "Wherever a branch of Era
our bank has been located where
there was an existing local
bank, it was permitted only
after careful investigation by
the banking department as to the
need of another bank in that
community. Whenever such a
condition existed you will find
in each case the then existing
bank has today increased
deposits equal to the total of
our branch operating in that
locality. Otherwise I contend a
branch of our bank in that
locality made a banking centre
and not only stimulated business
for itself but benefited its
competitor." (New York Times,
October 4, 1922, 7).
3. The CEO of Norwest Corp. Modern Era
observed: "It's (branch banking)
about allowing banks to serve
our customers wherever they are,
wherever they want to be, and
doing it faster, better and at a
lower cost. No matter where they
live, work, move or travel,
consumers will be able to bank
there, too." (Dahl et al. 1995).
4. Restrictions 1. The President of the First National Banking
hurt National Bank of Chicago, which Era
profitability was the second largest bank in
and bank the United States at the time,
stability. spoke before the Milwaukee
Bankers' Club and indicated that
a branching system would bring
more profits because deposits
could be used more efficiently
(New York Times, May 5, 1902,
1).
2. A scholar observed: "The Great Depression
branch bankers, on the other Era
hand, endeavored to prove that
intercommunity branch banking
had resulted in greater service
to the people of the state and
had contributed to the
remarkable record of safety of
banks in California during the
past fourteen years. They
favored permitting branches of
national banks not only in the
cities but on a basis of full
equality with the state banks."
(Preston 1924, 456)
3. A Spokesperson for Northwest
Bancoporation, the largest bank
in the U.S. in 1930:
"Communities will be insured
greater financial stability and
will be provided with broader
and better services than has
been possible by independent
unit banks. Regardless of
temporary depressions affecting
one locality or industry, every
bank will be able to render
uniform and continuous service."
(Neville 1930, 224).
4. Chairman and CEO of Banc One Modern Era
Corporation commented "[The
interstate banking] legislation
will lessen regional economic
downturns, such as the one that
hit New England several years
ago. It is clear that in New
England the downturn was made
much worse because weakened
banks were forced to shrink
their loan portfolios as their
capital levels fell off because
of losses. Interstate banking,
it is now recognized, would have
enables banks to better
withstand regional loan losses
and to continue providing credit
to job-creating businesses in
New England." (Wall Street
Journal, March 9, 1994, A12)
Table 3: Estimation Results for COMPETITION in National Banking Era
Variable Name Number of Banks
Operating Branches
Regression 1 Regression 2
UNIT (-1)
BRANCH (-1) 0.074 *** 0.154 ***
(1.76) (1.81)
NUMBER (-1) -0.004 ** -0.001 **
(2.06) (1.96)
YEAR 0.111 ** 0.202 ***
(2.08) (1.87)
CRISIS -0.578 *** -0.869 *
(1.68) (2.81)
%BUSFAILURE (-1) -0.416
(1.26)
Observations
Cross-section 25 25
Time-series 225 225
[R.sup.2] .576 .647
F-statistic 6.48 6.719
DW statistic 2.067 2.035
Variable Name Ratio of Unit
Banks to Total Banks
Regression 1 Regression 2
UNIT (-1) -4.11 *** -5.93 **
(1.72) (1.97)
BRANCH (-1)
NUMBER (-1) -0.001 ** -0.003 **
(1.94) (2.02)
YEAR 0.128 * 0.126 *
(2.73) (2.62)
CRISIS -0.529 ** -0.64
(2.08) (1.57)
%BUSFAILURE (-1) -0.475 *
(2.51)
Observations
Cross-section 25 25
Time-series 225 225
[R.sup.2] .614 .643
F-statistic 6.219 6.289
DW statistic 2.081 2.085
Dependent Variable: number of mergers. Independent variables:
"BRANCH" is the lag of the number of banks operating branches in
each state for each year. "NUMBER" is the lag of the total number of
banks in each state. "YEAR" controls for time-specific effects.
"CRISIS" is 0 for all years without a financial crisis, 1 for 1907.
"%BUSFAILURE" is the lag of the percent of commercial business
failures in each state. "UNIT" is the lag of the ratio of total
number of unit banks to total banks in each state. The first two
columns of results are for the legal branch status control and the
third and fourth columns are for the extent of branching control
specification. Regression 1 contains a control for the regulatory
environment and market structure, Regression 2 contains controls
from (1) and adds a control for the state economy. t-value in
parenthesis. * significant at 1% level. ** significant at 5% level.
*** significant at 10% level. Note that data availability was
limited for this era. Consequently, only 25 states are included in
this analysis: Connecticut, Florida, Georgia, Idaho, Illinois,
Indiana, Kansas, Louisiana, Michigan, Minnesota, Missouri, Nebraska,
North Carolina, Ohio, Oregon, Pennsylvania, Rhode Island, South
Carolina, South Dakota, Texas, Utah, West Virginia, Wisconsin,
Wyoming. For data sources, see Appendix A.
Table 4: Estimation Results for COMPETITION in Great Depression Era
Variable Name Number of Banks Operating Branches
Regression 1 Regression 2
UNIT (-1)
BRANCH (-1) 0.097 *** 0.101 ***
(1.69) (1.76)
NUMBER (-1) -0.001 ** -0.001 **
(1.84) (2.01)
YEAR -0.208 ** -0.216 **
(1.89) (1.97)
CRISIS 0.714 0.688
(1.30) (1.25)
%BUSFAILURE (-1) -0.829
(1.07)
Observations
Cross-section 47 47
Time-series 423 423
[R.sup.2] .664 .701
F-statistic 10.017 9.851
DW Statistic 2.107 2.107
Variable Name Ratio of Unit Banks to Total Banks
Regression 1 Regression 2
UNIT (-1) -1.732 *** -1.12 ***
(1.86) (1.81)
BRANCH (-1)
NUMBER (-1) -0.003 *** -0.002 ***
(1.81) (1.84)
YEAR -0.206 *** -0.215 ***
(1.76) (1.83)
CRISIS 0.67 0.652
(1.24) (1.19)
%BUSFAILURE (-1) -0.741
(0.958)
Observations
Cross-section 47 47
Time-series 423 423
[R.sup.2] .686 .698
F-statistic 9.893 9.926
DW Statistic 2.102 2.102
Dependent Variable: number of mergers. Independent variables:
"BRANCH" is the lag of the number of banks operating branches.
"NUMBER" is the lag of the total number of banks in each state.
"YEAR" controls for time-specific effects. "CRISIS" is 0 for all
years without a financial crisis, 1 for 1933. "%BUSFAILURE" is the
lag of the percent of commercial business failures in each state.
"UNIT" is the lag of the ratio of total number of unit banks to
total banks in each state. The first two columns of results are for
the legal branch status control and the third and fourth columns are
for the extent of branching control specification. Regression 1
contains a control for the regulatory environment and market
structure, Regression 2 contains controls from (1) and adds a
control for the state economy. t-value in parenthesis. * significant
at 1% level. ** significant at 5% level. *** significant at 10%
level. For data sources, see Appendix A.
Table 5: Estimation Results for COMPETITION in Modern Era
Variable Name Number of Banks Operating Branches
Regression 1 Regression 2
UNIT (-1)
BRANCH (-1) 2.451 ** 2.948 **
(2.05) (2.11)
NUMBER (-1) -0.0002 *** -0.0001 **
(1.86) (2.16)
YEAR 0.385 *** 0.825 **
(1.98) (2.06)
GSP (-1) -0.0002 *
(2.94)
Observations
Cross-section 50 50
Time-series 450 450
[R.sup.2] .604 .624
F-statistic 10.428 12.304
DW Statistic 2.045 2.921
Variable Name Ratio of Unit Banks to Total Banks
Regression 1 Regression 2
UNIT (-1) -1.144 -3.88
(1.21) (1.41)
BRANCH (-1)
NUMBER (-1) -0.0001 ** -0.001 **
(2.04)) (2.15)
YEAR 0.199 0.792 **
(0.94) (2.03)
GSP (-1) -0.0001 *
(3.00)
Observations
Cross-section 50 50
Time-series 450 450
[R.sup.2] .593 .602
F-statistic 10.894 11.081
DW Statistic 2.059 2.082
Dependent Variable: number of mergers. Independent variables:
"BRANCH" is the lag of the ratio of the number of banks operating
branches to total banks in each state. "NUMBER" is the lag of the
total number of banks in each state. "YEAR" controls for time-
specific effects. "GSP" is the lag of the gross state product in
each state. "UNIT" is the lag of the ratio of total number of unit
banks to total banks in each state. The first two columns of results
are for the legal branch status control and the third and fourth
columns are for the extent of branching control specification.
Regression 1 contains a control for the regulatory environment and
market structure, Regression 2 contains controls from (1) and adds a
control for the state economy. t-value in parenthesis. * significant
at 1% level. ** significant at 5% level. *** significant at 10%
level. For data sources, see Appendix A.
Table 6 Estimation Results for COMPETITION in Modern Era with
Alternative Branch Specification
Variable Name Legal Branch Status
Regression 1 Regression 2
INTRASTATE 0.684 0.369
(0.931) (0.82)
INTERSTATE 3.932 ** 2.728 **
(1.64) (1.69)
NUMBER (-1) -0.004 ** -0.003 *
(1.91) (2.73)
YEAR 0.065 0.684 ***
(0.723) (1.62)
GSP(-1) -0.0001 *
(2.72)
Observations
Cross-section 50 50
Time-series 450 450
[R.sup.2] .596 .603
F-statistic 10.792 10.901
DW Statistic 2.059 2.081
Dependent Variable: number of mergers. Independent variables:
"INTRASTATE" is 0 for years of no intrastate branching, 1 otherwise
for each state. "INTERSTATE" is 0 for years of no interstate
branching, 1 otherwise for each state. "NUMBER" is the lag of the
total number of banks in each state. "YEAR" controls for time-
specific effects. "GSP" is the lag of the gross state product in
each state. "UNIT" is the lag of the ratio of total number of unit
banks to total banks in each state. The first two columns of results
are for the legal branch status control and the third and fourth
columns are for the extent of branching control specification.
Regression 1 contains a control for the regulatory environment and
market structure, Regression 2 contains controls from (1) and adds a
control for the state economy. t-value in parenthesis. * significant
at 1% level. ** significant at 5% level. *** significant at 10%
level. For data sources, see Appendix A.
Table 7 Estimation Results for FAILURE in Great Depression Era
Variable Name Number of Banks Operating
Branches
Regression 1 Regression 2
UNIT (-1)
BRANCH (-1) -0.278 -0.263
(1.42) (1.39)
CAPASSET (-1) -19.93 -18.671
(0.97) (1.04)
LOANASSET (-1) 76.277 * 73.528 **
(2.39) (2.24)
DEPASSET (-1) -38.260 *** -37.046
(1.59) (1.47)
DIVERSITY (-1) -6.555 -30.851
(1.32) (1.32)
YEAR 2.255 ** 2.127 ***
(1.97) (1.76)
%BUSFAILURE (-1) 20.082 * 19.676 *
(2.63) (2.54)
CRISIS 61.765 * 62.081 *
(9.41) (9.37)
NUMBER (-1) 0.002 **
(2.21)
Observations
Cross-sectional 47 47
Time-series 423 423
[R.sup.2] .536 .593
F 6.391 6.419
DW Statistic 2.151 2.151
Variable Name Ratio of Unit Banks to Total Banks
Regression 1 Regression 2
UNIT (-1) 13.131 *** 13.633
(1.64) (1.45)
BRANCH (-1)
CAPASSET (-1) -18.512 -16.97
(1.04) (0.96)
LOANASSET (-1) 76.287 * 72.940 *
(2.41) (2.24)
DEPASSET (-1) -38.882 *** -37.431 ***
(1.61) (1.58)
DIVERSITY (-1) -11.296 -16.726
(0.95) (0.70)
YEAR 2.395 ** 2.240 **
(2.09) (1.87)
%BUSFAILURE (-1) 20.345 * 19.873 *
(2.67) (2.58)
CRISIS 61.613 * 61.998
(9.40) (9.39)
NUMBER (-1) 0.001 **
(1.93)
Observations
Cross-sectional 47 47
Time-series 423 423
[R.sup.2] .526 .532
F 6.561 6.641
DW Statistic 2.149 2.151
Dependent Variable: Number of bank failures in each state.
Independent Variables: "UNIT" is the lag of the ratio of number of
unit banks to total number of banks in each state. "BRANCH" is the
lag of the ratio of the number of banks operating branches to total
number of banks in each state. "CAPASSET" is the lag of the ratio of
total bank capital to total bank assets in each state. "LOANASSET"
is the lag of the ratio of total bank loans to total bank assets in
each state. "DEPASSET" is the lag of the ratio of total bank
deposits to total bank assets in each state. "DIVERSITY" is the lag
of the ratio of total number of bank branches to total number of
banks in each state. "YEAR" controls for time-specific effects.
"%BUSFAILURE" is the lag of the percent of commercial business
failures in each state. "CRISIS" equals one in 1933, otherwise 0.
"NUMBER" is the lag of the total number of banks in each state.
Regression 1 contains a control for the regulatory environment, bank
financial health controls, and the state economy. Regression 2
contains controls from (1) and adds a control for the market
structure. t-value in parenthesis. * significant at 1% level. **
significant at 5% level. *** significant at 10% level. For data
sources, see Appendix A.
Table 8: Estimation Results for FAILURE in Modern Bank Era
Variable Name Number of Banks Operating Branches
Regression 1 Regression 2
UNIT (-1)
branch (-1) -1.490 ** -2 931 ***
(2.29) (2.02)
CAPASSET (-1) -4.696 ** -9.495 **
(2.11) (2.23)
LOANASSET (-1) 0.594 0.966
(0.94) (1.02)
DEPASSET (-1) 0.004 0.071
(0.38) (0.59)
DIVERSITY (-1) -0.493 -0.388
(0.73) (0.94)
YEAR 1 739 *** 2.037 **
(1.98) (2.45)
GSP (-1) -0.0002 * -0.0001 *
(4.52) (4.72)
NUMBER (-1) 0.0001 ***
(1.99)
Observations
Cross-sectional 50 50
Time-series 450 450
[R.sup.2] .748 .766
F-statistic 22.384 23.847
DW statistic 1.903 1.956
Variable Name Ratio of Unit Banks to Total Banks
Regression 1 Regression 2
UNIT (-1) 6.651 *** 6.888 ***
(1.69) (1.72)
branch (-1)
CAPASSET (-1) -6.809 *** -11.561 **
(1.73) (1.96)
LOANASSET (-1) 0.296 0.257
(0.741) (0.82)
DEPASSET (-1) 0.0002 0.035
(0.24) (0.31)
DIVERSITY (-1) -0.328 -0.152
(0.81) (0.75)
YEAR 1.411 ** 1.562 **
(2.19) (2.34)
GSP (-1) -0.0003 * -0.0002 *
(5.25) (5.15)
NUMBER (-1) 0.0001 ***
(1.94)
Observations
Cross-sectional 50 50
Time-series 450 450
[R.sup.2] .779 .779
F-statistic 24.294 23.880
DW statistic 1.921 1.921
Dependent Variable: Number of bank failures in each state.
Independent Variables: "UNIT" is the lag of the ratio of number of
unit banks to total number of banks in each state. "BRANCH" is the
lag of the ratio of the number of banks operating branches to total
number of banks in each state. "CAPASSET" is the lag of the ratio of
total bank capital to total bank assets in each state. "LOANASSET"
is the lag of the ratio of total bank loans to total bank assets in
each state. "DEPASSET" is the lag of the ratio of total bank
deposits to total bank assets in each state. "DIVERSITY" is the lag
of the ratio of total number of bank branches to total number of
banks in each state. "YEAR" controls for time-specific effects.
"GSP" is the lag of the gross state product. "NUMBER" is the lag of
the total number of banks in each state. Regression 1 contains a
control for the regulatory environment, bank financial health
controls, and the state economy. Regression 2 contains controls from
(1) and adds a control for the market structure. t-value in
parenthesis. * significant at 1% level. ** significant at 5% level.
*** significant at 10% level. For data sources, see Appendix A.
Table 9: Estimation Results for FAILURE in Modern Bank Era with
Alternative Branch Specification
Variable Name Legal Branch Status
Regression 1 Regression 2
INTERSTATE -0.319 -0.325
(0.94) (1.10)
INTRASTATE 3.093 * 3.136 *
(2.59) (2.63)
CAPASSET (-1) -8.090 *** -6.856 ***
(1.59) (1.65)
LOANASSET (-1) 0.298 0.245
(0.46) (0.31)
DEPASSET (-1) 0.018 0.019
(0.51) (0.53)
DIVERSITY (-1) -0.292 -0.105
(0.93) (0.91)
YEAR 1.252 ** 1.426 **
(1.83) (2.01)
GSP (-1) -0.0003 * -0.0003 *
(5.71) (5.63)
NUMBER (-1) 0.0003 **
(2.04)
Observations
Cross-sectional 50 50
Time-series 450 450
[R.sup.2] .782 .784
F-statistic 24.310 23.922
DW statistic 1.941 1.973
Dependent Variable: Number of bank failures in each state.
Independent Variables: "UNIT" is the lag of the ratio of number of
unit banks to total number of banks in each state. "INTRASTATE" is 0
for years of no intrastate branching, 1 otherwise for each state.
"INTERSTATE" is 0 for years of no interstate branching, 1 otherwise
for each state. "CAPASSET" is the lag of the ratio of total bank
capital to total bank assets in each state. "LOANASSET" is the lag
of the ratio of total bank loans to total bank assets in each state.
"DEPASSET" is the lag of the ratio of total bank deposits to total
bank assets in each state. "DIVERSITY" is the lag of the ratio of
total number of bank branches to total number of banks in each
state. "YEAR" controls for time-specific effects. "GSP" is the lag
of the gross state product. "NUMBER" is the lag of the total number
of banks in each state. Regression 1 contains a control for the
regulatory environment, bank financial health controls, and the
state economy. Regression 2 contains controls from (1) and adds a
control for the market structure. t-value in parenthesis. *
significant at 1% level. ** significant at 5% level. *** significant
at 10% level. For data sources, see Appendix A.
Appendix A: Data Sources and Description
Great
National Bank Depression
Data Era (a) Era (b) Modern Era
Gross State NA NA Survey of
Product Current
Business,
various years
Commercial Historical Historical Not required.
failures Statistics of Statistics of
the U.S. the U.S.
Colonial Times Colonial Times
to 1970, each to 1970, each
year 1900- year 1924-
1910. 1934.
Total Number All Bank All Bank FDIC Historical
Commercial Statistics Statistics Statistics on
Banks 1896-1955, 1896-1955, each Banking, each
each year year 1924-1934 year 1984-1994
1900-1910
Total Number Calculated as Calculated as FDIC Historical
Unit Banks Total Number of Total Number of Statistics on
Banks minus Banks minus Banking, each
Number of Banks Number of Banks year 1984-1994
Operating Operating
Branches. Branches.
Number of Banks Banking and Banking and FDIC Historical
Operating Monetary Monetary Statistics on
Branches Statistics, Statistics, Banking, each
1914-1941, for 1914-1941, for year 1984-1994
years 1900 and years 1925 and
1910. Remaining 1930.
years Goldenweiser
extrapolated (1931) for
from regional 1931. Remaining
data in years
Goldenweiser extrapolated
(1931) from regional
data in
Goldenweiser
(1931)
Number of Banking and Banking and FDIC Historical
Branches Monetary Monetary Statistics on
Statistics, Statistics, Banking, each
1914-1941, for 1914-1941, for year 1984-1994
years 1900 and years 1925 and
1910. Remaining 1930.
years Goldenweiser
extrapolated (1931) for
from regional 1931. Remaining
data in years
Goldenweiser extrapolated
(1931) from regional
data in
Goldenweiser
(1931)
Number of Calomiris Federal Reserve FDIC Historical
Mergers (2000) provides Bulletin, 1937, Statistics on
the number of each year 1924- Banking, each
mergers and 1934 year 1984-1994
consolidations
for each state
for 1900-1909.
1910 data
extrapolated.
Total Bank All Bank All Bank FDIC Historical
Capital Statistics Statistics Statistics on
1896-1955, each 1896-1955, each Banking, each
year 1900-1910 year 1924-1934 year 1984-1994
Total Bank All Bank All Bank FDIC Historical
Assets Statistics Statistics Statistics on
1896-1955, each 1896-1955, each Banking, each
year 1900-1910 year 1924-1934 year 1984-1994
Total Bank All Bank All Bank FDIC Historical
Loans Statistics Statistics Statistics on
1896-1955, each 1896-1955, each Banking, each
year 1900-1910 year 1924-1934 year 1984-1994
Total Deposits All Bank All Bank FDIC Historical
Statistics Statistics Statistics on
1896-1955, each 1896-1955, each Banking, each
year 1900-1910 year 1924-1934 year 1984-1994
Number of Bank Annual Report Banking and FDIC Historical
Failures of the Monetary Statistics on
Comptroller of Statistics, Banking, each
the Currency 1914-1941, each year 1984-1994
(1910, 212-23) year 1924-1934
for national
banks and
Barnett (1911,
186-190) for
state banks
Notes: (a) The national bank era sample does not include Alaska,
Arizona, New Mexico, or Hawaii since these states were not yet
legally formed. Further, Oklahoma has been omitted since its
statehood was declared in 1907. (b) The Great Depression era does
not include Alaska or Hawaii since these states were not yet legally
formed.
