CHANGING BANKING MARKET STRUCTURE AMID
BANKING RESTRUCTURING: A CASE OF ST. LOUIS
 
Bin Zhou
Department of Geography
Southern Illinois University at Edwardsville
Edwardsville, IL 62026
 
1. INTRODUCTION

The period between 1980 and 1997 has seen significant changes in the American banking industry in the form of internal corporate reorganization, external corporate consolidation, and banking deregulation. (1, 4, 7, 8, 11, 12) The Depository Deregulation and Monetary Control Act of 1980 and the Gain-St. Germain Act of 1982 have dismantled many price and product restrictions on banking. (3, 7, 8) Banking laws in individual states and the Riegle-Neal Interstate Banking and Branching Efficiency Act of 1994 have significantly reduced geographical barriers to branch banking and interstate bank holding. (2, 12) Today banking firms have diversified their services to include the NOW accounts, real estate, insurance, securities, money market mutual funds, etc. in their portfolios. Two thirds of the banks have branched out, compared with the less than half in 1980. Many bank holding companies have evolved from a loosely affiliated bank network to a closely controlled and integrated banking organization. In addition, the nation has seen an unprecedented bank consolidation since the early 1980s. For example, between 1980 and 1994 an average of 423 bank mergers occurred each year, compared with only an average of 190 each year from 1960 to 1982. The number of banks in the nation had decreased to approximately 9,500 in 1996 from more than 14,400 in 1980. While the number of independent banks and thrifts decreased to 4176 in 1996 from 9180 in 1984, the number of multi-bank holding companies increased to 818 in 1996 from 730 in 1984. Between 1990 and 1996, a wave of mega bank mergers swept the nation. Gigantic bank mergers frequently caught the newspaper headlines, notably the mergers of NationsBank and Boatmen's in 1996, Wells Fargo and First Interstate, Chase Manhattan and Chemical, NBD Bancorp and First Chicago, and First Union and First Fidelity in 1995, NCNB and C&S/Sovran, BankAmerica and Security Pacific in 1991, etc.

There have been numerous studies on the national, regional, or state banking structural changes as a result of banking restructuring. (5, 9, 10, 13) Study at the individual city or metropolitan area level has been meager even though such study may provide insights to place specific structural changes and to structural changes at the local community level. This paper fills the void by focusing on banking structural changes in the St. Louis bank market. The investigation focuses on three main areas: out-of-market changes, in-market changes, and changes in two particular communities: East St. Louis and the Edwardsville-Glen Carbon area. It is hoped that findings in these areas shed light on issues concerning changes in banking performance as a result of changes in banking structure, and their different impacts on various local communities.
 

2. BANKING CONSOLIDATION IN THE ST. LOUIS BANK MARKET

The St. Louis bank market is a geographic region defined by the Federal Reserve St. Louis Bank following the Horizontal Merger Guideline (1992) by the Antitrust Division at the Department of Justice. According to the Merger Guideline, a geographic market is a region where a product is competitively priced. Any significant price increase at one location will result in a customer exodus into other locations, and therefore result in the price increase in other locations within the market. In the case of banking, this means that commercial banking as a cluster of services (taking deposits and issuing commercial loans) will be provided at the same price within the geographic market, but at significantly different prices in other geographic markets. Apparently, the notion of geographic market recognizes space or distance as a crucial factor in separating banking into location-specific services. In the case that the banking system in a geographic market is moving toward a concentrated structure, the possible anti-competitive effect (e.g. higher monopolistic prices over the entire geographic market) can be assessed.

The St. Louis bank market includes several geographic units in the Missouri and Illinois bi-state area. On the Missouri side, the market includes the City of St. Louis, St. Louis, St. Charles, and Jefferson counties, plus Calvey and Boles townships in Franklin county. On the Illinois side, the market includes Madison, St. Clair, and Monroe counties plus Sugar Creek and Looking Glass townships in Clinton county. Although, as shown in Figure 1, such a geographic extent differs to various degrees from the St. Louis metropolitan area definitions in both 1990 and 1996, the St. Louis banking market contains dominant shares of banking activities in the St. Louis metropolitan area. For example, bank deposits in the St. Louis banking market accounted for 95.5% of the total deposits in the St. Louis metropolitan area in 1990, and 93.6% in 1996. Therefore, a discussion of the banking industry in the St Louis banking market is largely a discussion of the St. Louis metropolitan banking industry.

FIGURE 1
THE ST. LOUIS BANK MARKET AND METROPOLITAN AREA
(omitted)
 

The Moody's Financial Manual, the Federal Reserve Board's National Information Center of Banking Information, and The Thomson Bank Directory (formerly Rand McNally) have recorded 134 non-FDIC assistant bank consolidations between 1980 and 1996 that were associated with the St. Louis market. These consolidations involved independently owned bank holding companies (BHCs) and Federal Reserve member banks (both state and national). These various sources confirm that this data set contained all major bank consolidations in the period and thus well represented the general trend. Although these 134 consolidations include mergers and acquisitions due to their various legal connotations, no distinction is made here since the legal aspect of consolidation is not the main focus of this paper.
 

FIGURE 2
THE NUMBER OF BANK MERGERS PER YEAR IN THE ST. LOUIS MARKET

As Figure 2 indicates, the mid 1980s and mid 1990s saw the two peaks of bank mergers in the St. Louis market. The 1990s is particularly an active period for bank mergers with the number of mergers climbing yearly from the early 1990s to the peak in 1995. Of the 134 mergers, 50 or 37% are in-market mergers, meaning both acquiring and acquired firms were headquartered and mainly operating in the St. Louis market. Seventy-five or 56% of the mergers were out-of-market type in which firms headquartered in the St. Louis market acquired firms headquartered and mainly operating outside the St. Louis market. Based upon the locations of the out-of-market firms' headquarters, out-of-market mergers can occur in bi-state metropolitan areas, bi-state non-metropolitan areas, and other states. The in-market mergers were fairly evenly distributed throughout the entire period. For example, the three time spans, 1980-1985, 1986-1990, and 1991-1996, saw 15, 19, and 16 in-market mergers respectively. However, for the three time spans bi-state metropolitan mergers stood at 15, 4 and 9; bi-state non-metropolitan mergers at 5, 3, and 9; and interstate mergers at 2, 0 and 26. Apparently, the early time span saw a slightly larger number of metropolitan mergers, while the later spans saw more non-metropolitan mergers, especially interstate mergers. To be sure, 26 out of 28 interstate mergers occurred during the 1990s. In-market mergers normally result in eliminating overlapping facilities in the same market and thus reducing banking cost. Out-of-market consolidations usually lead to market expansion. Thus, in-market consolidation and cost reduction had been a constant theme for the St. Louis banking industry, while market expansion progressed geographically from metropolitan, to non-metropolitan, and to out-of-state markets.

Explanations for such an out-of-market merger pattern can be at least partly found in banking regulatory structure. Both Illinois and Missouri historically operated under an unit bank system. As a compensatory mechanism, in-state multi-bank holding was allowed (in 1975 for Missouri and before 1960 for Illinois) long before 1986 when the interstate multi-bank holding was allowed in both states. In addition, most of the adjacent states in which St. Louis banking organizations made major acquisitions later did not open their borders to interstate banking until the late 1980s and early 1990s (e.g. Arkansas in 1989, Iowa in 1991, Kansas in 1992, New Mexico in 1989, and Oklahoma and Texas in 1987). In addition, inception of nationwide interstate banking in 1995 provided the momentum for a frenzy of interstate consolidations since 1995.

Due to their various corporate strategies, banking organizations responded to regulatory and market changes differently. Two major types of consolidations can be observed among major St. Louis BHCs. The first type of BHCs had been active throughout the period in question. Examples include Boatmen's Bancshares, Magna Group, and Mark Twain Bancshares. During the 1980s, Boatmen's made several major in-market mergers raising it to the nation's 40th largest BHC in 1990 from 100th in the early 1980s. During the 1990s, Boatmen's made a series of interstate mergers. In fact 15 out of 26 interstate mergers by St. Louis banking firms during the 1990s were made by Boatmen's. Of them were the two largest expansion mergers in St. Louis banking history prior to 1997: mergers with Albuquerque based Sunwest Financial in 1992 and Wichita based Fourth Financial Corp in 1996. In 1996 Boatmen's ranked as the 22nd largest BHC in the nation.

The second type of BHCs were exemplified by First Banks, Inc, and Mercantile Bancorporation. These BHCs had been inactive during the 1980s but made active mergers during the 1990s. In 1996, Mercantile stood as the 44th largest BHC of the nation. Such a position was achieved by a series of mergers since 1991, the largest among which were the mergers with St. Joseph based Ameribanc Inc. in 1992, St. Louis based United Postal Bancorp in 1994, and Iowa based Hawkey Bancorporation in 1996.

Of the 134 bank mergers from 1980 to 1996, only 9 were made by non-St. Louis BHCs to acquire banks in the St. Louis market. Most of targets in these mergers were small in assets. In early 1997 the largest bank merger in St. Louis was completed between NationsBank and Boatmen's. Only this time the Charlotte based NationsBank became the surviving organization. Boatmen's, the first banking firm west of the Mississippi, vanished from the banking landscape. In April, Mercantile and Mark Twain undertook an in-market merger. The surviving Mercantile becomes the largest St. Louis based BHC. These latest consolidations have brought further changes to the evolving structure of the St. Louis banking industry, that are discussed below.
 

3. OUT-OF-MARKET STRUCTURAL CHANGES

Prior to 1997, St. Louis had continuously increased its visibility on the nation's banking landscape. For example, from 1980 to 1995, major St. Louis BHC shares of assets in the national totals rose from 0.5% to 1.3%. The total assets controlled by major St. Louis BHCs ranked 17th among all metropolitan areas in 1980. In 1995 the rank rose to 13. Prior to 1997, as a result of its active out-of-market mergers (especially out-of-state mergers since the 1990s), the St. Louis banking industry had expanded its corporate network in an increasingly large geographic area.

A changing out-of-market structure can be clearly seen by examining the expansion path of the four largest St. Louis BHCs prior to 1997, Boatmen's, Mercantile, Magna, and Mark Twain. In 1996, these 4 BHCs ranked 22, 44, 90, and 130 respectively in the nation in their assets value. Together, they controlled 103 banks, 1077 banking offices, $63 billion assets, and accounted for over 80% of the total deposits controlled by all St. Louis based banking firms in 1996. Table 1 lists the geographical distribution of the deposits controlled by these four BHCs from 1981 to 1996. A common trend is the decreasing deposit share in the St. Louis market and increasing share in non-metropolitan, especially out-of-bi-state markets toward the end of the period.

Prior to 1996, Mark Twain operated exclusively in metropolitan markets. In 1996 it expanded into bi-state non-metropolitan, and out-of-bi-state markets. Magna was originally an Illinois based BHC. In 1989 it moved its headquarters to St. Louis. Therefore, Magna was the only major St. Louis BHC that had majority of its holdings in Illinois. Although Magna had no holdings outside the bi-state region, its holding growth in bi-state non-metropolitan areas had been significant in the study period. During the first half of the 1990s, Boatmen's and Mercantile followed a very similar growth path: a dramatic increase in out-of-bi-state holdings and substantial decrease in the relative shares of holdings in other types of areas. In 1996,
 

TABLE 1
DISTRIBUTION OF DEPOSITS OF THE 4 LARGEST ST . LOUIS BHCS
(UNIT: $10 MILLIONS; PERCENTAGES IN PARENTHESES )
Year BHC St. Louis 

Market

Bi-state 

Metro

Bi-state 

nonmetro

Out-of-bi- 

state area

Total
1981 Mark Twain 47 (84) 9 (16) 0 (0) 0 (0) 56 (100)
Magna 11 (100) 0 (0) 0 (0)  0 (0) 11 (100)
Mercantile 191 (73) 31 (12) 40 (15) 0 (0) 262 (100)
Boatmen's 83 (62) 50 (37) 2 (1) 0 (0) 135 (100)
1985 Mark Twain 80 (86) 13 (14) 0 (0) 0 (0) 93 (100) 
Magna 63 (100) 0 (0) 0 (0) 0 (0) 63 (100)
Mercantile 243 (68) 41 (12) 72 (20) 0 (0) 356 (100)
Boatmen's 212 (63) 73 (21) 53 (16) 0 (0) 338 (100)
1990 Mark Twain 147 (83) 31 (17) 0 (0) 0 (0) 178 (100)
Magna 123 (62) 54 (27) 20 (10) 0 (0) 197 (100)
Mercantile 328 (59) 91 (17) 134 (24) 0 (0) 553 (100)
Boatmen's 547 (50) 307 (28) 185 (17) 53 (5) 1092 (100)
1996 Mark Twain 157 (71) 42 (20) 1 (0*) 19 (9) 220 (100)
Magna 133 (56) 61 (26) 42 (18) 0 (0) 236 (100)
Mercantile 484 (36) 212 (16) 245 (18) 411 (30) 1352 (100)
Boatmen's 692 (22) 417 (14) 239 (8) 1732 (56) 3080 (100)
Sources: Summary of Deposits in All FDIC-Insured Commercial and Savings Banks and U.S. Branches of Foreign Banks, Washington D.C. FDIC. relevant years.
* Less than 1%.
 
FIGURE 3
DISTRIBUTION OF ST. LOUIS BANKING NETWORKS
(omitted)
 
Mercantile had over a third of its total deposits held in states outside Missouri. For Boatmen's this ratio reached 60%. Prior to 1997, Boatmen's was the largest banking firm operating not only in Missouri, but also in Kansas, Arkansas, and New Mexico. Out-of-state expansion became a major vehicle of expansion for the top St. Louis BHCs.

Figure 3 shows deposits holding locations of St. Louis-based banking firms in 1996. Measured by the percentage of total deposits in a county controlled by St. Louis-based banking firms, St. Louis had a prominent or moderate control over wide areas in central and southwestern Missouri, a south-north strip running through central Iowa, numerous counties throughout Kansas and Arkansas, northwestern Texas, southern and northern New Mexico, and a number of counties in Oklahoma and Kentucky. St. Louis banking firms had also established their presence on the west coast and in the Southeast. St. Louis banking firms' control of deposits outside the bi-state area was more than twice as large as non-bi-state banking firms' control of deposits in the St. Louis market. Such an overwhelming one way biased bank holdings, in conjunction with several top tier BHCs of the nation, indicated the role of St. Louis as a major regional banking corporate control center prior to 1997.

The 1997 NationsBank and Boatmen's merger ended all this. Due to the transfer of corporate control from Boatmen's to NationsBank, St. Louis banking firms have lost all their market in New Mexico, Tennessee, and Oklahoma, 70-87% of the market in Arkansas, Kansas, and Texas, and 32% of the market in Iowa. Seventeen percent of deposits in Missouri and 25% in the St. Louis bank markets have been taken over by NationsBank. Banking firms outside the St. Louis market have since held twice as much deposits in the St. Louis market as the St. Louis banks have held deposits outside the St. Louis market. Currently, St. Louis banks hold $5.2 billion deposits outside the bi-state area while non-bi-state banking firms hold over $7 billion deposits in the St. Louis bank market. In addition, St. Louis lost 41% of its market shares in the bi-state non-metropolitan counties, and 56% in the bi-state metropolitan areas. The current largest St. Louis BHC, Mercantile Bancshares, formed as the result of a merger between Mercantile and Mark Twain, is less than half the size of the prior merger Boatmen's. The combined deposits of the current three largest St. Louis BHCs, Mercantile, Magna, and First Banks, Inc., are only two thirds of the prior merger Boatmen's total deposits, and only 15% of the new NationsBank's total deposits. Once rising as a regional banking corporate control center, St. Louis has fallen into a subordinate position amid the latest round of bank consolidation in the nation.
 

4. IN-MARKET STRUCTURAL CHANGES

Consistent with the national trend, the in-market structure of the St. Louis area has seen a reduction in the number of independent banking firms and an increase in the number of multi-location banking organizations. For example, from 1981 to 1995, independent banking firms decreased from 118 to 65, independently owned unit banks reduced from 38 to 10, and branch offices increased from 317 to 531. An important question concerning in-market bank consolidation is its anti-competitive effect. Various studies have shown an increased level of concentration at the national and regional levels since the banking restructuring. (9, 10) What is controversial is the change in the local market. Numerous studies by bank regulators have shown no changes in local banking structure (4, 11), while independent studies concluded that the local market also experienced an increasing banking concentration. (6)

In measuring the degree of market concentration, commonly used indicators are the n-firm concentration ratio (CRn) and the Herfindahl-Hirschman Index (HHI). The former is the combined market share by the n (n=3, 4, ...) largest firms in the entire market, while the latter is the sum of squared market shares of all firms in a market. Table 2 lists the degree of concentration for the St. Louis bank market in various measures from 1981 to 1997. Between 1981 and 1985, HHI and CR3 indeed dropped. A decreased CR3 and an increased CRn (n=4,5,10,15,20) partly explained a dropping HHI as a result of the shift of deposits toward firms smaller than the top three. However, such a deconcentrating trend did not last long. By 1990 all indicators went up, all CRn had a significant increase of 6 to 9 percentage points. Between 1990 and 1995, CR3, CR4 and CR5 had a significant increase of 5 to 7 percentage points, while CR10, CR15 and CR20 had only a slight increase of 2 to 3 percentage points. Between 1995 and 1997 the trend was similar: the first three concentration ratios had a larger increase in percentage points than the latter three did. Therefore, while banking concentration has generally persisted from the mid 1980s to this day, banking concentration in the 1990s has been characterized by the increased dominance of the largest banking firms in the St. Louis market. Although the 1997 HHI was still lower than 1800, the threshold for a market to be considered over-concentrated by the Justice Department, the continuous increase in the degree of concentration in the St. Louis market should keep bank regulators vigilant against any possible anti-competitive effects brought about by bank consolidation.
 

TABLE 2
DEGREE OF CONCENTRATION OF THE ST. LOUIS BANK MARKET
Year HHI CR3 (%) CR4 (%) CR5 (%) CR10 (%) CR15 (%)  CR20 (%)
1981 783 43.9 48.4 52.3 66.1 73.3 77.6
1985 738 40.2 50.1 54.9  70.8 76.7 80.5
1990 1109 49.7 56.6 63.4 77.7 83.4 87.0
1995 1284 54.5 63.0 69.8 80.5 86.3 89.4
1997 1555 62.3 70.8 74.4 84.6 89.0 91.8
Sources: see Table 1.
 
5. EAST ST. LOUIS VS. EDWARDSVILLE-GLEN CARBON

Although the notion of a geographic market implies banking services of certain quality accessible for all communities in the market due to a high geographical mobility within a confined space, such an accessibility is not likely to be exactly equal for customers, especially individual customers and small businesses, in all communities. This is so because it is these customers that are confined within their immediate geographical environment and thus are conditioned by what is available in that immediate environment. Banking restructuring has changed the banking structure of various communities, and thus altered the immediate banking environments for small businesses and individual customers. Banking structural changes in two communities, East St. Louis and Edwardsville-Glen Carbon (the E-G area thereafter), in the St. Louis bank market help illustrate this.
 

TABLE 3
SOCIOECONOMIC INDICATORS IN EAST ST. LOUIS
AND EDWARDSVILLE-GLEN CARBON COMMUNITIES
1990 

Popula 

-tion

Growth 

rate 

since 

1980 

(%)

1990 

% of 

Black 

1990 

Median 

house 

value ($) 

1990 

Job- 

less 

rate 

(%)

1989 

Median 

household 

income 

($)

1989 

% of family below poverty 

level

E. St. Louis 40944 -26 98.1 26,400 25 12,627 39.5
Edwardsville 14579 17 7.3  64,200 4 32,733 6.1
Glen Carbon 5197 49 5.9 89,700 3 43,287 1.7
Sources: U.S. 1990 Census

As Table 3 shows, East St. Louis is an inner city and an African American community with a host of declining socioeconomic symptoms. On the other hand, physically adjacent and the functionally integrated City of Edwardsville and Glen Carbon Village together form a white dominant and thriving suburban community with all growing signs. A common feature in banking consolidation for both communities is the penetration of major St. Louis banking firms into the local communities. In East St. Louis, St. Louis-based Landmark Bancshares bought MidAmeica Bank and Trust Company in East St. Louis in 1987, which was in turn purchased by Magna in 1991. In the case of the E-G area, Mark Twain made inroads in Edwardsville in 1987 by purchasing Edwardsville National Bank and Trust Company. In 1994, Magna bought Eagle Bank of Madison County. Since the early 1990s, Mercantile Bancshares has established an in-store branch in Edwardsville.

However, banking consolidations in the two communities have created drastically different immediate banking environments. The E-G area has attracted an increasing number of banking firms and banking facilities since the early 1980s. For example, between 1981 and 1996 while there had been 3 independent banking firms that had banking facilities in East St. Louis, the number of independent banking organization that had banking facilities in the E-G area had risen from 3 to 6. The E-G area had attracted major banking firms from Springfield, Illinois (Firstbank of Illinois Co) as well as from St. Louis. In comparison, in the later 1980s, amid banking restructuring, Union Bank moved its headquarters out of East St. Louis to Swansea, bringing with it the entire corporate supporting staff.

During the same period, the number of banking locations, including both main and branch offices, rose from 7 to 9 in the E-G area while the number dropped from 5 to 4 in East St. Louis. The most drastic change occurred in the amount of deposits in the two communities. In the E-G area, between 1981 and 1996, the total deposits measured by current dollar value had increased by over 220%, from $147 million to $476 million. In East St. Louis, however, this 15 year period saw a 28% reduction in the total deposits, from $144 million to $104 million. When the effect of inflation is taken into consideration, the gap between the two communities is more readily seen. In 1981 both communities had a similar amount of the total deposits, measured in constant dollar value (1982-1984=100): $178 million for the E-G area and $175 million for East St. Louis. The amount had since constantly increased in the E-G area to reach $303 million in 1996, a 72% increase while it experienced a 62% slid in East St. Louis, to $66 million in 1996. Clearly, there is a concurrent movement of population and deposits in these two communities. Declining East St. louis has suffered depleting human and banking resources while the thriving E-G area has received influxes of both. Therefore, it seems that at the community level, banking restructuring is manifested as uneven banking environments and that thriving communities are the final beneficiaries of restructuring.

Banking restructuring in the E-G area has another unique aspect. Though facing the penetration of major banking organizations out of the community, the E-G area was able to maintain a strong locally-controlled banking segment. In 1981 the locally owned Bank of Edwardsville controlled 50% of the total deposits in the E-G area. In 1996 its deposit share in the E-G area rose to 66%. The total market shares of deposits controlled by locally-owned banking firms reached 76% in 1996. In comparison, in 1996 East St. Louis based banking firms controlled only 38% of the total deposits in the local market, down from 52% in 1981. Thus, despite consolidation as a general trend of banking restructuring, thriving communities, with their increased economic pie, may be able to provide opportunities for not only large out-of-community banking firms, but also for locally-held firms.
 

6. CONCLUDING REMARKS

Restructuring of the St. Louis banking industry has resulted in a rise and fall of the metropolis as a regional banking corporate control center, a more concentrated banking market, and in differentiated communities. Although banking restructuring in the U. S. largely reflects a response of the banking industry as a whole to a changing business environment and new technological and financial innovations in order to remain competitive, insights should not be lost as to the impacts of the banking restructuring on the quality of banking services at places of different scales. An increasingly concentrated banking market would necessarily cause concern about the anti-competitive effect of banking consolidation. A non-concentrating local market trend found nationwide should not disguise the concentrating tendency found in individual local markets. It may be too early to tell how a switch from a corporate control point to a subordinate center would affect the banking industry and economic development in the St. Louis area. However, its impacts at the community level should be closely monitored partly due to the fact that Boatmen's was reputed for its community oriented programs. Whether these programs will continue will be largely determined by an outside banking firm NationsBank. Appropriate regulatory frameworks may be necessary to maintain the continuation of these programs so that the financial accessibility of communities, especially those that are disadvantaged, is assured. In fact, since restructuring may sharply alter the local banking environment, there may be a need to strengthen the existing community oriented programs required of banks so that the quality of life of individual communities can be improved.
 

7. REFERENCES

1. Greenspan, A. 1995. Statements to the Congress. Federal Reserve Bulletin 81(4): 349-354

2. Holland, D., D. Inscoe, R. Waldrop, and W. Kuta. 1996. Interstate Banking-The Past, Present, and Future. FDIC Banking Review 6(1):1-17.

3. Kohn, M. 1991. Money, Banking, and Financial Markets. Hinsdal, Illinois: Dryden Press.

4. LaWare, J.P. 1993. Statements to the Congress. Federal Reserve Bulletin 79(8):772-777.

Lord, J.D. 1992. Geographic Deregulation of the U.S. Banking Industry and Spatial Transfers of Corporate Control. Urban Geography 13(1):25-48.

6. Lord, J.D. 1990. Impact of Banking Acquisitions on Local Market Concentration in Florida. Southeastern Geographer 30(1):1-16.

7. Miller, R.B. 1990 American Banking in Crisis. Homewood, Illinois: Dow Jones-Irwin.

Pierce, J.L. 1991. The Future of Banking. New Haven, Massachusetts: Yale University Press.

9. Rhoades, S.A. 1996. Bank Mergers and Industrywide Structure, 1980-94. Washington D.C.: Board of Governors of the Federal Reserve System

Rose, P.S. 1987. The Changing Structure of American Banking. New York: Columbia University Press.

11. Yellen, J.L. 1995. Statements to the Congress. Federal Reserve Bulletin 81(12): 1093-1101.

Zhou, B. 1997. Geographical Patterns of Interstate Banking: A Snapshot of the U.S. Metropolitan System in the Early 1990s. Urban Geography (forthcoming).

Zhou, B. 1997. Banking Geography of the United States: Structure, Conduct, and Performance of Commercial Banks in the U.S. Metropolitan System. New York: Garland Publishing, Inc.