Financial Restructuring and Regional Economic Activity

Financial Restructuringand Regional EconomicActivity

by Brian A. Cromwell Brian A. Cromwell is aneconomist at the Federal ReserveBank of Cleveland. The authorwishes to thank Randall Eberts,Joseph Haubrich, Louis Jacob-son, Elizabeth Laderman,Katherine Samolyk, and GaryWhalen for helpful comments andsuggestions.

Introduction

The relationship between the performance ofthe financial sector and economic activity hasreceived increasing attention from economists inthe past decade. The "credit view" holds that var-iation in the supply of financial services not cap-tured in monetary aggregates can help toexplain real economic activity. Empirical studiesof the importance of the banking sector havebeen conducted at both the macro and theregional level and in general support the viewthat financial structure and stress can have realeffects.

Interest in this view coincides with the dra-matic restructuring of the financial sector thathas accompanied both the advent of deregula-tion and interstate banking and the significantincrease in bank failures in the 1980s. By theend of 1988, all but three states permitted someform of interstate acquisition of their banks,14,600 offices of banking organizations existedoutside of the organizations' home state, and

more than half of these were permitted to offerall banking services.1 Bank failures, whichaveraged 75 per year in the 1970s, rose fromless than 50 per year in the early 1980s to morethan 200 per year by 1987. Financial-sectorrestructuring in the form of bank mergers,takeovers, or failures can affect investment andconsumption decisions by disrupting the linksbetween borrowers and creditors.

This paper explores the impact of financialrestructuring on economic activity using an alter-native data set that in some respects more com-pletely measures change in the local bankingsector than do data used in previous research.Restructuring in the local banking sector ismeasured by the birth, expansion, contraction,and death of banks within a standard metropol-itan statistical area (SMSA), as estimated by theSmall Business Administration using Dun andBradstreet files.

These data, known as the U.S. Establishmentand Longitudinal Microdata (USELM), attempt to

• 1 These figures come from a recent comprehensive review of inter-state banking by King, Tschinkel, and Whitehead (1989). Earlier surveysinclude Whitehead (1983a, 1983b, and 1985), and Amel and Keane(1986).

http://clevelandfed.org/research/review/1990 Q 3

Best available copy

record the location and employment levels ofall establishments in all industries. For multi-establishment firms, ultimate ownership of eachestablishment is tracked. With respect to bank-ing, an establishment equals a bank, a bank sub-sidiary, or branch office.2 (Although the USELMestablishment framework does not account forthe variations in bank organizations across states,its advantage is that it is applied consistently.)

The USELM data are aptly suited for examin-ing the disruption of credit relationships, sinceDun and Bradstreet collects these data for thepurpose of recording the creditworthiness offirms. A "death" is recorded if a firm fails or istaken over by management sufficiently differentfrom existing management to warrant a reexam-ination of the firm's credit. To the extent thattakeovers, management changes, and branchclosings affect local credit relations, the employ-ment effects from bank deaths can potentiallymeasure the disruption of credit links betweenborrowers and lenders.

The empirical analysis presented here usesemployment changes resulting from the birth,death, expansion, and contraction of small, mid-sized, and large banks as a proxy for restructur-ing in the banking industry. These measures arelinked to the local economic performance of217 SMSAs in the periods 1980-82 and 1984-86.If the credit view is supported, the impact ofemployment changes from a bank death shouldbe negative and significantly greater in mag-nitude than the impact of a bank contraction,since a bank closing should be more disruptiveto credit relations than simply a reduction in thebank's staff.

The results suggest that the deaths of mid-sized banks—those employing between 100 and500 employees—have a negative but short-livedimpact on economic activity. Exploration of thechannels for this impact indicates that bankdeaths affect employment in other midsizedfirms that presumably rely principally on localbanking markets and that are the most likelycustomers of midsized banks. The results controlfor overall financial restructuring and laggedeconomic activity, and are robust across severalspecifications.

I. Local EconomicEffects of FinancialStress

Restructuring due to financial stress—reflectedin bank failures, closings, and mergers—is poten-tially detrimental to local economic growth.3 Inthe case of a bank failure, several types of eco-nomic agents may be affected. Bank share-holders and uninsured depositors, for example,may suffer declines in wealth. For a local econ-omy, however, any wealth effect is likely to besmall. If the failed bank is merged with anotherbank, as is commonly the case, uninsureddepositors may suffer no losses. Moreover, evenif a failed bank is closed, these depositors gen-erally recover (over time) a high percentage oftheir funds when the bank's assets are liquidated.

Another effect on local economic activitytakes place through reductions in bank employ-ment when banks fail or are taken over. In addi-tion to this direct consequence for local employ-ment, unemployed bank workers suffering aloss of personal income will also likely reducetheir consumption expenditures, sending a fur-ther negative ripple (or multiplier) effectthrough the economy. The following sectionpresents direct measures erf the employmentlosses caused by bank deaths and looks foremployment effects in nonbank sectors. Moreimportant, it examines a direct measure ofemployment losses in banking due to bank con-tractions to see if the spillovers from bankemployment losses are different for failuresthan for contractions.

Credit Disruptions

In addition to these two potential effects, bankclosings can disrupt credit relationships. Thecredit-view literature holds that the principalchannel of a bank failure's economic impact isthe disruption of borrower-lender relationships.Each lender is assumed to have more informa-tion on its existing borrowers than do otherpotential lenders.4 When bank failure results inclosure rather than reorganization, borrowersare forced to seek credit from new sources. Dur-ing the period in which a new long-term credit

• 2 In practice. Dun and Bradstreet tracks all banking establishmentslisted in telephone directories, including branch offices.

• 3 This section in partfollows Gilbertand Kochin's (1990) presen-tation.

• 4 Gertler (1988) surveys the literature on credit and aggregateeconomic activity. Articles on the theory of financial intermediationinclude Diamond (1984) and Campbell and Kracaw (1980).


Best available copy

relationship is established, borrowers are likelyto face higher costs of credit or credit rationing.

Even if the failed bank is merged into a sur-viving bank, borrowers may encounter newloan policies and new senior management ifthey apply for extensions of their credit. Again,the terms of credit are likely to be less favorablethan those offered by the previous management.

The USELM data count bank takeovers andmergers as bank deaths only if there is achange in operating management sufficient forDun and Bradstreet to reexamine the neworganization's credit rating. To the extent thattakeovers, management changes, and branchclosings affect local credit relations, USELM'smeasures of change in bank employment dueto bank deaths can be used to estimate the dis-ruption of credit links between borrowers andlenders. Because we can also measure employ-ment changes due to bank contractions, weuse the difference between the impact of bankcontractions and bank deaths to distinguishbetween the direct employment effects of abank death and the impact of credit disruptions.^

Empirical Evidence

Empirical studies of the importance of the bank-ing sector at both the macro and the regionallevel generally support the view that financialstructure and stress can affect economic activity.

In a study of the macro effects of financialstress, Bernanke (1983) argues that extensivebank runs and defaults in the 1930-33 financialcrisis reduced the efficiency of the financial sec-tor in performing its intermediation function,and that this had adverse effects on real outputthrough either than monetary channels . Heexamines the effect of the real value of thechange in deposits of failed banks on thegrowth rate of industrial production. Usingregression analysis with monthly data for theyears 1919 through 1941, Bernanke finds thatbank failures have a negative and statisticallysignificant effect on industrial output. Samolyk(1988) conducts a similar test on British data,using corporate and noncorporate insolvenciesas proxies for the health of the financial sector,and also finds that credit factors matter empiri-cally in explaining output. Using Canadian data,Haubrich (1990) also determines that the credit

• 5 We maintain the assumption that the local economic effect of abank contraction results solely from the multiplier effect of reducedemployment, rather than from credit disruptions.

disruptions resulting from bank failures, asopposed to expansions or contractions, affecteconomic activity.

The credit view would also predict an impactof stress in local banking markets on local econo-mies. Calomiris, Hubbard, and Stock (1986)examine the impact of bank failures on real farmoutput. Using annual state data for farm output,they find that the number of bank failures laggedone year has a negative and statistically signifi-cant effect. Gilbert and Kochin (1990) test thehypothesis that bank failures have adverseeffects on sales subject to sales tax and oncounty employment using rural county-leveldata. They find that bank closings have a nega-tive impact on local sales and on nonagriculturalemployment.

II. Alternative Dataon FinancialRestructuring: TheUSELM File

Financial restructuring at the local level is ana-lyzed here by the birth, expansion, contraction,and death of banking establishments at theSMSA level, as measured in the USELM data forthe periods 1980-82 and 1984-86. The longitu-dinal establishment data files of the U.S. Estab-lishment and Enterprise Microdata (USEEM)were constructed primarily from data in the Dunand Bradstreet Duns Market Identifier Files(DMI). The Small Business Administration thenassembled more than 16 million establishmentrecords contained in the DMI files to constructthe USELM file. A team from the Brookings Insti-tution merged DMI data longitudinally and asso-ciated each establishment with its owners. Thelongitudinal detail in the data makes it possibleto measure employment change of establish-ments in a given size class. It also allows employ-ment change to be decomposed by establish-ment birth or death, or by the growth (or con-traction) of continuing establishments, andallows for the tracking of mergers, acquisitions,and divestitures.

The data base includes employment figuresand industry classifications for all establishmentsand enterprises, sales data for all enterprises andsubsidiaries, age of all nonbranch establish-ments, and organizational status and geographicdata for each establishment. In principle, everyestablishment in the United States is covered,except for federal agencies.

Dun and Bradstreet's principal business isprovision of credit ratings, which must be


Best available copy

updated to reflect discontinuances of businessmanagement. Among other activities, the firmtracks court proceedings in bankruptcy cases inorder to record the creditworthiness of estab-lishments. A death is recorded if the establish-ment is closed or is taken over by managementthat is sufficiently different to warrant reexamin-ing the firm's credit.'

In the case of banking, a death could repre-sent the takeover of a bank by another bank(recorded simultaneously as a death and anexpansion), a major change in ownership andmanagement (recorded simultaneously as adeath and a birth), or a true failure (recordedsolely as a death). As such, it is an imperfectmeasure of liquidations of financial institutions.To the extent that takeovers and managementchanges affect local credit relations, the USELMmeasure of bank deaths can proxy for the dis-aiption of credit links between borrowers andlenders.

Advantages of theUSELM Data

In an exploration of the impact of bank struc-ture on regional development, Bauer and Crom-well (1989) show that the private banking sec-tor appears to be systematically related to firmbirths. This study, however, is limited by theinadequacy of data on the location of financialinstitutions. Bank data were obtained from theFederal Financial Institutions ExaminationCouncil's Reports on Condition and Income,known as call reports, for 1980. For some finan-cial measures such as total loans, however, it isnot possible to detemiine where the loans weremade, even though their dollar value is known.For example, loans made by an Ohio bank tofirms in Florida and Ohio are counted in thesame way. An additional measurement problemis that a call report for a consolidated bankingunit may include data for branches not locatedin the SMSA. In states that allow branch bank-ing, activity at the branches may be reportedsolely in the headquarter's SMSA.

• 6 In practice, a death is identified when a firm appearing in the Dunand Bradstreet file in an initial year does not appear in an end-year file.The reason could be either that the firm was actually closed or that itsidentifying number was changed as a result of new management. Simi-larly, a birth is identified by the appearance of a firm in the end-year file.The Small Business Administration uses two-year intervals to track firms.

In principle, the USELM data report the loca-tion and employment levels of banking estab-lishments with a greater degree of accuracythan the call report data, which record bankstatistics at the firm (headquarters) level formany establishments (subsidiaries or branches).For example, a bank headquartered in Cincin-nati could report data for its Columbus andDayton branches in its call report, resulting in adistortion of the measured banking activity inCincinnati. For purposes of the USELM data,however, branches in Columbus and Daytonare recorded as establishments in those SMSAs.Out-of-state ownership of establishments is alsorecorded, allowing examination of the impactof interstate banking.

Disadvantages ofthe USELM Data

The USELM data set. while having advantagesfor this analysis, is not problem free. The majorreasons to question the validity of statisticsderived from DMI files stem from three charac-teristics of the underlying data-collection effort.First, the employment figures are self-reportedby establishments, usually in telephone inter-views. Second, employment figures are notroutinely updated; updates are primarily a resultof requests for credit checks. Jacobson (1985)reports that substantial lags can occur betweenthe date the file is extracted and the last time thefirm was surveyed. Employment statistics areoften more than two years out of date, whichmay lead to infrequent reports for smaller andslower-growing establishments with less needfor credit checks, and to delays in picking upshutdowns and status changes for these firms.Third, there are delays in recognizing the crea-tion of new establishments. A business may bein existence three to five years before it isrecorded as a birth in the USELM data.

To deal with some of these shortcomings,the Small Business Administration has modifiedthe Dun and Bradstreet data in several ways inconstructing the USELM file. Establishments forwhich employment data were missing wereassigned the state-level median employment fororganizations in their standard industrial classi-fication (SIC) code. Some 8.7 percent of estab-lishments in finance, insurance, and real estate(FIRE) received estimated employment figures in

7 See Harris (1983) and Armington and Odle (1983,1984).


Best available copy

~~nrT A B L E 1

Bank Employment Share:Small, Midsized, and Large Banks

Bank Size

Type 1:0 to 100 employees


Type 3:More than 500 employees,in-state HQ

Type 4:More than 500employees,out-of-state HQ

1980 1982 1984 1986

0.247 0.247 0.256 0.229(0.277) (0.269) (0.260) (0.224)

0.237 0.232 0.188 0.190(0.282) (0.271) (0.206) (0.196)

0.480 0.486 0.511 0.519(0.351) (0.344) (0.308) (0.294)

0.035 0.035 0.046 0.062(0.127) (0.120) (0.119) (0.145)

NOTE: Numbers arc expressed as SMSA means. Standard deviations are inparentheses.SOURCE: Authors calculations based on USELM data.

1980 (Harris [19831). Furthermore, when totalfirm-level employment exceeded the aggregateestablishment-level employment data, it wasassumed that the number of branches had beenunderreported. Branches were then imputedfrom this excess employment according to theaverage branch size for the particular industry-size classification. In 1980, 17 percent of theestablishments in FIRE were imputed. Theimputed branches lack geographic informationbeyond the state, however.

Finally, records that were not updatedduring the sample period were removed fromthe data set. The remaining updated recordswere then weighted to reflect the underlyingpopulation. Weights were assigned on the basisof the level of reporting in firm categories basedon size, organization type, and industry. Report-ing problems within the banking sector ledanalysts at the Brookings Institution to separatecommercial banking from the rest of FIRE,resulting in a much more accurately weightedpopulation in both parts of this sector. The levelof reporting problems is higher outside SMSAsthan within them (Armington and Odle [1983]),which does not present a problem for the analy-sis where the SMSA is the unit of observation.

In general, geographic errors from nonreport-ing of branches appear to be more severe forlarge firms with several establishments. Errors

resulting from records that are not routinelyupdated and from inaccurate geographic distri-bution of weighting are more severe for smallfirms that have infrequent credit checks underthe Dun and Bradstreet system. The employ-ment statistics and location of independent mid-sized firms with few establishments or brancheshowever, appear to be more accurately meas-ured. The midsized firms undergo more frequentcredit checks than small businesses and are lesslikely to be widely distributed geographically.The following estimation presents plausible andsignificant results for the economic impact ofmidsized banks on midsized firms, but weakerresults for small and large banks. Whether theseresults are due to the true importance of mid-sized banks or to measurement error is uncertain,so the findings should be interpreted withcaution.

III. ModelSpecification

This paper assumes that banking employmentlosses due to bank deaths, as measured in theUSELM data, are a reasonable proxy of creditdisaiptions during restructuring in local bank-ing markets. To control for the direct effects ofthe job losses of bank employees and the crediteffects of restructuring, I use the differencebetween the impact of bank contractions andbank deaths on local economic activity to iden-tify real economic effects resulting from disrup-tion of credit channels. Data were collected for217 SMSAs for the periods 1980-82 and 1984-86. The sample was limited to those SMSAs forwhich complete information was available.

Four types of establishments are identifiedthrough an extract of the USELM data base. Type1 establishments belong to independent firmswith fewer than 100 employees—typically single-establishment small businesses. Type 2 establish-ments belong to independent firms with 100 to500 employee;*—midsized finns that may havemore than one establishment. Type 3 establish-ments belong to finns with greater than 500employees headquartered within the same state.Type 4 establishments belong to firms with greaterthan 500 employees headquartered out of state.

The percentage of bank employees in thefour types of firm categories used in this study(averaged across SMSAs) is given in table 1. The

• 8 The Small Business Administration's time-series construction ofthe files compels the use of two-year periods.


Best available copy

T A B L E

Distribution of Bank Employmentby Asset Category, 1980

Asset Class($ millions)

0 to55 to 1010 to 2525 to 5050 to 100100 to 300300 to 500500 to 1,0001,000 to 5,0005,000 and more

Numberof Banks

8741,9374,6623,5531,9721,156

19815815737

Number ofEmployees

3,11316,28280,604

125,868144,710220,04779,216

112,331280,302419,908

Employeesper Bank

3.68.4

17.335.473.4

190.4400.1711.0

1,785.411,348.9

SOURCE: Federal Deposit Insurance Corporation. Annual Report, 1980.

L E 3

Bank Employment by AssetCategory over Time

Asset Class Employees per Bank($ millions) 1980

0to25 13.425 to 100 49.0100 to 1,000 272.21,000 and more 3,609-3

1986

11.333.3

150.12,710.3

PercentChange

13.232.144.924.9

SOURCES: Federal Deposit Insurance Corporation. Annual Report, 1980.and Statistics on Banking, 1986.

impact of financial restructuring is suggested bychanges in these employment categories overtime. Employment in small banks declined froman average 24.7 percent in 1980 to 22.9 percent in1986; employment in midsized banks declinedfrom 23.7 percent to 19.0 percent. Employment inlarge in-state banks, however, increased from 48.0percent to 51-9 percent, while employment in out-of-state banks almost doubled, from 3.5 percent to6.2 percent, reflecting the growing importance ofinterstate banking.

Bank Employmentas a Proxy forFinancial Structure

Before concluding that changes in bankemployment represent changes in financialstructure, one must assess the validity of bankemployment as a proxy for bank size, theimpact of labor productivity in the banking sec-tor, and changes in the size distribution of banks.

Table 2 reports the distribution of bankemployment in 1980 across banks categorizedby asset size. In general, the standard definitionof a midsized firm as having 100 to 500 employ-ees matches up well with the standard definitionof a midsized bank as having assets between$100 million and SI billion. Average employ-ment per bank ranges from 73 employees forbanks in the $50 million to $100 million categoryto 711 employees for banks in the $500 millionto $1 billion category. Similarly, our measures ofsmall and large firms match up with standarddefinitions of small and large banks.

The data suggest large improvements inlabor productivity (measured as employees perbank, controlling for assets) for all bank sizecategories over the 1980-86 period. As shown intable 3, average employment per bank declinedover the period for banks in all asset categories.Decreases ranged from 13 and 32 percent forsmall banks in the $0 to S25 million and S25 mil-lion to S100 million categories, respectively, to45 percent for midsized banks in the $100 mil-lion to $1 billion category. Productivity gains forlarge banks with assets of greater than $ 1 billionaveraged only 25 percent. Again, our definitionof a midsized firm is consistent with the defini-tion of midsized banks, which averaged 150employees per bank in 1986.

Table 4 reports shifts in the relative impor-tance of small, midsized, and large banks overthe 1980-86 period. In general, small banksdeclined in both number and relative impor-tance, midsized banks were little changed, andlarge banks increased in importance.

Small banks with less than $25 million inassets constituted 50.8 percent of all banks in 1980but held only 5.1 percent of all assets. By 1986,they had declined to 34 percent of all banksand their share of assets stood at 2.4 percent.Declines in the share of assets also occurred inbanks in the S25 million to $50 million and $50million to $100 million categories. The numberand share of assets of midsized banks were littlechanged over the 1980-86 period. As a percentageof all banks, those in the $100 million to $300 mil-lion category increased from 7.9 percent to 13-4


Best available copy

^ ^ ^ ^ m T A B

Distribution of Banksby Asset Category

Asset Class($ millions)

1980:0to2525 to 5050 to 100100 to 300300 to 500500 to 1,0001,000 and more

Total

1986:0to2525 to 5050 to 100100 to 300300 to 500500 to 1,0001,000 and more

Total

L E •4

Number Percentof Banks of Banks

7,4733,5531,9721,156

19815S194

14,704

4,8233,6852,8991,903

33421634C

14,200

50.824.213.47.9

: 1.3! 1.1

1.3

100.0

34.026.0

' 20.413.42.41.5

i 2.4

100.0

Percentof Assets

5.16.87.39.84.15.8

61.1

100.0

2.44.56.8

10.44.35.1

66.5

100.0

SOURCES: Federal Deposit Insurance Corporation, Annual Report, 1980,and Statistics on Banking,

^ ^ ^ H TAB

1986.

L E !

Nonbank Employment Rates:Small, Midsized, and Large Firms

Firm Size

Total employment



Type 3:More than 500employees, in-state HQ

Type 4:More than 500employees,out-of-state HQ

1980 1982 1984 1986

0.385 0.385 0.390 0.411(0.064) (0.066) (0.065) (0.072)

0.133 0.138 0.130 0.136(0.027) (0.030) (0.026) (0.025)

0.056 0.053 0.053 0.053(0.017) (0.016) (0.015) (0.014)

0.080 0.077 0.077 0.080(0.045) (0.044) (0.044) (0.048)

0.119 0.112 0.110 0.114(0.050) (0.046) (0.046) (0.046)

NOTE: Numbers are expressed as SMSA means. Standard deviations are inparentheses.SOURCE: Author's calculations based on USELM data.

percent, but their share of assets rose only from9.8 percent to 10.4 percent. Banks in the $300million to $500 million category increased theirshare of assets slightly, while those in the $500million to $1 billion range showed a smalldecline in asset share. Banks with assets ofgreater than $1 billion constituted only 1.3 per-cent of all banks in 1980 but accounted for 61.1percent of assets. By 1986, the asset share oflarge banks rose to 66.5 percent.

In general, the changes in bank structure sug-gested by the employment shifts in table 1reflect transformations observed in the distribu-tion of banks by asset size. Small banks declinedin importance, while large banks gained. Thedrop of employment in the midsized banks,however, is most likely the result of strong laborproductivity gains, which exceeded those ofboth the small and large banks, rather than adecline in their importance in terms of assets.Bank employment at any particular time, how-ever, does appear to track closely with asset size.Thus, the use of bank employment losses due tobank deaths appears to be a reasonable proxy ofcredit disruptions. In addition, the definitions ofsmall, midsized, and large firms used here cor-respond with standard definitions of small, mid-sized, and large banks.

Dependent Variableand Specification

County-level employment data from the Bureauof Labor Statistics aggregated to the SMSA levelare used to measure local economic activity.Bank employment, as reported in the USELMdata, is subtracted from the aggregate employ-ment. An alternative proxy for output (personalincome) yielded qualitatively similar results tothose reported here, but in order to comparetotal employment rates with the employmentrates in firms of various size classes, it is notused. Thus, specifications are also estimatedwith employment in small, midsized, and largefirms (nonbank) as dependent variables. Theseaverage employment rates (employmentdivided by population) are reported in table 5.Total employment rose over the period, whichbegan in recession, from 38.5 percent in 1980 to41.1 percent in 1986. Employment in small andmidsized firms was essentially flat. Employmentin large firms headquartered within the statechanged little over the period, as did employ-ment in out-of-state firms.

The effects of bank deaths on local eco-nomic activity are estimated using regression


Best available copy

T A B L E

Summary Statistics:Independent Variables

WAGE

TAX

EXPAND 1

EXPAND 2

EXPAND 3

EXPAND 4

BIRTH 1

BIRTH 2

BIRTH 3

BIRTH 4

9.256(1.875)0.404

(0.038)

0.033(0.073)0.020

(0.055)

0.057(0.151)

0.009(0.101)

0.015(0.032)

0.025(0.068)

0.099(0.302)

0.024(0.173)

6

CONTRACTl

CONTRACT!

CONTRACTS

CONTRACT A

DEATH I

DEATH 2

DEATHS

DEATH 4

0.011(0.035)0.014

(0.047)

0.034(0.075)

0.005(0.034)

0.020(0.056)

0.011(0.046)

0.040(0.110)

0.001(0.004)

NOTE: Numbers are expressed as SMSA means. Standard deviations are inparentheses.SOURCE: Author's calculations based on USELM data.

analysis. The dependent variable is the employ-ment rate at the end of the period. (Usingemployment levels and including population asan independent variable yielded similar results,as did using changes in employment andchanges in employment rates.)

The following measures potentially affectingemployment growth are included as independ-ent variables:

1) IWAGE = average (log) wage ofproduction workers in the SMSA as measuredby the Census of Manufacturers.

2) TAX = effective corporate tax rate forthe state.

3) DUM86 = a dummy variable equaling1 for the 1984-86 period.

4) BIRTH 1-4 = percent change in bankingemployment due to the births of banks of types1 through 4.

5) EXPAND 1-4 = percent change in bank-ing employment due to the expansion of banksof types 1 through 4.

6) CONTRACT 1-4 = percent change inbanking employment due to the contraction ofbanks of types 1 through 4.

7) DEATH 1-4 = percent change in bank-ing employment due to the deaths of banks oftypes 1 through 4.

8) lagged employment rates.The means and standard deviations for these

variables are given in table 6. Financial restruc-turing is measured by the percent change inbank employment due to births, expansions,contractions, and deaths (BIRTHi, EXPANDi,CONTRACTi, and DEATHi; i = 1, ...4 ) for banksof types 1 through 4. Note that the credit-disrup-tion hypothesis cannot explain why an expan-sion or birth of a bank would have an effect onnonbank employment. All components ofchange in banking are included for complete-ness, however. In particular, the expansion andcontraction variables appear separately in casethe multiplier effect of changes in bank employ-ment on local economies is not symmetric. Dif-ferences in the estimated coefficients of CON-TRACT'and DEATH variables are meant tomeasure the impact of credit disruptions. Onaverage, 2.0, 1.1, 4.0, and 0.1 percent of SMSAbank employment is lost over a two-yearperiod, due to deaths of bank types 1 through4, respectively.

As suggested in the literature on firm loca-tion, wages and tax rates are included to controlfor their impact on economic growth. A dummyvariable for the 1984-86 period controls for anyfixed effect associated with this period ofeconomic expansion.

Following previous empirical studies, thespecification includes lagged values of thedependent variables as independent variables tocontrol for the possibility of a spurious rela-tionship between bank deaths and employment.Suppose the causality between bank deaths andemployment actually runs from employment tobank deaths. Banks tend to close after periods ofrelatively slow regional economic growth. If thelagged values of the dependent variables werenot included as independent variables, the coeffi-cients on the bank death variables would tend tobe negative and significant even if bank deathshad no true effect on employment. The regres-sion results thus indicate whether lagged bankdeaths explain employment after accounting foremployment in the past year.

Timing problems in the data make it impos-sible to entirely discount a spurious correlation.


Best available copy

The dependent variables are 1) overall employ-ment data for the SMSA—an average rate for theending year—and 2) employment rates forvarious-sized finns reported in the USELM data,based on end-of-calendar-year employment.The bank death data are employment changesdue to deaths that occur in the two-year periodfrom December of the beginning year toDecember of the ending year. The laggeddependent variables are employment rates atthe beginning and middle of the two-yearperiod. Bank deaths at the end of the two-yearperiod (part of our independent variable) arethus potentially caused by adverse economicconditions at the end of the period (our depend-ent variable).

To explore the simultaneity problem further,however, we examine the effect of bank deathson economic activity in the year following thetwo-year period. We also examine the impact ofeconomic conditions at the beginning of theperiod on bank deaths. Finally, we test whetherour results are driven by adverse shocks occur-ring in the oil states. The results, while not con-clusive, suggest that our measure of the impactof banks deaths on regional activity is not beingdriven by simultaneous-equation bias.

IV. EstimationResults

The model was estimated on the pooled sampleof 434 observations using ordinary leastsquares, which are presented in table 7.

Model 1 (in column 1), which uses the totalemployment rate as the dependent variable,suggests that high wages have a negativeimpact on total employment, while taxes haveno significant effect. Lagged employment rateshave a significant effect, as does the 1984-86dummy variable.

The expansion and births of large in-statebanks {EXPAND 3 and BIRTH5) have a smallbut statistically significant effect on nonbankemployment. The coefficient on EXPAND 3 is0.012 and is significantly different from zero atthe 95 percent confidence level, which indicatesthat a 10 percent increase in bank employmentfrom the expansion of large banks raises thenonbank employment rate by 0.12 percentagepoint. A 10 percent increase from the birth oflarge banks raises the employment rate by 0.07percentage point.

The contraction of bank employment(CONTRACT! through CONTRACT A) does nothave a statistically significant effect on nonbank

employment for any of the bank types. Thedeath of midsized banks (DEATH2), however,has a statistically significant negative effect onthe nonbank employment rate. The estimatedcoefficient is -0.053 with a standard error of0.020, suggesting that a 10 percent decrease inbank employment from the death of midsizedbanks reduces the nonbank employment rateby 0.53 percentage point. Given that the aver-age employment rate in the sample is 39.8 per-cent, this represents a drop in nonbank employ-ment of 1.3 percent. The estimated coefficientfor midsized bank contraction (CONTRACT!) is-0.004 and is statistically not significantly dif-ferent from zero. The difference between theestimated coefficients of DEATH2 and CON-TRACT! suggests that the effect of midsizedbank deaths on employment is almost entirelydue to credit disruptions as opposed to the mul-tiplier effect of lost bank jobs. An F-test rejectsthe hypothesis that the coefficients are identicalat the 90 percent confidence level.

The coefficients for DEATH 3 and DEATH4are negative, and the coefficient for DEATH 1 ispositive, but all are statistically insignificant.These insignificant effects for deaths of types 1,2, and 3 banks could result from the relativeimportance of midsized banks to local econo-mies, from a difference in the type of deathsthey represent (for example, small-bank deathsbeing takeovers rather than true failures), orfrom the relative accuracy of the midsized bankdata previously discussed. Results should thusbe taken as positive evidence for the impor-tance of midsized banks, rather than asevidence for the lack of importance of smalland large banks.

To investigate further the local economicimpact of bank deaths, the model was reesti-mated with the small-, midsized-, and large-firmemployment rates as dependent variables inmodels 2, 3, and 4, respectively (shown incolumns 2, 3, and 4 of table 7). In model 2, theexpansion of midsized banks has a positiveeffect on small-business employment, while thecontraction of small banks has a negative effect.Bank deaths do not have a significant effect onsmall businesses, which is contrary to the com-mon view that small firms are the first to sufferthe effects of a credit cainch. It is possible, how-ever, that these firms, many of which are rela-tively new or small "mom-and-pop" operations,rely more on informal sources of capital—suchas loans from friends and relatives, retainedearnings, and personal savings—than on fundsfrom commercial banks. This would make smallfirms less likely to be affected by bank deaths. It


Best available copy

Regression Results: Impactof Financial Restructuringon Employment Rates

Coefficient

LWAGE

TAX

EXPAND I

EXPAND 2

EXPAND 3

EXPAND 4

BIRTH 1

BIRTH 2

BIRTH 3

BIRTH 4

CONTRACT 1

CONTRACT!

CONTRACT 3

CONTRACTS

DEATH1

DEATH 2

COTotal

Employment Rate

-0.013a

(0.005)0.029

(0.024)

0.009(0.013)0.002

(0.017)0.012a

(0.006)

-0.009(0.009)-0.032(0.031)-0.001(0.014)

0.007a

(0.003)-0.002(0.005)-0.030(0.027)-0.004(0.019)0.007

(0.012)0.017

(0.028)0.022

(0.017)

-O.O53a

(0.020)

IHHHH

(2 )

Small-FirmEmployment Rate

0.001(0.001)

0.003(0.007)0.0002

(0.0038)0.010a

(0.005)0.002

(0.002)0.002

(0.003)-0.006(0.009)0.003

(0.004)0.001

(0.001)-0.002(0.002)

-0.013b

(0.008)-0.002(0.006)0.002

(0.004)-0.001(0.008)0.002

(0.005)-0.006(0.006)

1

(3)

Midsized-Firm: Employment Rate

-0.001(0.001)0.0002

(0.0060)

-0.005(0.003)-0.003(0.004)-0.0002(0.0015)0.006a

(0.002)-0.002(0.008)-0.001(0.003)0.0005

(0.0008)0.001

(0.001)

-0.009(0.007)-0.004(0.005)-0.004(0.003)-0.0003(0.0070)-0.001(0.004)-0.017a

(0.005)

(4)

Large-FirmEmployment Rate

0.005(0.004)0.038a

(0.018)0.001

(0.010)0.011

(0.013)0.004

(0.005)-0.007(0.007)-0.009(0.023)-0.004(0.010)0.004b

(0.002)

0.003(0.004)-0.016(0.020)-0.025b

(0.014)

0.003(0.009)0.011

(0.021)

0.005(0.013)0.001

(0.015)

is also possible that these results are driven bythe errors in the data for small firms discussedabove.

The impact of financial restructuring on mid-sized firms is measured in model 3. Midsizedfirms are more likely to rely on local commer-cial banks than on national credit markets (suchas with large firms) or on informal sources ofcapital (such as with small firms); see Ellie-hausen and Woken (1990). They are thus morelikely to be affected by stress in the financial sec-tor. The results support this conclusion. The

estimated coefficient for DEATH2 is -0.017 witha t-statistic of 3.40, indicating that a 10 percentdecrease in bank employment from bankdeaths reduces the employment rate of mid-sized firms by 0.17 percentage point. With anaverage employment rate of 5.3 percent, thisrepresents a 3-2 percent drop in midsized firms'employment. Again, an F-test rejects at the 90percent confidence level the hypothesis that thecoefficient for DEATH2 equals the coefficientfor CONTRACT2.


Best available copy

T A B L E 7 c o n t i n u e d

Regression Results: Impactof Financial Restructuringon Employment Rates

CD (2) (3) (4)

Coefficient

DEATH 5

DEATH A

CONSTANT

DUM 86

EMPRTB

EMPRTA

SEMPRTA

MEMPRTA

GEMPRTA

Log likelihoodR2

Mean of the

TotalEmployment Rate

-0.002(0.009)-0.008(0.254)0.036a

(0.016)0.010a

(0.002)1.624a

(0.088)-0.697"(0.092)

—————

1122.7

0.9330.398

dependent variableNumber of

observations434

Small-FirmEmployment Rate

0.003(0.002)0.020

(0.074)-0.002(0.005)-0.002a

(0.001)0.308a

(0.026)

-0.291a

(0.027)0.975a

(0.013)———

1661.8

0.9650.137

434

Midsized-FirmEmployment Rate

-0.003(0.002)-0.062(0.063)0.002

(0.004)0.002a

(0.001)0.047a

(0.022)-0.036(0.023)

—

0.885a

(0.019)—

1726.30.912

0.053

434

Large-FirmEmployment Rate

-0.005(0.006)0.026

(0.189)-0.040a

(0.012)0.007a

(0.002)0.001

(0.066)

0.031(0.069)

————

0.970a

(0.017)

1250.90.912

0.079

434

a. Significant at the 95 percent confidence level.b. Significant at the 90 percent confidence level.NOTE: Numbers are expressed as estimated coefficients. Standard errors are in parentheses.SOURCE: Author's calculations.

The impact of midsized bank deaths on mid-sized firms is small but statistically significant.This result is plausible because these firms aremost likely to rely on local banks and becausemidsized commercial banks are more likely toconcentrate on lending to such firms. This find-ing controls for lagged total employment ratesand for the lagged midsized-firm employmentrate. A spurious correlation is still possible, how-ever, if these lagged measures are inadequatecontrols due to timing problems in the data.

Model 4 measures the impact of financialrestructuring on large firms. These firms are morelikely to have access to national credit markets

and thus may be less affected by local restructur-ing. The results suggest that the expansion orbirth of large banks has a positive impact onlarge-firm employment, while the contraction ofmidsized banks has a negative impact. Thesefindings are significant at the 90 percent confi-dence level, but not at the 95 percent level,implying that changes in local bank structure donot have powerful effects on large firms.

To explore the robustness of the results andthe potential for simultaneous-equations bias,other specifications of the model were alsotested. First, the employment rate in the year fol-lowing the two-year period used in the USELM


Best available copy

file was used as a dependent variable, withlagged employment rates for the beginning,middle, and end of the period included asindependent variables. The estimated coeffi-cient on DEATH2 was -0.013 (compared withan estimated coefficent of -0.053 in model 1)with a t-statistic of 1.024. This smaller (and statis-tically insignificant) effect suggests that eitherthe effect of bank deaths dampens out quicklyover time, or that the original result was drivenby simultaneity bias. This specification, how-ever, allows bank deaths to have an effect up tothree years after they occur. Such long-terminfluences of credit disruptions are unlikely ifbanking markets are competitive.

Second, the impact of local economic condi-tions on bank deaths was explored by usingmidsized bank deaths as a dependent variableand the beginning-of-period economic condi-tions as an independent variable. The economicconditions (total employment rates and employ-ment rate of midsized firms) had no explanatoryeffect on DEATH"1. (T-statistics of the economicvariables in various specifications were neverlarger than 0.50.) This suggests that bank deathswere not statistically driven by our measures oflocal economic conditions.

Finally, robustness of the results was testedby including geographic dummy variables tocontrol for effects from economic distress in oil-producing states, which had an especially largenumber of bank failures. In particular, we testedwhether the findings were solely due to bankfailures in Oklahoma, Louisiana, and Texasbetween 1984 and 1986. Controlling for thesestates did not affect the results. The coefficienton DEATH! in model 1 remained above -0.040in the various specifications tested, and thet-statistic did not fall below 2.60.

In sum, the structure of the data in thisexperiment does not permit standard tests ofGranger causality or simultaneity bias. The alter-native specifications tested, however, suggestthat the results are not being driven by theoccurrence of bank deaths in economically dis-tressed states or by any obvious feedback ofeconomic conditions on bank deaths.

V. Conclusion

Restructuring of the financial sector in the formof bank mergers, failures, or takeovers poten-tially affects investment and consumption deci-sions by disaipting the links between borrowersand creditors. Empirical evidence at both themacro and regional levels has shown that finan-cial structure and stress can have real economic-effects.

This paper further explores the impact offinancial restructuring on local economies usinga data set that measures change in the localbanking sector by the birth, expansion, contrac-tion, and deaths of banks at the metropolitanlevel.

The empirical analysis suggests that, control-ling for overall financial restructuring andlagged economic activity, the deaths of mid-sized banks—employing between 100 and 500employees—have a negative but short-livedimpact on local economic activity. Furthermore,employment in other midsized firms appears tobe most directly affected by these deaths. Thesefirms presumably rely on local banking marketsand are the most likely customers of midsizedbanks. The results are robust across severalspecifications.

The strongest effects of bank deaths arefound for midsized banks and firms, but thisshould not be interpreted to mean that thedeaths of small and large banks have no effect.Measurement problems, which exist for all sizecategories, are least severe for midsized banksand firms, which can account for the statisticalsignificance of these results and the statisticalinsignificance of the results for the other twotypes. Nonetheless, the results suggest that mid-sized banks are an important source of fundsfor midsized finns and that a disruption of thislink through financial restructuring can havenegative short-ain local economic effects.

The effects of credit disruption appear to beshort run. In particular, the impact of bankdeaths on employment rates appears to die outafter two years. One would expect such a resultif banking markets were competitive and con-testable. In such markets, firms that found theirsource of credit disrupted by a bank deathwould quickly be able to establish credit rela-tions with another institution. Thus, the resultspresented here, while consistent with the credit-view theory that disaiption in financial marketscan have real effects, also suggest that theseeffects are short-lived, as one would expect in acompetitive environment.


Best available copy

References

Amel, Dean F. and Keane, Daniel G., "StateLaws Affecting Commercial Bank Branching,Multibank Holding Company Expansion,and Interstate Banking," Issues in BankRegulation, Autumn 1986, 10, 30-40.

Armington, Catherine and Odle, Marjorie,"Weighting the 1976-1980 and 1978-1980USEEM Files for Dynamic Analysis ofEmployment Growth," Business MicrodataProject. Washington, D.C.: The BrookingsInstitution, April 1983.

, "U.S. Establishment LongitudinalMicrodata (USELM), The Weighted Inte-grated USEEM 1976-1982 Sample," BusinessMicrodata Project. Washington, D.C.: TheBrookings Institution, June 1984.

Bauer, Paul W. and Cromwell, Brian A., "TheEffect of Bank Structure and Profits on FirmOpenings," Economic Review, FederalReserve Bank of Cleveland, 1989 Quarter 4,25, 29-39.

Bernanke, Ben S., "Nonmonetary Effects of theFinancial Crisis in the Propagation of theGreat Depression," American EconomicReview, June 1983, 73, 257-76.

Calomiris, Charles W., Hubbard, R. Glenn,and Stock, James H., "The Farm Debt Crisisand Public Policy," Brookings Papers onEconomic Activity, Washington, D.C.: TheBrookings Institution, 1986, 2, 441-79.

Campbell, Tim S. and Kracaw, William A.,"Information Production, Market Signalling,and the Theory of Financial Intermediation,"Journal of Finance, September 1980, 35,863-82.

Diamond, Douglas W., "Financial Intermedia-tion and Delegated Monitoring," Review ofEconomic Studies, July 1984, 51, 393^14.

Elliehausen, Gregory E. and Wolken, JohnD., Banking Markets and the Use of Finan-cial Services by Small and Medium-SizedBusinesses. Staff Studies 160, Washington,D.C.: Board of Governors of the FederalReserve System, 1990.

Gertler, Mark, "Financial Structure and Aggre-gate Economic Activity: An Overview," Jour-nal of Money, Credit, and Banking, August1988 (Part Two), 20, 559-88.

Gilbert, Alton R. and Kochin, Levis A., "LocalEconomic Effects of Bank Failures," Journalof Financial Services Research, 1990 (forth-coming) .

Harris, Candee S., "USEEM, U.S. Establishmentand Enterprise Microdata, Version 3," Busi-ness Microdata Project. Washington, D.C.:The Brookings Institution, April 1983-

Haubrich, Joseph G., "Nonmonetary Effects ofFinancial Crises: Lessons from the GreatDepression in Canada," Journal of MonetaryEconomics, March 1990, 25 223-52.

Jacobson, Louis, "Analysis of the Accuracy ofSBA's Small Business Data Base," WorkingPaper 1958, Center for Naval Analyses,Alexandria, Va., October 1985.

King, B. Frank, Tschinkel, Sheila L., andWhitehead, David D., "F.Y.I.: InterstateBanking Developments in the 1980s,"Economic Review, Federal Reserve Bank ofAtlanta, May/June 1989, 71, 32-51.

Samolyk, Katherine A., "In Search of theElusive Credit View: Testing for a CreditChannel in Modern Great Britain," EconomicReview, Federal Reserve Bank of Cleveland,1990 Quarter 2, 26, 16-28.

Whitehead, David D., "Interstate Banking:Taking Inventory," Economic Review, Fed-eral Reserve Bank of Atlanta, July 1983b.

, "Interstate Banking: Probabilityor Reality?" Economic Review, FederalReserve Bank of Atlanta, March 1985, 6-19.


Best available copy

Documents

Financial Restructuring and Regional Economic Activity