COMPARATIVE BUSINESS FAILURES OF FOREIGN ... - SSRN papers

COMPARATIVEBUSINESS FAILURESOF FOREIGN-CONTROLLED FIRMS IN THE UNITED STATES Jiatao Li* and Stephen Guisinger** Universityof Texasat Dallas Abstract. We explore the comparative business failures of foreign-owned or controlled firms and domestically owned firms. Original data are collected regarding foreign-controlled firms in the U.S. that filed for bankruptcyprotection,were involuntarily liquidated or ceased operations mainly due to poor financial performanceduring the 1978-1988period. Our results show that foreign-controlledfirms fail less often than domestically owned firms. The patterns of foreign-controlledbusiness failures and the impacts of entry modes, ownershiptypes, and national culture on the failures of foreign-controlledfirms are also examined. Interestin the performanceof foreign-controlledfirmsin the UnitedStates has increasedas the inwardflow of foreign investmenthas grown. Many fear that foreign firms will come to dominate certain industry groups. However,sinceU.S. firmshavea historyof failure-through bankruptcyor involuntaryliquidations-one can expect the U.S. subsidiariesof foreign firms to fail, as well. The issue that we explorein this studyis the comparative rateof businessfailurebetweenthese two groupsof firms. The rates of failurediffer betweenthese two and we examinesome of the reasons for this difference. THEORETICALFOUNDATIONAND HYPOTHESES

Before examiningthe relationshipbetween foreign control and performance, we should be more precise about our definitions. Foreigncontrol exists when a foreign firm controls ten percentor more of the equity of a subsidiary.Failureoccurswhen a firm entersinto Chapter11or a forced liquidation.Thereare,of course,alternativedefinitionsfor both failureand *JiataoLi is a doctoralcandidatein Organizations,Strategy,and International Managementat the Universityof Texasat Dallas. **StephenGuisingeris Professor of InternationalManagementStudies at the Universityof Texasat Dallas, wherehe also servesas Chairman,Joint Centerfor China-U.S.ManagementStudies. The authorswishto thankJeffreyArpan,EdwardFlowersandDavidRicksfor suggestingthe idea for this researchin theirpaper[1981]andthe threeanonymousreviewersfor tileirinsightfulcommentsand suggestions. Received:April 1990;Revised:July 1990;Accepted:October1990. 209

Palgrave Macmillan Journals is collaborating with JSTOR to digitize, preserve, and extend access to Journal of International Business Studies ® www.jstor.org

210 JOURNAL OF INTERNATIONALBUSINESS STUDIES, SECOND QUARTER 1991

control,but these definitions,being widelyused throughoutthe literature, facilitatecomparisonswith other research. We examinefive categoriesof factorsthat can explainboth the failureof firms and differences in failure rates between domestic and foreigncontrolledfirms. These factors lead logically to five distinct hypotheses that we test in a subsequentsection. OwnershipAdvantagesand BusinessFailures

Dunning [1977, 1988] developedthe eclectic paradigmto explain internationalproduction.Subsidiariesof multinationalfirms can manufacture successfullyin foreignmarketsonly if they possess ownershipadvantages. Theseownershipadvantagesmustbe sufficientto compensatefor the costs of settingup and operatinga foreignvalue-addingoperationbeyondthose faced by indigenousproducers.These ownershipadvantagesmay include intangibleassetsandadvantagesof commongovernance.Empirproprietary ical investigationshave identifiedownershipadvantagesin both U.S. and foreign-basedmultinationalfirms [Lall 1980; Pugel 1981;Swendenborg 1979;Dunning 1981;Kimura1989].However,these studiesdid not answer the questionwhetherthese advantagesare sufficientto offset the costs of operatingin countriesalien to the firm, separatedby both geographyand culture. The patternof inwardforeigndirectinvestment(FDI)may differfromthat of outwardFDI. Lall and Siddharthan[1982]pointed out that, in some industries,the ownership-specificadvantagesthat enable U.S. firms to invest abroad successfullymay be so extensiveor relativelyso dominant that foreign firms generally cannot compete successfully in the U.S. market.They suggest this may be the case in the computerindustry.In other industries,foreign firms may have built up proprietaryintangible assets that enableinvestmentin the U.S. on a largescale. This may be the case for watchesand clocks [Pugel 1985]. Arpan and Ricks [1974] identified several economic forces promoting increasedforeigndirectinvestmentin the United States. One factor is the increasedconfidence of large foreign MNCs in their ability to competL with U.S. firms in the U.S. market.Researchby Franko[1976]and Stopford [1980]also attributedthe strongforeigndirectinvestmentgrowthin the U.S. to the increasingsize and capabilitiesof foreign-basedcompanies. However,because so little researchhas been done on the comparative performanceof foreign-ownedor controlledfirms in the U.S., thereis no conclusiveevidencethat foreign-ownedfirms have discoveredbetterways of operatingin the U.S. [Arpan, Flowersand Ricks 1981]. Little [1982]studiedthe impacton financialhealthof foreignacquisitions of U.S. non-financialfirms. The author finds that at the time of their purchase firms bought by foreignerswere generallymarkedby belowfirmsacquired debt. Manufacturing averageprofitabilityand above-average by foreignerswere particularlyweak. As for the impact of the foreign

FAILURESOF FOREIGN FIRMS IN THE U.S.

211

acquisitions,the resultsshow an apparentdeteriorationin the U.S. firms' relativeprofitabilityand an apparentaccelerationin the growthof salesand assets. Apparently,foreignownerswereable to provideadequatecapital, technology or managementskills to help acquired firms expand their marketshares [Little 1982]. Studies of foreign divestmentby Boddewyn [1979], Wilson [1980] and Casson [1986]provideadditionalinsightsinto the performanceof foreign investments.Foreign divestmentdiffers from initial entry in two ways. First, divestmentoccurswhen a firm loses one of the ownership,location or internalizationadvantages[Dunning 1988:22]. Second, there may be certain exit barriersthat do not correspondto barriersto entry [Porter 1980;Harrigan1982]. Recentevidencehas shown that the rate of divestmentby multinationalfirmshas beenincreasing[Davidsonand McFetridge 1984]. Severalkey factors such as poor pre-investmentanalysis, adverse environmentalconditions,bad communications,organizationaland structural factorsmay lead to poor financialperformance,which in turn may contributeto divestmentdecisions [Boddewyn1979]. Twocases can be consideredwhereownershipadvantagesarerelatedto the firms.The first comparativeperformanceand failuresof foreign-controlled is when foreign firms possess specific ownershipadvantagesover their competitorsin the U.S. These foreign firms can utilize these ownership advantagesby either setting up new subsidiariesor even acquiringweak U.S. companies.This is consistentwith Little's [1982]finding that firms bought by foreignersweregenerallymarkedby below-averageprofitability and above-averagedebt. The second case is when foreign firms may not possess a particularownershipadvantageand can acquiresuch advantage by buying a U.S. firm. Based on the ownershipadvantagesargumentfor internationalproduction,we would expect foreign-controlledfirms in the U.S. to have a lowerbusiness failurerate than domesticallyowned firms. We examine the following hypothesis for all non-financialindustriesin aggregateand for major industrygroups (manufacturing). Hi: The business failurerate of foreign-controlledfirms in the U.S. is significantlylower than that of domesticallyowned firms. Liability of Newness

Aggregateanalyses[Altman1983]showthat overone-half of all failuresin the U.S. occur within the first five years of a firm's existence.Research in organizational ecology [Stinchcombe 1965; Freeman, Carroll and Hannan 1983]showsthat new organizationssuffer a liabilityof newness, a greaterrisk of failurethan older organizations,becausethey depend on the cooperationof strangers,have low levelsof legitimacy,and are unable to compete effectivelyagainst establishedorganizations.Further,Stinchcombearguedthat new organizationsof a new form are morelikelyto fail than new organizationswith an establishedform. Empiricalstudies (e.g.,

212 JOURNAL OF INTERNATIONALBUSINESS STUDIES, SECOND QUARTER1991

Freemanet al. 1983)found that thereis a liabilityof newness:death rates at early ages are much higher than those at later years. They have also found that the effectsof newnesscan be separatedfromthose of smallness. Lupo et al. [1978]found that the profit rates of foreignaffiliates of U.S. MNCsin 1966werestronglyrelatedto the age of subsidiaries,aftercontrolling for the countryand industryin which the subsidiaryis located. This age-profitabilityrelationshipis largelydue to the higherpercentageof new subsidiariesrunninglosses. A corollaryholds that the averageprofits of new foreignsubsidiariesshould be low and the failurerate shouldbe high [Caves1982].Our second hypothesisexaminesthe liabilityof newnessof foreign-controlledfirms in the United States. H2: NewU.S. affiliatesof foreignparentssuffera higherbusiness failurerate than more establishedU.S. affiliates. Entry Modes

The choice among acquisition,joint ventureor greenfieldentryis related to business failuresbecause they differ both in expectedriskinessand in the importanceof various coordinationcosts. Michalet and Delapierre [1976]suggestedthat acquisitionis morelikelyin sectorswherethe advantages of MNCsrestin generalorganizationalabilityand not technologyor other specificassets. New ventureshave the disadvantageof greaterriskiness [Caves 1982, pp. 81-82]. One might expect that the survivalrate of subsidiariesfounded by acquisitionwould exceedthat of newly founded subsidiaries,because of risk differences. However,the data from the HarvardMultinationalEnterpriseProjectshowthe opposite:foreignsubsidiariesof U.S. MNCs foundedby acquisitionhavea higherrateof exit than those founded as new ventures[Curhanet al. 1977;Wilson 1980]. Acquisitionand greenfieldcan be consideredas representingalternative entry modes, with joint venturesbeing only a question of the degreeof ownership[Cavesand Mehra1986;Kogutand Singh 1988].A joint venture is formedby the pooling of assets in a commonand separateorganization by two or more firms who share joint ownershipand control over the pooled assets. Both acquisitionsand greenfieldinvestmentscan take the form of joint ventures.In this study we treatjoint venturesas a separate entrymode becausetheypossessuniquecharacteristicsthat affect business failures.Coordinationof and conflicts betweenjoint ventureparents,as well as the potentialunintendeddisclosureof proprietaryknowledge,can generatetransactioncosts associatedprincipallywith uncertainty,opportunistic behavior and asset specificity [Williamson 1975; Anderson and Gatignon1986].Killing[1983]assertedthat, among his threejoint venture categories,dominantpartnerjoint venturesaremorelikelyto be successful. Totest thishypothesis,Killingmeasuredperformancethroughmanagement's assessmentof the joint venture'sperformance,as well as throughliquidations and reorganizations.Kogut [1988]studiedthe life cycle of domestic and internationaljoint venturesin the U.S. and found that mortalityrates


213

differ for these two groupsof joint ventures.Wewould expectthat foreign investmentsenteredthroughjoint venturesaremorelikelyto fail thanthose enteredthrough greenfields.However,the differencesin the impacts of acquisition entries and joint ventureentries on business failure are not clear.Thus our subhypothesis[H3(b)]takesa null form. The thirdhypothesis examinesthe relationshipbetweenentry modes and businessfailures. H3: Foreign-controlledfirms entered through acquisitions are more likely to fail than those enteredthroughgreenfields. (a): Foreign-controlled firmsenteredthroughjoint ventures are more likely to fail than those entered through greenfields. (b): There is no difference in the failure rates between foreign-controlledfirms enteredthrough acquisitions and those enteredthroughjoint ventures. Ownership1)Ipes

The types of foreign ownershipmay also relateto business failures. We consider three types of foreign ownership: MNC, individual, and government-owned enterprise.Firmsof differentownershiptypes differin the following factors: (1) the interests and constraints of owners and managers;and (2) the abilities of these parties to obtain resourcesfrom productmarketsand factormarkets,suchas capital,management,and technical talent [Mascarenhas1989].Government-owned firms havegenerated muchcontroversyabout "unfair-competition"in such diverseindustriesas airlines, telecommunications,mining, aircraft manufacturingand steel [Waltersand Monsen 1979].Foreignparentsof differentownershiptypes maypossessdifferentownershipadvantages.If ownershipadvantagessuch as managementand marketingskills, technology,and capitalplay crucial roles in the success of foreign-controlledfirms [Dunning 1977, 1988], we would expecta lowerfailurerate from U.S. affiliatescontrolledby foreign MNCsthan those controlledby foreignindividuals.However,the natureof comparativefailuresof firmscontrolledby foreigngovernmentsis not clear when compared to those controlled by foreign individuals. Thus our subhypothesis[H4(b)] takes a null form. H4: The U.S. affiliatescontrolledby foreignMNCs areless likely to fail than those controlledby foreign individuals. (a): The U.S. affiliatescontrolledby foreignMNCs areless likelyto failthanthosecontrolledby foreigngovernments. (b): Thereis no differencein the failureratesbetweenthe U.S. affiliates controlled by foreign individualsand those controlledby foreign governments, CulturalDistance

Contraryto the usual perception,surveysof internationalbusinessfailures [Altman 1983]revealthat, comparedto the U.S., business failureratesin


culturallydissimilarcountries(e.g., Japan)are not significantlydifferent fromthose of culturallysimilarcountries(e.g., Canadaor UK). The Japanese rate of business failurein recentyears is at least as large as that of in smallin Japanareconcentrated the UnitedStates.Althoughbankruptcies and medium-sizefirms,especiallythose that do not enjoyprotectionof an affiliatedgroupof companies(keiretsu),a numberof largefirmslisted on the first section of the TokyoStock Exchangehavealso failed (about fifty from 1963to 1978) [Altman 1983].1Japan is also an importantcase for on FDI behavior testinghypothesesabouteffectsof nationalcharacteristics and performance.The culturaldistanceof Japan from Westernindustrial countries has been identified as a negative factor in its participationin foreign directinvestment[Yoshino1976;Ozawa 1979]. Previousstudieshavesuggestedthat the abilityof foreignfirmsto manage the local operationsof subsidiariesmay be influencedby two considerations. One concernsthe relativeculturaldistancebetweenhome and host countries.The secondconcernsthe absoluteculturalattitudestowarduncertainty avoidance.Even with the internationalexpansionof organizational boundaries,managersin thesefirmsarelikelyto be influencedby the dominant countryculture[Hofstede 1980;Johanson and Vahlne1977]. Kogut and Singh [1988]providedevidencethat culturaldistance and attitudes towardsuncertaintyavoidanceinfluencethe selectionof entrymodes. The following hypothesisrelatesnational cultureto business failures. H5: U.S. affiliates whose foreign parents are from culturally dissimilarcountriesare more likely to fail than those from culturallysimilarcountries. DATA AND RESEARCH METHODOLOGY

Data and Sample Selection

The definitionof businessfailuresin this studyincludesfirmsthat filed for bankruptcyprotection,wereinvoluntarilyliquidatedor ceased operations mainlydue to poor financialperformance.Weused two sourcesto compile a list of companies that filed for bankruptcyprotection under either Chapter11 or Chapter7 of the U.S. bankruptcycode duringthe period 1978-1988:F & S Indexof CorporateChanges(F & S) and the WallStreet JournalIndex(WSJ).We collecteddata on foreignownershipinterestsof firms on the bankruptcy list through three sources: Department of Commercedocumentson FDI in the U.S.; Directoryof foreignmanufacturersin the U.S. by Arpan and Ricks [1974, 1979, 1985 editions]; and Mergers & Acquisitions (various issues).2

By comparingcompany names from the two lists, we createda list of companiesthat filed for bankruptcyduringthe 1978-1988 foreign-controlled period. For 1978-88, the F&S Index contains 1,470 non-financialfirms under bankruptcy, of which 94 firms were foreign-controlled.We recheckedthe 94 firms against relatedreportsin the WallStreetJournal


215

and announcementsin the M &A journaland foundthat only 55 had effective foreigncontrolbefore filing for bankruptcy.The other 39 firmseither came underforeigncontrolafterthe firm declaredbankruptcyor werethe divestedassetsor divisionsof the U.S. firmsthat laterdeclaredbankruptcy. Wethen examinedforeign-controlledfirms in the U.S. that had been involuntarily liquidated or ceased operations mainly due to poor financial performance.Data on liquidationswerecollectedfrom the F & S Indexof CorporateChanges,again for 1978-1988.Among the 690 firms in liquidation reportedin the F & S Index for 1978-88,we identified 35 firms that were foreigncontrolledbefore liquidation.We then examinedreasonsfor liquidation.Thirtyfirmswereinvoluntarilyliquidatedor ceasedoperation due to financialfailures.The remainingfive whichwereidentifiedas divestments of partial assets of U.S. affiliates, were excludedfrom our list of failed firms. Thus, the total numberof foreign-controlledbusinessfailures is 85: 55 firms under bankruptcyplus 30 firms under liquidation. BusinessFailureRate

Wecollectedaggregatebusinessfailureand bankruptcystatisticsin the U.S. from Dun & Bradstreet's(D&B) annual Business Failure Record. The D&B business failurerate index reportsthe numberof failuresrecorded per 10,000firmsthat D&Bcovers.The D&Bdefinitionof businessfailures includesbusinessesthat filed for bankruptcyprotectionor ceased operations with losses to creditors.This definitionis consistentwith the one we use in this study. We developeda foreign-controlledbusinessfailurerate (FFR) to compare with the U.S. domestic business failurerate (DFR). The FFR is defined as the numberof foreign-controlledbusiness failuresper 10,000 foreigncontrolledfirmsin the UnitedStates.Data on the total numberof foreigncontrolled firms in the U.S. were collected from publications of U.S. Departmentof Commerce.We use the Dun & Bradstreetbusiness failure rateas our DFR index. Owingto the small samplesize, we conductednonparametrictests (chi-square)to determineif the FFR is significantlylower than the DFR. Parent Guarantee

Since this study concernsfailuresof foreign-controlledfirms, the issue of parentguaranteesof bank loans to foreignaffiliatesbecomesimportant.If the parentguaranteesthe debt of its (wholly or majority-owned)subsidiaries, a subsidiarycannot go bankruptindependentof the parent [Caves 1982]. Stobaugh[1970]reportedthat only one of thirty-sevenU.S. MNCs would let a subsidiarydefaulton its debt (evenif it was not formallyguaranteed). A large portion of bank lending to foreign affiliates includes explicit or implicit guaranteesby the parent firm in the loan agreement. Shapiro[1986]reportsthatMNCsarereluctantto guaranteethe debtof their subsidiaries,even when a more favorableinterestrate can be negotiated.


They arguethat affiliates should be able to stand alone. Accordingto the surveyby Robbinsand Stobaugh [1973],large U.S. MNCs prefernot to guaranteetheiraffiliates'loans.Theyoften adopt a specific"no guarantee" rule,in partbecausethey find it easierto be consistentthanto try to differentiateamong affiliates. Small MNCs also preferno guarantee.Mediumsized firms, however,are often willing to give a guarantee.In the case of joint ventures,when one partneris unableor unwillingto provideits share of the guarantee,banks may requirehigher interestrates to be paid. A recentstatementfrom GeneralMotors Company [Winters1990]notes that "as a matterof policy,GeneralMotorsdoes not provideguaranteesfor borrowingby its local subsidiaries."Even though recenteconomic conditions in Brazilhavehad a severeimpacton GM's Braziliansubsidiary(GM do Brasil), the parent company has not providedloan guaranteesas a matterof policy.Unfortunately,dataarenot availableregardingthe current practiceof the parentguarantee,especiallythe practicesof foreign-based multinationalfirms.Theparentguaranteewouldaffectthe failuresof some MNCs' wholly or majority-ownedsubsidiaries.In our sample, foreigncontrolled firms are defined as firms with lOWoor more foreign control interests.Wholly or majority-ownedsubsidiariesof MNCs account for a small proportionof the sample.Indeed,our studysheds some light on the currentparentguaranteepracticeof foreign-basedMNCs: the failuresof wholly or majority-ownedsubsidiariesin our sample show that foreign multinationalparentsmay not offer debt guaranteesto their subsidiaries. Effects of Firm Size

Studieshaveshownthat the size of a firm has an importanteffect on business failure [Freeman,Carrolland Hannan 1983]. Smaller firms have a higher failure rate than large firms. In the 1980 Benchmarksurvey of foreign direct investmentin the U.S. by the Departmentof Commerce, U.S. affiliates wererequiredto file a completereportif total assets, sales, or net income were at least $1 million. Partialreportshad to be filed if total assets, sales and net income were less than $1 million. In the 1980 Benchmarksurvey,partial reports filed by smaller firms accounted for 38.7% of the total. In this study,we haveadjustedthe numberof foreigncontrolledfirmsfor the proportionof partialreports.The lowerboundaries of the size of firms in the U.S. affiliates' population and in the firms covered by Dun and Bradstreetare identical. Except for severalrecent large foreign acquisitions,foreign-controlledfirms are generallysmaller, especiallycomparedto large U.S. multinationalfirms coveredby Dun & Bradstreet.Because of their smaller sizes, one would expect foreigncontrolledfirms to fail more often than domesticallyowned firms. FINDINGSAND DISCUSSIONS

Ideally, we could test our hypotheses simultaneouslyin a multivariate model. Unfortunately,both the nature and numberof our observations


217

precludethis approach.Nevertheless,the resultsof our bivariatehypothesis tests are rewardingenough to believethat more researchleading up to a full multivariateexaminationis clearlyjustified. Twoother limitationsof our study should be noted. First, measuresof both domesticand foreigncontrolledfailureratescontain unknownerrors.In compilingour list, we reliedon publishedsourcesof businessnews.However,some foreignbankruptciesundoubtedlymanagedto avoidnotorietyand wereleft off our list. Similarerrorsno doubt occur in the U.S. business failure rate. Second, manyof the foreign-controlledfailuresin our list had foreignparentsthat wereeitherunwillingor unableto rescuetheir troubledU.S. affiliates.We are not sure what percentageof failed U.S. firms had U.S. parents that were similarlydisinclinedto bail out their failing subsidiaries.A much higher rate of parental rescue (either a higher proportion of parental controlledsubsidiariesor a higher propensityto rescue)could explain a lowerforeign-controlledfailurerate.Data are simplylackingon this issue, but this does not implythat comparisonsof failureratesbetweendomestic and foreign-controlledfirms are without interest.3 ComparingBusinessFailureRates

businessfailuresrepresentabout 3.94%of the total nonForeign-controlled financialbusiness failuresduringthe period 1978-88in the U.S. This can be comparedto the 10.78%of total U.S. assets and 8.71%of total U.S. equityheld by foreign-ownedor controlledU.S. affiliatesduringthe same period [U.S. Departmentof Commerce1988]. This suggeststhat foreigncontrolledfirms in the U.S. have a lowerbusinessfailureratethan domestically owned firms. We comparethe failurerates of foreign-controlledfirms in the U.S. with those of domesticallyowned firmsboth in aggregateand by industrytype. In the aggregateanalysis for all non-financialindustries,the chi-square tests supportthe hypothesis(HI) that the failurerateof foreign-controlled firmsis significantlylowerthan that of domesticallyowned firms (at a .05 level for 1978 and at a .01 level for 1979-1988,see Table1). The FFR has a mean of 9.4 while the DFR has a mean of 81. In manufacturing,chisquaretests showthat the FFRis significantlylowerthanthe DFR for 198088 (at a .05 level for 1980-81and at a .01 level for 1982-1988).For 1978 and 1979, the results are significant only at a .15 level. This can be explainedby the unusuallylow U.S. failurerates duringthese two years.4 The mean of the FFR is 22 while the mean of the DFR is 85. Wecan observethat the FFR's mean in manufacturingis twice as high as the FFR's mean in all non-financialindustries.This suggests there may exist some industry concentrationamong failures of foreign-controlled firms. Ownershipadvantagesof foreign parents may result in a lower failurerate in one industrysector relativeto the other. Sincethe U.S. affiliates'proportionof total U.S. assetsis higherthan their shareof total U.S. equity,foreign-controlled firmsin the U.S. appearto use

218 JOURNAL OF INTERNATIONALBUSINESS STUDIES, SECOND QUARTER1991

TABLE 1 Comparative Foreign-Controlled Business All Non-Financial

Year

Number of U.S. Affiliates1

Number of Failures

1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988

5,456 5,839 7,079 7,490 7,901 8,147 8,359 8,491 8,815 9,990 11,190

3 4 2 7 11 9 9 11 11 9 9

Mean Total

Failure Rates 1978-1988

Industries

Manufacturing

FFR DFR2 (Per 10,000) 6** 7*** 3*** 9*** 14 11*** 11*** 13*** 12*** 9*** 8***

24 28 42 61 88 110 107 115 120 102 98

9.4

81

Number of U.S. Affiliates1

Number of Failures

1,700 1,769 2,034 2,171 2,308 2,424 2,507 2,561 2,719 3,346 3,730

3 3 2 5 7 7 4 10 5 7 7

85

FFR DFR2 (Per 10,000) 18* 17* 10** 23** 30*** 29*** 16*** 39* ** 18*** 21*** 19*** 22

44 43 53 67 97 84 124 119 114 95 98 85

60

Chi-square test: ***p