University of Nebraska-Lincoln College of Business Administration

How Corporate Managers View Dividend Policy

Author(s): H. Kent Baker and Gary E. Powell
Source: Quarterly Journal of Business and Economics, Vol. 38, No. 2 (Spring, 1999), pp. 17-35
Published by: University of Nebraska-Lincoln College of Business Administration
Stable URL: http://www.jstor.org/stable/40473257
Accessed: 29-10-2015 08:40 UTC

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/
info/about/policies/terms.jsp

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content
in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship.
For more information about JSTOR, please contact [email protected].

University of Nebraska-Lincoln College of Business Administration is collaborating with JSTOR to digitize, preserve and
extend access to Quarterly Journal of Business and Economics.

http://www.jstor.org

This content downloaded from 128.111.121.42 on Thu, 29 Oct 2015 08:40:35 UTC
All use subject to JSTOR Terms and Conditions
How Corporate Managers View
Dividend Policy
H. Kent Baker*
The AmericanUniversity

Gary E. Powell
Hood College

Thisstudyinvestigates theviewsofcorporatemanagersabouttherelationship
betweendividend policyand value;explanationsofdividend relevanceincluding
thebird-in-the-hand, and agencyexplanations;and
signaling,tax-preference,
howfirmsdetermine theamountofdividendstopay. Wealso examinewhether
theresponseson thesetopicsdifferamongthreeindustry groups(manufacturing,
Weobtaindatafroma mid-1997 mailsurvey
trade,andutilities).
wholesale/retail
sentto603 chief ofU.S.firmslistedon theNYSE.Basedon 198
financialofficers
usableresponses,theempiricalresults showthatmostsurvey respondents believe
thatdividendpolicyaffects firmvalue. Of thefourexplanations for dividend
relevance,therespondents generallyexpressthehighest levelofagreement with
statementsaboutsignaling.Theresultsalso showthatmanagersare concerned
aboutthecontinuity ofdividendswhensettingdividend payments. Finally,the
respondentsfromthethreeindustry groupssurveyed generally holdsimilarviews
aboutdividend policyissues.

Introduction
Dividend policy is one of the most controversialsubjects in finance. Finance
scholarshave engaged in extensivetheorizingto explainwhycompaniesshouldpay or
not pay dividends. Otherresearchershave developed and empiricallytestedvarious
models to explain dividend behavior. Some researchershave surveyed corporate
managersand institutionalinvestorsto determinetheirviews about dividends.Despite
extensivedebate and research,the actual motivationforpaying dividends remainsa
puzzle.1

* thesurveyandtheKogod
assistanceinconducting
thankNirajGuptaforresearch
The authors
at The AmericanUniversity
Collegeof BusinessAdministration forresearchsupport.
(1996) fora discussionofthedifficulty
iSee Black(1976) andBernstein the
ofunderstanding
dividendpuzzlefromtheviewpoint offirmsandinvestors.
17
300/0017/$2.00
0747-5535/99/1
University
Copyright - Lincoln
ofNebraska

This content downloaded from 128.111.121.42 on Thu, 29 Oct 2015 08:40:35 UTC
All use subject to JSTOR Terms and Conditions
18 Bakerand Powell

In thispaper,we investigatetheviews of corporatemanagersof major U.S. firms

about three topics: (1) the relationshipbetween dividend policy and value, (2)
explanations of dividend relevance including the bird-in-the-hand, signaling, tax-
preference, and agency explanations, and (3) how firms determine the amount of
dividendsto pay.2We also examinewhethertheresponseson thesetopicsdifferamong
threeindustrygroups(manufacturing, trade,and utilities).We examine
wholesale/retail
the level of agreementthat corporate managers express about 26 theoreticaland
empiricalissues about dividendpolicy.
Investigatingthese issues is importantto determineto what extent corporate
managers agree with the various messages contained in the academic literatureon
dividends.Understandingthebeliefsof managerswho are involvedin settingdividend
policymay contribute of whyfirmspay cash dividends.This study
to our understanding
also contributesto the growing body of survey research on dividend policy. For
example, the currentstudynot only expands and updates previous surveyresearchby
Baker, Farrelly,and Edelman (1985) among othersbut also asks different questions
especially relatedto explanationsof dividendpolicy.
The studyaddresses fourresearchquestions.First,do corporatemanagersbelieve
thatdividendsare relevant?Second, what explanationof dividendsdo managerstend
to favor?Third,how do firmsset theamountof dividendsthattheypay? Finally,do the
views of managersabout dividendissues differamong different industrygroups?
Some findingsprobablyare predictable,while othersare surprising.Overall, the
evidence shows that:
• Most respondentsbelieve thatdividendpolicy affectsfirmvalue;

• The respondentsgenerallyshow the highest level of agreementwith

about thesignalingexplanationof dividendrelevance;
statements
• Managers' views on settingdividendpaymentstodayare consistentwith
those reportedby managers interviewedby Lintner (1956) and his
behavioralmodel of dividendpolicy; and
• Few statistically existamongtheresponsesfromthe
differences
significant
threeindustries.

LiteratureReview
The RelationshipBetween DividendPolicyand Value
The questionof whetherdividendpolicy affectsthevalue of the firmhas puzzled
researchersand corporatemanagersformanyyears.Dividend policy is one of themost
widely researched topics in finance. Yet, researchershave differentviews about
whetherthepercentageof earningsthata firmpays out in dividendsmateriallyaffects

2See,forexample,Ang(1987) and Barclay,Smith,andWatts(1995) fora reviewofcorporate

dividendtheoriesandevidence.

This content downloaded from 128.111.121.42 on Thu, 29 Oct 2015 08:40:35 UTC
All use subject to JSTOR Terms and Conditions
QJBE,Spring1999,Vol. 38,No. 2 19

its long-termshare price. Some empirical studies appear to support Miller and
Modigliani's (1961) classic dividendirrelevanceproposition[e.g., Black and Scholes
(1974), Miller and Scholes (1978), Jose and Stevens (1989)]; othersdo not [e.g., Long
(1978), Sterkand Vandenberg(1990)]. In addition,surveyresearchby Farrelly,Baker,
and Edelman (1985) shows thatcorporatemanagers typicallybelieve thatdividend
policy affectsa firm'svalue and thatan optimal level of dividend payout exists. In
practice,most firmspay cash dividends,althoughpayingdividendsis costlyin various
ways. Thus, empiricalevidenceon whetherdividendpolicy affectsa firm'svalue offers
contradictory advice to corporatemanagers.Today, manyacademicians and corporate
managers stilldebate whetherdividendpolicy matters.

Explanations of Dividend Relevance

Given thatmanagerstypicallybelieve in dividendrelevance,the second question
is "What explanations of dividends do managers tend to favor?" Researchers have
offered four common explanations of dividend relevance: the bird-in-the-hand,
signaling,tax preference,and agency explanations.

The Bird-in-the-HandExplanation
One argumentthata relationshipexists between firmvalue and dividendpayout
is thatdividends representa sure thingrelativeto share price appreciation.Because
dividendsare supposedlyless riskythancapitalgains, firmsshould set a highdividend
payout ratio and offera high dividend yield to maximize stock price. Miller and
Modigliani (1961) disagree and call the theorythata high dividendpayout ratiowill
maximize a firm'svalue thebird-in-the-hand fallacy. Bhattacharya(1979) also argues
thatthereasoningunderlyingthebird-in-the-hand explanationfordividendrelevance
is fallacious. The riskiness of a project's cash flows determinesa firm's risk. An
increase in dividendpayout today will resultin an equivalent drop in the stock's ex-
dividendprice. Thus, increasingthedividendtodaywill not increasea firm'svalue by
reducingthe riskinessof futurecash flows.

The Signaling Explanation

Anotherpossible reason forpaying dividends is the use of dividend policy to
communicateinformationabout a firm's futureprospects to investors.3Miller and
Modigliani (1961) realize thatin the real world a change in the marketprice often
follows a change in the dividendrate.According to the informationcontentof divi-
dends or signaling explanation, cash dividends announcementsconvey valuable

3Ross(1977) first
developedthetheoretical
analysisofdividends as a signaling
device.Various
modelsdevelopedbyBhattacharya
dividend-signaling (1979, 1980),JohnandWilliams(1985),
Millerand Rock (1985), and Oferand Thakor(1987) posita positiverelationship among
dividendpolicychanges,equity,values,andsubsequent performance.

This content downloaded from 128.111.121.42 on Thu, 29 Oct 2015 08:40:35 UTC
All use subject to JSTOR Terms and Conditions
20 Bakerand Powell

informationabout management'sassessmentof a firm'sfutureprofitability thatother

means cannot fully communicate. Informationasymmetrysuggests that corporate
managers have an informationadvantage over outside investors.If managers have
informationthatinvestorsdo not have, managersmay use a change in dividends as a
way to signal thisprivateinformationand thusreduce information asymmetry. In turn,
investorsmay use dividend announcementsas informationto assess a firm's stock
price.4 On balance, much empirical evidence supportsthe view of dividends as a
signalingdevice.5
Although managers can use dividend actions to convey useful information,
dividend changes may not be perfectsignals. According to Easterbrook (1994),
dividendincreasesmaybe ambiguoussignalsunlessthemarketcan distinguish between
growing firms and disinvesting firms, i.e., those with a lack of investment
opportunities.6For example,Soter,Brigham,and Evanson (1996) notethatFPL Group,
the parent company of Florida Power & Light Company, announced a 32 percent
reductionin its quarterlydividendon May 9, 1994 forstrategicreasons,notproblems
in cash flow.The stockmarket'sinitialreactionto FPL's announcementwas negative,
withan initialdrop of about 20 percentin value. Aftercarefullyreviewingthereasons
for the reduction,analysts concluded that the action was not a signal of financial
distress. Instead, the dividend decrease was a strategic decision designed by
managementto improvethefirm'slong-termfinancialflexibility and growthprospects.
Afterthe financialcommunityadopted thisview, FPL's stockbegan to recover.

The Tax-PreferenceExplanation
Another explanation of why dividend policy mattersinvolves the tax effect.
Accordingto thetax-preference theory,investorsmay favorretentionof fundsover the
payment of dividendsbecause of tax-related of capital
reasons.7The favorabletreatment
gains over dividends may lead investorsto prefera low dividend payout a high
to
payout.This theorysuggeststhatfirmsshouldkeep dividendpaymentslow iftheywant
to maximize prices.

4Although unexpected dividendchangestypicallyresultin stockpricechanges,thisdoes not

meanthatdividendpolicyperse is relevant.The stockpricelevelmaychangefasterduetothe
dividendchange,but eventuallyearningsreportswould providethe same information as
containedinthedividendannouncement.
5Forevidenceof signalingeffectsof dividendannouncements, see, forexample,Woolridge
(1982, 1983),AsquithandMullins(1983), Benesh,Keown,andPinkerton (1984), Ghoshand
Woolridge(1988), HealyandPalepu(1988), Bajaj andVijh(1990),Christie (1994), Michaely,
Thaler,andWomack(1995), andImpson( 1997).
6Bornand Rimbey(1993) examinetherelation betweenpriorfinancing andthemarket
activity
responseto initialdividendsand findevidenceconsistentwiththeEasterbrook (1984) agency
costmodel.
7Brennan (1970) andStapleton (1972) amongothersdevelopan optimaldividendpolicybased
on thetaxdifferentialbetweencapitalgainsanddividends.

This content downloaded from 128.111.121.42 on Thu, 29 Oct 2015 08:40:35 UTC
All use subject to JSTOR Terms and Conditions
QJBE,Spring1999,Vol. 38,No. 2 2'_
Because thetax effectdiffersamong various typesof investors,investorsmay be
attractedto firms that have dividend policies appropriateto their particular tax
circumstances.Researcherscall thisnotionthetax clienteleeffect.Otherthingsbeing
equal, stocks with low payouts should attractinvestorsin high tax brackets,leaving
highpayoutstocksto investorssubjectto low or zero tax rates.The empiricalevidence
on the tax-preferenceexplanationof dividendsis inconclusive.8

The AgencyExplanation
Anotherpopular view of dividendrelevance,advanced by Jensenand Meckling
(1976), and extendedby Rozeff(1982) and Easterbrook(1984), is agency theory.This
theoryderives fromtheconflictof interestsbetweencorporatemanagers(agents) and
outside shareholders(principals). For example, managementcan consume excessive
perquisites out of undistributedcorporate earnings and invest the retained funds
suboptimally.This conflict leads to agency costs. Agency theoryposits that the
dividendmechanismprovidesan incentiveformanagersto reduce the costs relatedto
the principal/agentrelationship.One way to reduce agency costs is to increase
dividends. Paying larger dividends reduces the internal cash flow subject to
managementdiscretionand forcesthe firmto seek more externalfinancing.Raising
costly outside capital subjects the firmto the scrutinyof the capital marketfornew
fundsand reducesthepossibilityof suboptimalinvestment. This monitoringby outside
suppliersof capital also helps to ensurethatmanagers act in thebest interestof outside
shareholders.Thus, dividendpaymentsmay serveas a means of monitoringor bonding
managementperformance.
Several empiricalstudiesprovidesupportfortheagencyexplanationfordividends.
For example,Rozeff(1982) findssupportfortherole of dividendsin resolvingagency
costs in minority-manager-controlledfirms.His analysisshows a negativerelationship
betweendividendpayoutand thepercentageof insiders.Given a lower percentageof
outsiders,less need exists to pay dividends to reduce agency costs. Crutchleyand
Hansen (1989) and Moh'd, Perry,and Rimbey (1995) conclude thatmanagers make
financialpolicy tradeoffssuch as payingdividendsto controlagency costs.

SettingDividendPayments
Researchers have developed many differentmodels for explaining dividend
behavior.In Lintner's(1956) classic study,managersperceivedthatshareholderswere
entitledto a fairshare of the firm'searningsthroughdividends.Althoughmanagers
advocated a long-rangetargetpayoutratio,theybelieved thatshareholderspreferreda
steady increase of dividends. Managers sought to avoid making changes in their
dividendratesthatmighthave to be reversedwithina yearor so. Therefore,theytended

on theroleoftaxesanddividendpolicy.
oftheliterature
8Ang(1987) providesa summary

This content downloaded from 128.111.121.42 on Thu, 29 Oct 2015 08:40:35 UTC
All use subject to JSTOR Terms and Conditions
22 Bakerand Powell

to make partialadjustmentstowarda targetpayoutratioratherthandramaticchanges.

Managers smootheddividendsin the shortrunto avoid frequentchanges.
Lintner'sbehavioralmodel suggeststhatthechange in dividendsis a functionof
thetargetdividendpayoutless thelastperiod's dividendpayoutmultipliedby thespeed
of an adjustmentfactor.The targetdividendpayoutis a fractionof thecurrentperiod's
earnings. Lintnertests his propositionsand findsthatthe partial adjustmentmodel
predicteddividendpaymentsmore accuratelythan"naive" models.
Otherstudieshave confirmedthedividendpolicybeliefsof managers,as described
by Lintner.For example,Joseand Stevens(1989) findthatthemarketvalues stableand
steady growing dividends per share ratherthan stable payout ratios. Based on an
extensiveempiricalanalysis of changes in dividends,Benartzi,Michaely, and Thaler
(1997, p. 1032) conclude that "...Lintner's model of dividends remains the best
descriptionof the dividend settingprocess available." In separate surveys,Baker,
Farrelly,and Edelman (1985) and Pruittand Gitman(1991) findcontinuedsupportfor
Lintner'sresults.

SurveyDesign
Our sample consistsof U.S. corporationslistedon theNew York Stock Exchange
(NYSE) thatpaid a cash dividendin at least one yearduringthe 1994-1995 period.The
primarybusiness of thesecorporationsis manufacturing (SIC 20-39), wholesale/retail
trade (SIC 50-58), or utility(SIC 49). We examine threeindustrygroups because
research by Michel (1979) and Baker (1988) among others suggests that dividend
policies varyacross industries.9The totalsample is 603 firms:392 manufacturing, 98
wholesale/retailtrade, and 113 utilities.As Table 1 shows, utilities,on average,
representa high dividend-payoutindustry,whereas firms in manufacturingand
wholesale/retailtradehave, on average, more moderatedividendpayoutratios.
We use a mail survey to obtain informationabout the respondents' views on
various dividend policy issues and a profileof the respondentsand theirfirms.We
group these issues into five broad explanations of dividend policy: (1) dividend
irrelevance/relevance, (2) bird-in-the-hand,(3) tax preference,(4) signaling,and (5)
agency costs. We also ask respondentsabout how firmsset dividendpayments.The
surveyasks therespondentsto indicatetheirgeneralopinionabout each of 26 closed-
end statementsbased on a five-pointresponse scale. This scale is as follows: -2 =
definitelydon't agree, -1 = probablydon't agree,0 = neitheragree nor disagree,+1 =
probablyagree,and +2 definitelyagree. We pretestedtheinitialsurveyamong a small
group of financeacademicians.

9RozefF(1982)concludesthata company'sindustrydoesnothelptoexplainitsdividendpayout
ratio.This conclusionmaynotapplyto utilitiesbecausehe intentionally
excludesregulated
companiesfromhisanalysisbecausetheirregulatory statusmayinfluence
theirpolicies.

This content downloaded from 128.111.121.42 on Thu, 29 Oct 2015 08:40:35 UTC
All use subject to JSTOR Terms and Conditions
QJBE,Spring1999,Vol. 38,No. 2 23

We mailed a cover letterrequestingparticipationin this study along with a

stamped,self-addressedreturnenvelope and thesurveyinstrument to thechieffinancial
officer(CFO) of the603 firmsin mid-April 1997. The cover letterrequestedthatifthe
recipientswere not activelyinvolved in determiningtheirfirm'sdividendpolicy that
theygive the surveyto someone in theircompanywhom was involved. We mailed a
second surveyto nonrespondentsin mid-May, 1997, to increasetheresponserateand
to reduce potentialnonresponsebias. The surveycontaineda code numberto identify
therespondentsand to avoid includingduplicateresponses in the analysis.
By the end of June 1997 the survey had yielded 198 usable responses (a 32.9
percentresponse rate). The divisionof theresponsesamong thethreeindustrygroups
was: manufacturing117 of 392 (or 29.8 percent),wholesale/retailtrade28 of 98 (or
28.6 percent),and utility53 of 113 (or 46.9 percent).Some respondentsdid notanswer
everyquestion. The most commonpositionor titleof therespondentswas CFO (63.2
percent),followedby vice president(16.6 percent),treasureror assistanttreasurer(8.8
percent),directorof investorrelations(4.7 percent),and other(6.7 percent).About 95
percentof theserespondentssaid thattheywere activelyinvolvedin determining their
firm'sdividendpolicy.

Limitations
Beforepresentingtheresults,we wantto pointout several limitingaspects of this
thesample to threebroad industry
research.First,we restrict groupsof U.S. firmslisted
on theNYSE. We also concentrateon firmsthatpaid cash dividendsand exclude those
firmswith other types of dividend policies. Omittingfirmsthat chose not to pay
dividendsfromthesurveymay bias our findingsin severalinstances.Therefore,readers
should be carefulin generalizingthe findingsto otherindustrygroups and to firms
whose characteristicsdifferfromthose of thecurrentsample.
Second, we obtained the views about dividendpolicy fromone managerwithin
each firm.Because identifying all participantsinvolved in makinga firm'sdividend
is
decisions impractical, we use the CFO or otherindividualfamiliarwith the firm's
dividendpolicy as a proxyfordividendpolicy makers.
Third,we limitthelengthand scope of thesurvey.We did notwantto discourage
potentialrespondentsby makingthesurveytoo lengthyor by askingmanyopen-ended
questions.Therefore,thesurveyinstrument involvesa tradeoffbetweentheinformation
needed and the response rate. Although our studyprovides informationabout how
dividend policy makers view certainfactorsin settingdividend policy, it does not
provide information about whytherespondentshold theseviews.
Fourth,despite effortsto make questionsclear,respondentscould misinterpret or
misunderstandthesurveyquestions.We are mostconcerned with the two questionsin
the agency area (S25 and S26). Althoughthesequestions were understandableto our
pilot testgroup of academics, theymay not have the same meaning to managers. In
retrospect,determiningifmanagersgenerallyinterpreted the specificquestions asked

This content downloaded from 128.111.121.42 on Thu, 29 Oct 2015 08:40:35 UTC
All use subject to JSTOR Terms and Conditions
24 Bakerand Powell

as we intendedwould have enhanced our abilityto draw valid conclusions fromthe

responses provided. Therefore,we are carefulin drawingconclusions based on the
responses to these two questions.
Finally,nonresponsebias could affectthe findings.To reduce thispossibility,we
took several precautionarystepsto increasetheresponseratesuch as havingmultiple,
personalized mailings and assuring anonymityand confidentialityof individual
responses. Nonresponse bias is potentiallygreaterbetween the manufacturingand
wholesale/retailtradegroupsbecause of theirlower responseratescomparedwiththe
response rate fromthe utilitygroup. The effectof nonresponse on survey results
depends on thepercentagenotrespondingand theextentto whichthosenotresponding
are biased.10
We conduct t-testsfordifferencein means to determinewhetherselected char-
acteristicsof therespondingfirmsdiffersignificantly fromthoseof thenonresponding
firms.We use COMPUSTAT to computefivecharacteristics(total assets, totalsales,
marketvalue of equity,dividendpayoutratio,and dividendyield) forall firmsin the
fullsample forwhich data are available in 1994. To reducepossible scale problemsin
conductingthe t-tests,we converttotal assets, sales, and marketvalue of equity to
logarithms.As Table 1 shows, no statisticallysignificantdifferencesexist at the 0.05
!
level forany of the characteristicsby thethreeindustrygroups.1

Survey Results
Table 2 providestherespondents'opinionsabout26 closed-endstatements relating
to dividend policy. We compute descriptivestatisticsforeach of the 26 closed-end
statementsand rankeach statementby its mean score. We use chi-squareteststo test
forsignificantdifferencesin the level of agreementamong the threeindustrygroups.
To account forinadequate cell sizes in the chi-squaretests,we collapse the response
scale fromfive to threecategories. These categories are: definitelydon't agree and
probablydon't agree (-2 and -1), neitheragreenordisagree(0), and probablyagree,and
definitelyagree (+1 and +2).

The Relationship Between Dividend Policy and Value

Our firstresearchquestionwas "Do corporatemanagersbelieve thatdividendsare
relevant?"Accordingto dividendirrelevancetheoryof Miller and Modigliani (1961),
dividendpolicy does not affecteithertheprice of a firm'sstock or its cost of capital.

lOFowler(1984) statesthatonewayto studynonresponsebias is tocomparethosewhorespond

immediatelywiththosewhorespondaftertakingfollow-up steps.We comparethesurvey results
ofthefirstandsecondmailingandfindfewsignificant differences betweenthetwogroupsof
respondents.
1Althoughnotshown,themediansofeachvariablearegenerally betweentheresponding
similar
andnonresponding firms.Theonlyexception payoutratio,whichis largerforthe
is thedividend
nonresponding tradegroupcomparedwithitsresponding
wholesale/retail counterparts.

This content downloaded from 128.111.121.42 on Thu, 29 Oct 2015 08:40:35 UTC
All use subject to JSTOR Terms and Conditions
QJBE, Spring 1999, Vol. 38, No. 2 25

Table 1- Characteristicsof Responding and NonrespondingFirms

This table shows the means of the respondingand nonrespondingfirmson five characteristicsfor the
manufacturing, wholesale/retailtrade,and utilitygroupsincludedin thefullsample. COMPUSTAT is used to
computetotalassets, sales, marketvalue of equity,dividendpayoutratio,and dividendyield in 1994. T-tests
in meansare conductedto determineiftherespondingand nonresponding
of difference firmsdiffersignificantly
on each characteristic.Because outliersaffectthemeans of thetotalassets,sales, and marketvalue of equity,
each of thesevariablesis convertedto logarithmsbeforeconductingthet-tests.
Mean of Finn Characteristic
Market
Total Value of
Assets Sales Equity Dividend Dividend
Industry ($ millions) ($ millions) ($ millions) Payout Yield
Panel A: Manufacturing
Responding $5,000.75 $4,965.01 $4,718.71 36.67% 2.27%
Nonresponding 4,021.68 3,599.43 4,009.19 47.37 2.48
t-statistic -0.03 -0.00 0.02 -0.09 0.03
Numberresponding 105 105 96 105 102
Numbernonresponding 274 274 235 273 271

Panel B: Whole/RetailTrade
Responding $2,750.43 $5,166.33 $4,152.81 23.00% 1.97%
Nonresponding 3,805.19 6,040.62 2,672.32 52.61 2.44
t-statistic 0.08 -0.01 0.12 -0.20 -0.32
Numberresponding 28 28 25 28 28
Numbernonresponding 60 60 54 60 58

Panel C: Utility
Responding $5,085.34 $1,951.64 $2,777.89 77.24% 6.22%
Nonresponding 4,897.41 1,945.54 2,534.92 70.71 6.21
t-statistic 0.05 0.03 0.00 0.13 0.00
Numberresponding 52 52 48 52 52
Numbernonresponding 58 58 50 58 58
None of thet-valuesis statisticallysignificantat the0.05 level

Panel A of Table 2 shows thatmostrespondentsagree thata change in dividendpolicy

affectsnotonly firmvalue (SI) but also its cost of capital (S10). More than90 percent
of the total respondentsalso agree with two otherstatements:an optimal dividend
policy strikesa balance between currentdividendsand futuregrowththatmaximizes
stockprice (S5) and a firmshould formulateits dividendpolicy to produce maximum
value for shareholders (S6). These findings are clearly contraryto Miller and
Modigliani' s propositionthatin a world withouttaxes no optimal dividend policy
exists. This view suggests thatone dividend policy is as good as any other. Such
evidence is not surprising because Miller and Modigliani's (1961) restrictive
assumptionsdo nothold in thereal world.This findingis also consistentwithprevious
surveyresearchby Baker, Farrelly,and Edelman (1985) involvingmanagementviews
of dividendpolicy.

This content downloaded from 128.111.121.42 on Thu, 29 Oct 2015 08:40:35 UTC
All use subject to JSTOR Terms and Conditions
26 Bakerand Powell

2 I •§ 3 ¿P^.-acSPfltffScfiPoágcSPoágíPuáfSÍPoís

3B.I 3 P £1 SI 3| SI S
S i ¿ ® ®' ®' ®' c5' °
g
Sei
f1 1
'_]3 c £ je
Q r-Tj-r-ooONQ-^t^tínOíNvo^-^or- - v>
o5&) oohi^oooívpcfl^oo-qqqOi^fSiSN '< ~ - -I ö - - -' -'
^peo 3 ÖÖÖÖÖÖÖÖÖ-

tP^o § m'or-vor^r^Tfo<Nro(Noo<N - - o *r> oo

TZ-ti- a> ' ' '
iriTtr^ro«-«rirn>^TtoooqaNoor^aN--<Np
2 - ' -"-'-' - - - - - Ö Ö Ö Ö Ö Ö © Ö Cp
5, G H

00^-^B T "fr po. oo^^^i^^. ^t °^lOri'^". ^^^^

■+■ ©«oodaovtN^Hòoôrioòrivovioíiooô
S3^<t3

** ■"
*■ ^" *~" ""^ "~" ^ "™" *""" *~-
O O

^ ^ g a>~ ^ooo-ooM^oi^-^'O^^fi^q*

S^Q) iS^ «îPpoopp^PP^^P^. ^ oq « r ^

1 SM
5 c h es ^ co .ha

if.1 e -1 1 I
I |f| o| | 1 |
|
111 II I Î .1 .1
il If |.í ¡i lili
IH! II it '■%-aî à *
iîxi 11 | 8 1 i 1
O
líll 11 II i I I I

'1 ^ Ï 1 |^ «8 fi ?•£ |o
ti s
g "g -1 'S §
■»oëv"5 1? 'S '-3 -ote

CO

Xîlîl <I| fl i I I if
S f^i! l£« «s o; - - »

This content downloaded from 128.111.121.42 on Thu, 29 Oct 2015 08:40:35 UTC
All use subject to JSTOR Terms and Conditions
QJBE,Spring1999,Vol. 38,No. 2 27

| S £5 2 £5 2^52^52^52^52^52^5
.è1
•rs oo no tj-i ml o ^- ol oo
.O ro OO
_§ P "~ ro|
Pi
Vì|
PI
n©
°°.
©
P
OM
PI
On
^".
O O © ©I ©I © © ©I ©

q
1-1 r->©r^rorooo ONfSoor^©^ovoiN(NTf©vOTtm^-oor^oo
Ososoq©©ON '
oq©r*;ONa?sNOrniNfOON©vq©oqoo©-^- h
3 ©©©-'-'© © - «' ©©'©'© ^ - *-' © ^-" ö -'©©-'-' -h'

S « h ^ oo a i1 Tj-r^'--'r^r^oo<NONfOforo«r>vo©i- m-^^o
Ö ©©VirOfO© ' '
fO©«nON©rnï^r^*- vOTf»- ^;vqON<N<N©
s ©'©"©©'©© - - ^ © - «' «-' © ©'-'©©'-*©©©©©©

-
îl oqsqr-v-jvooN TfrNO'vq(Nw-ir>:TrTfrnrnr---Pr^:ror-

T"
+ **i ^ ^ °i ^ ^ *0 "*. ^t ^ ^ "*. ^ *"*P ^ ^ ^t ^ ^ P ^" ~* ^
r^ooTtTf^t© ^'-yeoònvONKtinò-òvoooNN
n M 1^ - -* n fO<N(N«Ov£>TtfOfO'- tnmvO«nvovOrororo

fSiNoôootrioN ^"iion'noNooOi/iOÎoôoîd^oifS^vi
ro ^ (N fO fi co (N ^^ »-^ (N fS »-< (N >~^ »-^

Sb S ^ ^ R P! 22 fî K ^ ^ rn ö r-:¿ os oc jq - H ¿ ^ a H S
« 'o oo ^ f^ t oo©©u-ivo©rovoin«^©©(N©ONTtNqNq
q<n

I
r^roroV>"^-oÑ ^©©Ttro©<^ror^fO©©N¿©- ^Tf-roOÑ

42 ^ ^

a
s li ! 1 i í i I
i
I II Si I 1 I !
III i j Ii, I1 [i ¡ i|
Q

U 111 I .g
| || 1 ||
SjL s If . I ■& -8.2? I!
Jail le 1 1^ I -§^. •§|: 1 . I ^
Ì IéjÏ || g| f ¡i || || || ||
I g II
g « |
S<3
Tf
£&
ro "g < '•!
(X ¿
JSË.
Ü
< .^
'
< '«5
'
nS-K
*"*
S ai
^

This content downloaded from 128.111.121.42 on Thu, 29 Oct 2015 08:40:35 UTC
All use subject to JSTOR Terms and Conditions
28 BakerandPowell

rS <N *- ^ -^ ^o »n
.0 m - r- -* r- vo
*§ O Ö Ö Ö Ö Ö

«ft »O Vi ^ »û »û
Jn^^'O"rt»tt^nroVì

3 -'-'©© © ö ö ©-<'©© ö ööö-^ö^

C/3

ci oovoONt^r-o»n^^Ti-Tto<N -^oor^r^fso
û ypTf^tvi^o^pqpfn ^ ^ "O ^ °9 ■":
S òòòooòcp^òòòo ^H"^^ocpo

8
Sb
2?¡S2¿¿^¿^vci^^
vo
¿ $Söö~
^T *?Pririr^^aiv£îTÎ: P1^.^ ^ ^. ~ °° ~* Tt
Tj-íNTj-^-mrO(N(N(N(N - - * m Pi -

<Nf**><NrnrOTfTj-rorOTj-u-)m - mrorn

CK ¿ .n CK ¿ ^ vo wo VO o ^ r^ ro H
g' rq jq jn £
J
Q <N
^«n vq Os tj- o p r- ^ r~; ^- p uo on p p H ^p «^
nfo^tdòar^vitoK 6 ò d ^ 2S 3
ro (N <N

| I
| ||
Î
I ! fi l:-8 i li
I i I I ||
•S *^ j§ S S s £
|¡ II
i§ >ço ^ ä
1- s S •§ il € "8 J B§ g2

feu* T3 »73 w" ß ¿j -g E >rt Ü

X
éfj Iî I tï <lî p
§ g<^ £-5 « i £€ I h ^ p e
á £2 8 H ^ ÎS S

This content downloaded from 128.111.121.42 on Thu, 29 Oct 2015 08:40:35 UTC
All use subject to JSTOR Terms and Conditions
QJBE,Spring1999,Vol. 38,No. 2 29

îS ool col OO ^ © v©
•£> ml - I «n on« r* «n
co ool oq - ro oq
tt|
"§ ©ï öl ö ö o o

>
f*' ©voONvO<NV">r--co -
i»/">u*ioocoi--rNro©'^i
1-1 © vO ON oq On 0O ON On On O^ On On <N •- © CN <N <N g
3 -* ^ Ö Ö Ö Ö Ö Ö Ö Ö Ö Ö -' - -''-''-''-' &,

S ©ONoo«/->ON<Noo«/->Tft^"<*<NO»r>>/"ìxOTrr-- .ti M
o ' "
TtoqropoqONpt^ONu^vqONTt(Nr*:pP'-
- ö - ! - ö ö -^"ö o ö ö o ö ö ö ö ö ö S*rt
2 *£ |S
•o 5
^ II rt
3^+ i:t^o'-'--Tf:oÑ(^d(Nr^v¿o'Tt't^r>:in - +J2S
ÇÎ vOrovOmfNeNrn-Hrn-^ <N - -^<N-^ »-h -tì-ii
^ g g
^ T ^. n ^ ^ ^ °i ^ °. ^ ^. °i ^ ^ ^ t ^ ^ rt. - >>
+ .-«"oÑvooÑt^or^©-^ oorifS^-rirnriioiN â^S
r^rnfSTf»n»rjro»r»TfTtTf»r)rorOTÍ-rornro tt'S

JO i
^

Oû ^^ - -^íN»- "íN-^ roil«/-)

cd •- i>

c
o S g -a I t ^ Si«
,0,

Í I" g1 S£ 1 1 sI |S2
2 cö
Io £~
S
çq
'S
^^ ^^%
lj-
C cd 11

> 1 I ï 1 I »Ha !
| f 1 1 | f* l||î I
O
4 gî I 1 î • fi HffVf
Ms « « ! Sì li -iîlis
fii 1 l.f || il HI 111
X
n sili < <1 ßi <î <l |?ì|l!
ff

This content downloaded from 128.111.121.42 on Thu, 29 Oct 2015 08:40:35 UTC
All use subject to JSTOR Terms and Conditions
30 BakerandPowell

The respondentshold views that conflictwith a primarytenet of corporation

financetheoryinvolvingtheseparationof a firm'sinvestment, financing,and dividend
decisions. Originally proposed by Modigliani and Miller (1958), the separation
principle asserts that,in the presence of perfectcapital markets,a firm's value is
completelyindependentof themannerin whichthefirmfinancesitsproductiveassets.
More than88 percentof therespondentsbelieve thata firm'sinvestment, financing,and
dividenddecisions are related(S9).12 The respondents,
however,hold widelydivergent
views about whethera firmshould view cash dividends as a residual afterfinancing
desired investmentsfromearnings(S8). An implicationof theMiller and Modigliani
(1961) argumentis thatthedividenddecision is a residualdecision. Taken as a whole,
the evidence suggests that most of the respondentsare proponents of dividend
relevance.

Explanations of Dividend Relevance

The second researchquestionwas "What explanationsof dividendsdo managers
tendto favor?"Panel B of Table 2 providestheresultsforfourpopularexplanationsof
dividend policy: the bird-in-the-hand,signaling, tax preference, and agency
explanations.

The Bird-in-the-HandExplanation
Two statementsreflectthe explanation of dividend relevance: investorsprefer
certain,currentdividends to possibly higherbut riskierfuturedividends (S24) and
investorsprefera certaindividendstreamto an uncertainprice appreciation(S23). The
responsesproduce mixed resultson thesestatements.About a thirdof therespondents
neitheragree nor disagree witheitherstatement.

The Signaling Explanation

Six statementsinvolve informationcontentor signalingeffects(S2, S3, SI 1, S 12,
S 13, and S 14). The respondentsagree, on average, that investorsregard dividend
changes as signals about a firm'sfutureprospects(SI 2) and investorsuse dividend
announcementsas informationto assess a firm'sstock value (SII). The respondents
also agree withconclusions of empiricalstudies thatdividend changes convey some
unanticipatedinformation to themarket(S2 and S3). The respondentsshow a highlevel
of agreementwith the notion thata firmshould adequately disclose to investorsits
reasonsforchangingdividendpolicy(S14). The responsesvaryaboutwhetherdividend

l2The empirical literatureon Modigliani and Miller's (1958) investment-financingirrelevance

propositionsreaches divergentconclusions. Some researchers(Petersonand Benesh, 1983) find
a significantlinkage between the investment,financing,and dividend decisions of firmswhile
otherresearchers(Smirlock and Marshall,1983 ) do not.

This content downloaded from 128.111.121.42 on Thu, 29 Oct 2015 08:40:35 UTC
All use subject to JSTOR Terms and Conditions
 1999,Vol.38,No.2
QJBE,Spring 3]^
increases are ambiguous. Taken as a whole, however,the responses suggest general
agreementthatchanges in dividendshave signalingeffects.

The Tax-Preference Explanation

The fourstatementsinvolvingthe tax-preference explanationof dividends (SI 9,
S20, S21, and S22) commandmixed agreements.Most respondentsagree thata firm
should be responsive to the dividend preferencesof its shareholders(SI 9) and that
investorsare attractedto firmsthathave dividendpolicies appropriateto theinvestors'
particulartax circumstances(S20). Yet, themosttypicalresponseis uncertainty (neither
agree nordisagree).Respondentsare uncertainaboutwhetherstocksthatpay high(low)
dividendsattractinvestorsin low (high) tax brackets(S22). They are also unsureabout
whetherinvestorspreferthata firmretainsfundsover payingdividendsbecause of the
way capital gains are taxed as comparedwithdividends(S21).13

The Agency Explanation

The surveycontainstwo statements thatprovidean agencyexplanationforpaying
dividends (S25 and S26). More than90 percentof the totalrespondentsthinkthatthe
paymentof dividendsforcesa firmto seek moreexternalfinancing,whichsubjectsthe
firmto scrutinyof investors(S25). The respondentsgenerallydisagreethatthepayment
of dividends serves as a bonding mechanism to encourage managers to act in the
interestof outside shareholders(S26). If respondentswere unfamiliarwith agency
theory,thenthe phrase bonding mechanismmightconnote thatbonding has a legal
meaning or somethingstrongerthan a simple economic incentive. This potential
miscommunicationmightpromptsome managersto disagree withthe statement.The
responsesconcerningtheagency explanationforpayingdividendsappear mixed. Yet,
theyshould not be discardedor discountedbased solely on the survey's results.

Setting Dividend Payments

Our thirdresearchquestion was "How do firmsset the amountof dividendsthat
they pay?" Several statementsin the survey reflectthe dividend policy beliefs of
managers,as describedby Lintner(1956), and Lintner'sbehavioralmodel of dividend
policy,namely,S4, S7, S15, S16, S17, and S18. Taken as a whole,thesurveyresponses
were generallyconsistentwithLintner'sfieldworkand dividendmodel.
The resultsshow thatseveralstatements commandeda highlevel of agreement.For
example, almost 85 percentof the totalrespondentsbelieve thata firmshould avoid

l3Omittingfirmsthatchose not to pay dividendsfromthe surveycould bias the findingsabout

thetax-preference explanation.For example,a highpercentageof respondents,especially among
tax rateson dividendincome and capital gains are not an
utilities,indicatesthatthe differential
important considerationfor investors(S21). Firmsthatchose notto pay dividendsmay view this
factoras an importantconsideration.

This content downloaded from 128.111.121.42 on Thu, 29 Oct 2015 08:40:35 UTC
All use subject to JSTOR Terms and Conditions
32 Bakerand Powell

changingitsregulardividendifthatchangemighthaveto be reversedin a yearor so

(SI 6). Aboutthreequarters ofthetotalrespondents also agreethata firmshouldstrive
tomaintain or
steady modestly growing dividends (SI 8) andshouldstrivetomaintain
an uninterrupted recordofdividendpayments (SI 7).
The currentfindingsalso confirm managerial perceptions reportedbyLintner that
themarket valuesstabledividends thanstablepayoutratios(S4).14Less thanhalf
rather
therespondents, however,agreethata firm'snew capitalinvestment requirements
generally havelittleeffecton modifying itspattern ofdividends (7). Thislatterfinding
is inconsistentwithLintner' s fieldwork.
Lintnersuggests that
a partialadjustment modelcanincorporate dividendbehavior.
According to his model, a firm should have a target dividend payout ratio and
periodicallyadjustthepayouttowardthetarget (SI 5). Thesurveyresults providesome
supportforLintner'sbehavioralmodelofdividendpolicy.

IndustryInfluence on Dividend Policy

The finalresearchquestionis "Do theviewsof managersaboutdividendissues
differamongdifferent industry groups?"The resultsofthechi-squaretestsshowthat
theresponsesofthethreegroupsdiffer at the0.05 levelforonlyfourof
significantly
the26 statements (SI, S2, S 14,andS24). Thisevidencesuggeststhata firm'sindustry
typehas little
influence on theviewsthatmanagers haveabouttheoretical andempirical
issues involvingdividendpolicy.The currentfindingscontrastsharplywiththe
conclusionofBaker,Farrelly, andEdelman(1985) thattheopinionsoftherespondents
fromutilitiesdiffer markedly fromthoseof themanufacturing and wholesale/retail
tradeindustries.Theseresultsmaybe due to thegreater competitiveenvironment in
whichutilitieshavebeenthrust the
during past decadeor more.As Soter,Brigham, and
Evanson(1996) note,theeconomicenvironment forutilities
has beenchangingover
time.Due toregulator actions,theutility industryhasbecomea riskier place inwhich
to operateand invest.Today, utilitiesfind themselvesincreasinglysubject to
competition.
The respondents fromtheutilities showa higherlevelofagreement on all fourof
thesestatements thantherespondents fromtheothertwoindustry groups.Analysisof
thesestatements suggeststhattheresultsmaybe dueto industry characteristicsrather
thanregulation.Given thatthe utilitieshave higherpayoutsthanthe othertwo
industries(manufacturing and wholesale/retail trade),such differences are notsur-
prising.The surveyfindings showthatmanagersof utilitiesare sensitiveto how an
unexpected changeindividendpolicy(SI andS2) mayaffect a firm'sstockpriceand
shareholdervalue.In addition, mostutility respondentsstronglyagreethata firm should
adequately discloseto investors itsreasons forchangingdividendpolicy(S14). This

i4Joseand Stevens (1989) empiricallyfindsupportforthe notionthatthe marketvalues stable

dividends ratherthanstable payoutratios.

This content downloaded from 128.111.121.42 on Thu, 29 Oct 2015 08:40:35 UTC
All use subject to JSTOR Terms and Conditions
QJBE,Spring1999,Vol. 38,No. 2 33

findingis not surprisingconsideringthemarketreactionto thedividendcut of Florida

Power & LightCompanyreportedby Soter,Brigham,and Evanson (1996). Finally,the
high level of agreementamong utilitymanagersthatinvestorsprefercertain,current
dividendsto possiblyhigherbutriskierfuturedividendsmay reflectthecharacteristics
of utilitiesincludinginvestorclienteles.

Conclusions
The findingsof thissurveylead to severalconclusionsabout dividendpolicy. First,
most respondentsbelieve thatdividend policy affectsfirmvalue. Although Baker,
Farrelly,and Edelman (1985) reachedthesame conclusionmorethana decade ago, the
currentstudyprovidesadditionalsupportformanagerialbeliefsabout therelationship
between dividendpolicy and value.
Second, of thefourexplanationsfordividendrelevanceexaminedin thisstudy,the
respondentsgenerallyhad the highestlevel of agreementwith statementsinvolving
signaling. The respondentstypicallyare most uncertain(neitheragree nor disagree)
about statementsinvolving the tax-preferenceand bird-in-the-hand explanations of
dividendrelevance.As previouslynoted,however,omittingfirmsthatchose notto pay
dividends from the survey could bias the findingsinvolving the importance of
differentialtax rates. Although the responses are inconsistentinvolvingthe agency
explanation of paying dividends, respondentsmay not have fully understood the
meaningof these questions.
Third,theresultsshow thatmanagers' views on settingdividendpaymentstoday
are consistentwith those reportedby managers interviewedby Lintner(1956). In
particular,the respondentsare highlyconcerned about continuityof dividends. The
surveyresultsalso provide some supportforLintner's behavioral model of dividend
policy.
Finally,the opinions of therespondentsfromthethreeindustrygroups show few
statistically differences
significant This findingcontrastswith
among the26 statements.
earliersurveyresearchby Baker,Farrelly, and Edelman (1985). The differences
among
industriesmay have diminishedover time because of the changing economic and
competitiveenvironmentforutilities.

References
1. Ang,J.S.,Do DividendsMatter?A ReviewofCorporate andEvidence,
DividendTheories (New
York:SalomonBrothers New YorkUniversity,
CenterfortheStudyofFinancialInstitutions, 1987).
2. Asquith,P., and D.W. Mullins,"The Impactof Initiating on Shareholder
DividendPayments
Wealth"JournalofBusiness,56,no. 1 (January1983),pp. 77-96.
3. andtheInformation
Bajaj, M.,andA. Vijh,"DividendClienteles ofDividendChanges,"
Content
JournalofFinancialEconomics,26, no. 2 (August1990),pp. 193-219.
4. Baker,H.K.,"TheRelationship BetweenIndustry andDividendPolicy,"Southern
Classification
BusinessReview,14,no. 1 (Spring1988),pp. 1-8.

This content downloaded from 128.111.121.42 on Thu, 29 Oct 2015 08:40:35 UTC
All use subject to JSTOR Terms and Conditions
34 Baker and Powell

5. Baker, H.K., G.E. Farrelly,and R.B. Edelman, "A Survey of Management Views on Dividend
Policy," Financial Management, 14, no. 3 (Autumn 1985), pp. 78-84.
6. Barclay, M.J., C.W. Smith, and R.L. Watts, "The Determinantsof Corporate Leverage and
Dividend Policy," Journal of Applied Corporate Finance, 7, no. 4 (Winter 1996), pp. 4-19.
7. Benesh, G.A., A.J. Keown, and J.M. Pinkerton,"An Examination of Market Reaction to
Substantial Shiftsin Dividend Policy," Journal of Financial Research, 7, no. 2 (Summer 1984), pp. 131-
140.
8. Benartzi, S., R. Michaely, and R. Thaler, "Do Changes in Dividends Signal the Futureor the
Past?" Journal of Finance, 52, no. 3 (July 1997), pp. 1007-1034.
9. Bernstein,P.L., "Dividends: The Puzzle," JournalofApplied CorporateFinance, 9, no. 1 (Spring
1996), pp. 4-15.
10. Bhattacharya,S., "ImperfectInformation, Dividend Policy,and 'The Bird in the Hand' Fallacy,"
Bell Journal of Economics, 10, no. 1 (Spring 1979), pp. 259-270.

11. Bhattacharya,S., "NondissipativeSignaling Structuresand Dividend Policy," QuarterlyJournal

of Economics, 95 (August 1980), pp. 1-14.
12. Black, F., "The Dividend Puzzle," Journalof PortfolioManagement,2, no. 2 (Winter 1976), pp.
5-8.
13. Black, F., and M. Scholes, "The Effectsof Dividend Yield and Dividend Policy on Common
Stock Prices and Returns,"Journal of Financial Economics , l,no. 1 (May 1974), pp. 1-22.
14. Born, J.A., and J.N. Rimbey, "A Test of the Easterbrook Hypothesis Regarding Dividend
Paymentsand Agency Costs," Journal of Financial Research, 16, no. 3 (Fall 1993), pp. 25 1-260.
15. Brennan,M., "Tax Reformand the Stock Market: An Asset Price Approach," American Eco-
nomic Review, 23, no. 4 (December 1970), pp. AM All.
16. Christie,W.G., "Are Dividend Omissions Trulythe Crudest Cut of All?" Journal of Financial
and QuantitativeAnalysis,29, no. 3 (September 1994), pp. 459-480.
17. Crutchley,C.E., and R.S. Hansen, "A Test of the Agency Theory of Managerial Ownership,
CorporateLeverage, and CorporateDividends," Financial Management,18, no. 4 (Winter1989), pp. 36-46.
18. Easterbrook,F.H., "Two Agency-CostExplanationsof Dividends,"AmericanEconomic Review,
74, no. 3 (September 1984), pp. 650-658.
19. Farrelly,G.E., H.K. Baker, and R.B. Edelman, "Corporate Dividends: Views of the Policy-
makers,"AkronBusiness and Economic Review, 17, no. 4 (Winter 1986), pp. 62-74.
20. Fowler, Jr.F.J.,SurveyResearch Methods (Beverly Hills, CA: Sage Publications, 1984).
21. Ghosh, C. and J.R. Woolridge, "An Analysis of Shareholder Reaction to Dividend Cuts and
Omissions," Journal of Financial Research, 1 1, no. 4 (Winter 1988), pp. 28 1-294.
22. Healy, P.M. and K.G. Palepu, "Earnings InformationConveyed by Dividend Initiationsand
Omissions," Journal of Financial Economics, 21, no. 2( May/September1988), pp. 149-176.
23. Impson, M., "Market Reaction to Dividend-Decrease Announcements: Public Utilities vs.
UnregulatedIndustrialFirms,"Journal of Financial Research, 20, no. 3 (Fall 1997), pp. 407-422.
24. Jensen,M.C. and W.H., Meckling,"Theoryof theFirm:Managerial Behavior,AgencyCosts and
Capital Structure,"Journal of Financial Economics, 3, no. 4 (October 1976), pp. 305-360.
25 . Jose,M .L. and J.L. Stevens,"Capital MarketValuation of Dividend Policy,"Journalof Business
Finance & Accounting, 16, no. 5 (Winter 1989), pp. 651-662.
26. Lintner,J.,"Distributionof Incomes of CorporationsAmong Dividends, Retained Earningsand
Taxes" American Economics Review,46, no. 2 (May 1956), pp. 97-1 13.

This content downloaded from 128.111.121.42 on Thu, 29 Oct 2015 08:40:35 UTC
All use subject to JSTOR Terms and Conditions
QJBE,Spring1999,Vol. 38,No. 2 35

27. Long,Jr.,J.B.,"The MarketValuationof Cash Dividends:A Case to Consider,"Journalof

1978),pp. 235-264.
FinancialEconomics,6, no. 2/3(June/September
28. Michaely,R., R.H. Thaler,and K.L. Womack,"PriceReactionsto DividendInitiations
and
JournalofFinance,50, no. 2 (June1995),pp. 573-608.
or Drift?"
Omissions:Overreaction
29. Michel,A., "IndustryInfluence 8, no. 3 (Autumn
on DividendPolicy,"FinancialManagement,
1979),pp. 22-26.
"
30. Miller,M.,andF. Modigliani,"DividendPolicy,Growth, ofShares, Journal
andtheValuation
ofBusiness,34, no. 4 (October1961),pp.41 1-433.
3 1. Miller,M.,andK. Rock,"DividendPolicyUnderAsymmetric Journal
Information," ofFinance,
40, no. 4 (September1985),pp. 1031-1051.
32. Miller,M., and M.S. Scholes,"DividendsandTaxes: Some EmpiricalEvidence,"Journalof
PoliticalEconomy,90 ( 1982), pp. 1118-114 1.
33. Finance,and theTheoryof
Modigliani,F., and M.H. Miller,"Cost of Capital,Corporation
Investment,"American EconomicsReview,48 (June1958),pp. 261-297.
34. Moh'd, M.A., L.G. Perry, of theDynamicRelationship
and J.N.Rimbey,"An Investigation
BetweenAgencyTheoryand DividendPolicy,"FinancialReview,30,no. 2 (May 1995),pp. 367-385.
35. Ofer,A., and A. Thakor,"A Theoryof StockPriceResponsesto Alternative Cash
Corporate
Methods:StockRepurchases
Disbursement JournalofFinance,42,no. 2 (June1987),pp.
andDividends,"
365-394.
36. Peterson,P.P., and G.A. Benesh,"A Reexaminaron of theEmpiricalRelationship Between
Investment andFinancing Analysis,18,no.4 (December
Decisions,"JournalofFinancialand Quantitative
1983),pp. 439-453.
37. Pruitt,S.W., and L.J.Gitman,"The InteractionsBetweenthe Investment, Financing,and
DividendDecisionsofMajorU.S. Firnis,"FinancialReview,26, no. 3 (August1991), pp.409-430.
38. Ross,S., "The Determination The Incentive
ofFinancialStructures: SignalingApproach," Bell
JournalofEconomics,8 (1977), pp. 23-40.
39. Rozeff,M.S., "Growth, Betaand AgencyCostsas Determinants of DividendPayoutRatios,"
JournalofFinancialResearch,5, no. 3 (Fall 1982),pp. 249-258.
40. Smirlock,M., and W. Marshall,"An Examination Betweenthe
oftheEmpiricalRelationship
Dividendand Investment Decisions:A Note,"JournalofFinance,38, no. 5 (December1983), pp. 1659-
1667.
41. andP. Evanson,"The DividendCutHeard'roundtheWorld:The Case
Soter,D., E. Brigham,
ofFPL/1JournalofAppliedCorporateFinance,9,no. 1 (Spring1996),pp.4-15.
42. Stapleton, Value,"Economic
R.C.,"Taxes,theCostofCapitaland theTheoryof Investment
Journal,82 (December1972),pp. 1273-1292.
43. "The MarketValuationof Cash Dividendsand theTax
Sterk,W.E., and P.A. Vandenberg,
FinancialReview,25,no. 3 (August1990),pp.
TheoryofDividendPolicy:A Case Revisited,"
Differential
441-455.
44. Woolridge,J.R.,"The Information Contentof DividendChanges,"Journalof Financial
Research,5, no. 3 (Fall 1982),pp. 237-247.
45. Woolridge,J.R.,"DividendChangesand SecurityPrices,"Journalof Finance,38, no. 5
(December1983),pp. 1607-1615.

This content downloaded from 128.111.121.42 on Thu, 29 Oct 2015 08:40:35 UTC
All use subject to JSTOR Terms and Conditions