SHELBY D. HUNT and JOHN R. NEVIN^

This article empirically assesses the relationship between power and thesources of power in a channel of distribution that has a well-defined powersource: the franchisor-franchisee channel. In addition, the consequences offranchisors utilizing coercive versus noncoercive sources of power are examined.

Power in a Channel of Distribution: Sources

and Consequences

INTRODUCTIONThe concept of power is considered central in

understanding the means by which one channelmember can change or modify the behavior of anothermember within its channel of distribution [1,3,5,7,8,11,14,17,18,19,21]. The recent article by El-Ansaryand Stern is the only published effort to measureempirically power in a channel of distribution [5].El-Ansary and Stern tested the hypothesis that thepower of any given channel member is a functionof the sources of power available to him at any giventime. In testing this hypothesis, no significant relation-ship was found between power and the sources ofpower. El-Ansary and Stern attributed this findingto the fact that a well-defined power structure didnot exist in the specific channel of distribution theystudied.

This article will empirically assess the relationshipbetween power and the sources of power in a channelof distribution that has a well-defined power source:the franchisor-franchisee channel. In addition, theconsequences of franchisors utilizing coercive versusnoncoercive sources of power are examined.

THEORETICAL EOUNDATIONS

Power in a Channel of DistributionPower, in its most general sense, refers to the ability

of one individual or group to control or influence

*Shelby D. Hunt is Associate Professor of Business, and JohnR. Nevin is Assistant Professor of Business, Graduate School ofBusiness, University of Wisconsin-Madison. They acknowledge theSmall Business Administration for financially supporting the studyon the economic effects of franchising [10] which served as adata base for this research, and William H. Peters for his assistancewith MCA analysis.

the behavior of another. Dahl, for example, definespower as the ability of one individual or group toget another unit to do something that it would nototherwise have done [4, p. 203]. El-Ansary and Sternapplied this notion to distribution channels by opera-tionally defining power as the ability of a channelmember to control the decision variables in the mar-keting strategy of another member at a different levelin the channel of distribution [5, p. 47].

In the franchisor-franchisee channel of distribution,the direction of power for most decisions has alreadybeen built into the relationship by means of thefranchise contract. Previous analysis has suggestedthat most franchisees enter the negotiations at anextreme bargaining disadvantage vis-a-vis the franchis-or and that franchisors go to great lengths in develop-ing the contract to protect their own prerogatives [10,p. 261]. Therefore, power in a franchisor-franchiseechannel usually refers to the ability of the franchisorto control the business decision variables of thefranchisee.

Power as a Eunction of Sources of PowerSimon has suggested that an index of power might

be determined from the magnitude of the powersources [16]. El-Ansary and Stern applied this notionto distribution channels by indicating that the powerof any given channel member is probably a functionof the sources or bases of power available to himat any given time [5, p. 48]. Although there are manypossible sources of power, French and Raven havedefined five which seem common and important:reward, coercive, legitimate, referent, and expertpower [6]. A model of power in a channel of distribu-tion relying on the work of Simon and French andRaven could be depicted as follows:

186

Journal of Marketing ResearchVol. XI (May 1974), 186-93

POV/ER IN A CHANNEL OF DISTRIBUTION 187

where:

Pj, = power of channel member i over /,Rjj = reward power of i over /, i.e., power based

on the belief by / that i has the ability tomediate rewards for him,

Cjj = coercive power of /over / , i.e., power basedon the anticipation of the part of /of possiblepunishment by i if he fails to yield to theinfluence attempt,

Ljj. = legitimate power of i over /, i.e., poweroriginating from internalized values in /which dictate that i has a legitimate rightto influence /and that /is obligated to acceptthis influence,

Fjj. = referent power of i over /, i.e., power basedon the identification of / with i whereidentification means a feeling of onenessor a desire for such an identity, and

£,,. = expert power of i over /, i.e., power basedon the extent of knowledge which / attri-butes to I within a given area.

In an empirical case, coercive power can be dif-ferentiated from the others because it, alone, involvespotential punishment. For all the other noncoercivesourcesof power, i.e., reward, legitimate, expert, andreferent, the individual willingly (rather than begrudg-ingly) yields power to another. Because the variouspower sources dichotomize meaningfully into coerciveand noncoercive sources, and because other re-searchers [3, p. 113] have pointed out the extremedifficulty of empirically differentiating the variousnoncoercive sources of power, the research reportedhere tested the following power model:

(2) P.. = f(C..,N..)

where:

P,j = power of channel member i over member

i,C,j = coercive power of / over /, andNjj = noncoercive power of i over / , i.e., power

that / willingly yields to i because / believesthat: (1) i has the ability to mediate rewardsfor him; (2) i has a legitimate right toprescribe behavior for him; (3) he has anidentification with (,or (4) has some specialknowledge or expertise in a given area.

This modified model of power is shown schematicallyin the figure.

Personal interviews with franchisees during theexploratory phase of the research served as a basisfor suggesting the following research hypothesis:

H I: Franchisors will rely primarily on coercive sourcesrather than noncoercive sources to achieve powerover their francisees.

MODIFIED MODEL OF POWER

REWARD

EXPERTISE

LEGITimCY

REFERENT

COERCIVESOURCES

NONCOERCIVESOURCES

POWER

Consequences of Exercising the Sources of Power

The franchisor-franchisee relationship is essentiallya channel of distribution with a unilateral dependencyrelationship. The franchisor, being the strong channelmember, has coercive and noncoercive sources ofpower that can be exercised to control the franchisee.No published research has empirically examined theconsequences of exercising coercive versus noncoer-cive sources of power in a channel of distribution.Some authorities, however, have suggested that thenoncoercive sources of power appear to be betteralternatives for enhancing the satisfaction of the weakchannel members [3,6]. If this hypothesis is true,franchisors can attain higher levels of satisfactionamong franchisees by using noncoercive sources thanby using coercive sources of power. This issue isimportant because franchisees who are more satisfiedwiil probably: (1) have higher morale; (2) cooperatewith the franchisor better; (3) be less likely to termi-nate the contract voluntarily; (4) be less liJcely to fileindividual suits against the franchisor; (5) be less likelyto file class action suits, as in the Chicken Delightcase [15]; and (6) be less likely to seek protectivelegislation lil:e the "Franchise Full Disclosure Actof 1970"orthe "Franchise Competitive Practices Act"proposed by Senator Philip Hart of Michigan [10,pp. 1-4]. The preceding discussion suggests the fol-lowing hypothesis:

H,: Franchisors who rely less on coercive sources andmore on noncoercive sources will have franchiseeswho are more satisfied with the franchise relation-ship.

188 JOURNAL OF MARKETING RESEARCH, MAY 1974

METHODEmpirical measures of power, coercive sources,

noncoercive sources, and satisfaction of franchiseewere needed to test the two research hypotheses. Thedata used to develop these empirical measures camefrom a recently completed study on the economiceffects of franchising conducted at the University ofWisconsin-Madison [ 10]. The data base is a probabilitysample of approximately 950 completed questionnaires(26% response rate) from franchisees in the fast foodrestaurant area. Only 815 fast food franchisees wereused in this study after deleting subjects for noresponse.

Measuring Power

Power does not refer to the objective ability of oneindividual or group to control or influence the behaviorof another, but rather to the potential ability of thecontrolling or influencing agent as perceived by thecontrollee or influencee. March expressed thisviewpoint when he suggested that attributed influencebe used to measure power directly [9]. Raven ex-pressed a similar viewpoint when he stated that:

. . . it is not the objective ability of the influencingagent to mediate such rewards and punishments thatis important, but rather the potential rewards andpunishments as perceived by the influencee [13, p.373].

Therefore, power in this study refers to the abilityof the franchisor, as perceived by the franchisee, tocontrol the decision variables of the franchisee. Sevenfranchisee decision variables were used to developan index of perceived power: (1) determining the hoursof operation; (2) specifying the bookkeeping system;(3) adding or deleting items from the product line(menu control); (4) approving the content and mediafor local advertising; (5) setting the retail prices forproducts and services; (6) determining the appropriatestandards of cleanliness; and (7) determining thenumber of employees. The extent of the franchisor'spower would seem to be related to the degree of controlhe was perceived to have over these seven key decisionareas.

Each franchisee indicated the degree to which hehad "responsibility for" (a euphemism thought to besemantically preferable to "control over") each ofthe decision areas. In each case, a rating scale rangingfrom one through six was used as a measuring tool(see Appendix). An index of power (V,), defined asthe mean of the franchisee's responses for each ofthe seven decision areas, was used as an empiricalmeasure of power with large numbers indicating highfranchisor control.

Noncoercive Sources of Power

In an empirical case, noncoercive sources of powercan be distinguished from coercive sources of power

in that they involve a willingness on the part of oneindividual or group to yield power to another. In thefranchisor-franchisee channel, franchisors providefranchisees with several types of assistances designedto influence franchisee behavior. To the extent thatthese assistances are of high quality, they establishthe franchisor as an expert in the eyes of the franchisee;they legitimize the franchisor's efforts to gain power;and they help to get the franchisee to yield powerwillingly to the franchisor. The franchisor assistancesused as measures of noncoercive sources of powerin this study include: site location assistance (X,);formal training at a central location (Xj); "on-the-job''training (Xj); equipment sold to franchisee by franchi-sor (X4); operating manual (Xj); bookkeeping assis-tance (Xg); methods of product preparation (X^); fieldsupervisors (Xg); supplies sold to franchisee byfranchisor (X,); deletions and additions to productline (X,o); day-to-day business advice (X,,); pricingassistance (X,2); national advertising program (X^);and local advertising assistance (X,^).

Each franchisee indicated the quality of each ofthe assistances provided by his franchisor on a ratingscale ranging from one to five (see Appendix). Thefranchisee's quality ratings for his franchisor's assis-tances were used as empirical measures of thesenoncoercive sources of power with large numbersindicating high quality franchisor assistance.

Coercive Sources of Power

In an empirical sense, coercive sources of powercan be differentiated from the noncoercive sourcesof power in that they involve potential punishment.A franchisor has coercive power over a franchiseewhen the franchisee anticipates possible punishmentif he fails to yield to the franchisor's influence attempt.There are various coercive sources of power that afranchisor can use to get a franchisee begrudgingly(rather than willingly) to yield power. A franchisor'scoercive sources of power are frequently built intothe franchisor-franchisee relationship during the nego-tiation phase which precedes the granting of a franchiseto a potential franchisee.

This research employed six measures to attemptto identify empirically potential coercive sources ofpower:

1. Control of building {X^^)•. A dummy variable coded" 1 " if the building was owned or controlled by thefranchisor, and coded " 0 " if not.

2. Control of land (X,J: A dummy variable coded " 1 "if the land was owned or controlled by the franchisor,and coded " 0 " if not. .

3. Threaten to revoke franchise (X,,): A dummy variablecoded " 1 " if a franchisor has threatened to revokea franchisee's franchise, and coded "0" if not.

4. Need for legislation (X.^): Each franchisee indicatedthe extent of need for federal legislation that wouldrestrict the right of franchisors to terminate a franchiseon a rating scale having values from one through five

POWER IN A CHANNEL OF DISTRIBUTION 189

Table 1MCA ANALYSIS

Table 1 (continued)

PERCEIVED POWER AS A FUNCTION OFNONCOERCIVE SOURCES OF POWER''

Noncoercive sources of power

Num- Name andber response category

Y Grand meanX|4 Local advertising

1 &2345

X|2 Pricing assistance1 &23

Adjustedcoeffi-cients

2.30

.37-.17-.06-.21

-.21-.11

P

.205

.190

F-ratio

13.77"

11.80"

Noncoercive sources of power

Num- Name andber response category

X|3 National advertising1 &23

X, Formal training1 & 2.345

Adjusted R^

Adjustedcoeffi-cients

- .04.03

-.04C\A

.08- .03

01.00

= .171

P F-ratio

.032 .33

.030 .28

45

X^ Equipment1 &2345

X,g Deletions & additions1 &2345

X, Product preparation1 &2345

Xg Supplies1 & 2 & 345

Xj Bookkeeping1 &2345

Xg Field supervisors1 &2345

X| | Day-to-day advice1 &2345

Xj "On-the-job" training1 &2345

Xj Operating manual1 &2345

X| Site locationI &2345

.36

.42

.34

.13

.02

.26

.09

.13

.20

.30

.23

.14

.05

.22

.15

.03

.25

.03

.07

.02

.38

.10

.08

.25

.17

.12

.19

.19

.59

.10

.05

.04

.06

.18

.02

.01

.03

.05

.00

.07

.03

.149 7.21'

.142 6.64"

.132 5.66"

.124 7.55"

.121 4.81"

.117 4.51"

.082 2.19

.051 .83

.044 .63

.032 .34

"The order in which the independent variables are arranged isby descending Beta (P) values.

" p < .01.

with large nutnbers indicating high need (see Appendix).5. Right to sell franchise (X,,): Each franchisee indicated

the extent of his right to sell his franchise on a ratingscale, ranging from one through five with large numbersindicating that the franchisor has greater control overthe resale of the franchise (see Appendix).

6. Fairness of agreement (X^g): Each franchisee indicatedthe extent to which the agreement favored either thefranchisor or the franchisee on a rating scale havingvalues from one through six with large numbers indi-cating more weight in favor of the franchisor (seeAppendix).

Control of the building or land by the franchisoris an indicant of coercive power because of thefranchisor's right to dispossess the franchisee phys-ically. Similarly, threats to revoke the franchise andneed for legislation are indicants of coercive powerbecause they measure the franchisee's real and per-ceived fear of having his franchise revoked. Limitinga franchisee's right to sell his franchise is also anindicant of coercive power in that the franchiseeunwillingly yields control of this decision to thefranchisor. Fairness of the agreement is a globalmeasure indicating coercive power in that a franchiseagreement weighted in favor of the franchisor alsoforces the franchisee to yield some control of hisdecision areas unwillingly to the franchisor.

Franchisee's Level of Satisfaction

The overall level of satisfaction of the franchiseewith the franchisor was measured by asking thefranchisee what he would do "if he had it to do overagain." The franchisee's level of satisfaction (Vj) isa dummy variable coded " 1 " if the franchisee wouldstill choose to be a franchisee with this franchisesystem, and coded "0" if not.

190 JOURNAL OF MARKETING RESEARCH, MAY 1974

TESTS AND RESULTS

Multiple Classification Analysis (MCA)

Multiple Classification Analysis (MCA) was usedin testing the research hypotheses [2]. The MCAprogram is a convenient variant of dummy variablemultiple regression, which, like all dummy variableregression programs, does not require intervally scaledindependent variables. The model assigns a coefficientto each category of each independent variable andexpresses the dependent variable as the sum of thegrand mean, the coefficients for each category, andan error term. Thus,

(3)

where:

Y •

Y •

Y= Y+ ... + e

e =

a particular value of the dependent variable,the grand mean of the dependent variable,the effect of membership in the ith categoryof independent variable X,,the effect of membership in the jth categoryof independent variable Xj, anderror.

Results of the MCA AnalysisIn testing H, , three separate MCA analyses were

run with the index of perceived power {Y^) as thedependent variable. Table 1 shows the results of theMCA analysis with perceived power (Y^) as thedependent variable and the noncoercive sources ofpower (X,, X2, ..., X,4)as the independent variables.The variables are arranged in order of descendingBeta (3) values. The p statistic measures the abilityof the independent variable to explain variation inthe dependent variables after adjusting for the effectsof all other independent variables.' The independentvariable X^^ (local advertising) has the greatest abilityto explain variation in Y^ and is displayed first. Inaddition, the table shows the adjusted coefficientsfor each category of each independent variable andF-ratios with their associated levels of significancefor each independent variable.^ An adjusted coefficientis an estimate of the effect of a category alone holdingconstant all other categories in the analysis. TheF-ratio is used to determine whether an independentvariable explains a significant portion of the varianceof the dependent variable holding constant all otherindependent variables. The results of the MCA analysiswith perceived power (Y,) as the dependent variable

' In the article by Peters [ 12], Beta ((3) was defined properlybut a typographical error improperly reported p as (3̂ .

^ Some categories of independent variables were combinedto meet the 40 observation minimum category size that hasbeen personally recommended by other users of the program[12]. Failure to meet the minimum size requirement canproduce results which are likely to be meaningless [2].

Table 2MCA ANALYSIS

PERCEIVED POWER AS A FUNCriON OF

COERCIVE SOURCES OF POWER"

Coercive sources of power

Num- Name andber response category

Y Grand meanX|^ Land

01

X2ff Fairness123456

X|, Right to sell1234 & 5

X|5 Building01

X|8 Legislation1 &2345

X|, Threaten01

Adjusted R^

Adjustedcoeffi-cients

2.30

- .32.46

- .52- .32- .44- .13

.20

.23

- .50- .09

.19

.85

- .09.13

.06

.00

.04- .05

.01- .06

= .344

.321

.224

.213

.090

.040

.020

F-ratio

127.81"

12.46"

18.81"

10.05"

.65

.50

"The order in which the independent variables are arranged isby descending Beta ((3) values.

>< .01.

and the coercive sources as theindependent variables are shown in a similar formatin Table 2.

Hypothesis 1 predicted that franchisors would relyprimarily on coercive sources rather than noncoercivesources to achieve power over franchisees. Table 1shows that noncoercive sources explain only 17.1%of the variance in the index of perceived power. Inaddition, three of the eight significant noncoercivesources (including the variable that does the greatest"explaining") have adjusted coefficients that decreaseover the range of categories (one through five) whenall of them were hypothesized to be positively relatedto perceived power. According to the adjusted coeffi-cients, as the quality of local advertising assistance,supplies sold by franchisors, and equipment sold byfranchisors increases, perceived power decreases. Tosuggest that franchisees would more willingly yieldpower when their franchisors' assistance is of lowquality than when it is of high quality seems to becontrary to reason and suggests that the observed

POWER IN A CHANNEL OF DISTRIBUTION 191

relationships may be spurious.The coercive sources of power as shown in Table

2 explain 34.4% of the variance in perceived power.Allot the significant coercive sources (X|g, XJQ, X^g,Xij) have adjusted coefficients that increase fromthe lowest through the highest response category ashypothesized. This means that the perceived powerof the franchisor increases: when the land and/orbuilding is controlled by the franchisor; as the franchiseagreement is considered weighted in favor of thefranchisor; and as the franchisor's right to sell isrestricted.

When a third MCA analysis was run with perceivedpower as the dependent variable and both the non-coercive and coercive sources of power as independentvariables, R- was only increased to .420. The additionof noncoercive sources to coercive sources providesonly a small improvement in the percent of varianceexplained. Taken collectively, the above results suggestthat franchisors, as hypothesized, primarily employcoercive sources of power to achieve power over theirfranchisees.

Three MCA analyses with franchisee satisfaction(Yj) as the dependent variable were run in testingHj. Table 3 shows the results of the MCA analysiswith franchisee satisfaction (Yj) as the dependentvariable and the noncoercive sources of power(X|, Xj, ..., X14) as the independent variables. Theresults of the MCA analysis with franchisee satisfac-tion (Yj) as the dependent variable and the coercivesources of power (X,^, X,^, ..., XJQ) as the indepen-dent variables are shown in Table 4.

Hypothesis 2 predicted that franchisors relying lesson coercive sources and more on noncoercive sourceswould have franchisees who are more satisfied withthe franchise relationship. Table 3 shows that all sixof the noncoercive sources significant at the 1% level

Table 3MCA ANALYSIS

FRANCHISEE SATISFACTION AS A FUNCTION OF

NONCOERCIVE SOURCES OF POWER"

Noncoercive sources of power

Table 3 (continued)

Num- Name andber response category

Adjustedcoeffi-cients F-ratio

Grand meanEquipment

I &2345

Site locationsI &2345

National advertisingI & 2

.65

.17

.06

.06

.04

.12

.02

.07

.09

.01

.155 8.49"

.150 8.01"

.122 5.31*

Noncoercive sources of power

Num- Name andber response category

345

"On-the-job" training1 &2345

Local advertising1 &2345

Pricing assistanceI &2345

Deletions & additionsI &2345

Product preparation1 &2345

Field supervisors1 &2345

Formal training1 &2345

Supplies1 & 2 & 345

Operating manual1 &2345

BookkeepingI &2345

Day-to-day advice- 1 &2

345

Adjustedcoeffi-cients F-ratio

-.06.09.08

- . 1 1

.02

.02

.04

-.07.02.08.00

- . 0 6- . 0 1

.09

.05

-.07.03

-.05

.07

.00

.03

.05

.02

.04

.03

.05

.01

.03

.05

.01

.04

.00

.05

.05

.01

.04

.02

.02

.01

.05

.03

.01

.00

.01

.06

.112 4.42"

.109 4.23"

.097 3.36"̂

.091 2.93'

.080 2.24

.074 1.94

.069 1.72

.065 2.25

.065 1.49

.042 .63

.042 .63

Adjusted R- = .234

"The order in which the independent variables are arranged isby descending Beta (P) values.

" p < .01.'p< .05.

192 JOURNAL OF MARKETING RESEARCH, MAY 1974

Table 4MCA ANALYSIS

FRANCHISEE SATISFACTION AS A FUNCTION OF

COERCIVE SOURCES OF POWER"

Coercive sources of power

Num- Name andber response category

Y Grand meanX20 Fairness

123456

X|7 Threaten01

X|g Legislation1 &2345

X|, Right to sell1234 &5

X,5 Building01

X|^ Land01 •

Adjusted R^

Adjustedcoeffi-cients

2.30

.02

.02

.02

.01

.04-.25

.03-.19

.03-.02

.05-.05

-.09.00.01.10

-.01.02

-.01.02

= .207

P

.388

.150

.097

.069

.028

.028

F-ratio

30.96"

23.01"

3.24"=

1.62

.80

.80

"The order in which the independent variables are arranged isby descending Beta O) values.

*'p< .01.= p < .05.

have increasing adjusted coefficients as hypothesized.This means that as the quality of these franchisorassistances (noncoercive sources of power) increases,so does the level of franchisee satisfaction. For ex-ample, as the quality of the equipment sold by thefranchisor to the franchisee improves, the level offranchisee satisfaction increases. The noncoercivesources of power explain 23.4% of the variance infranchisee satisfaction.

As hypothesized, the two coercive sources of powersignificant at the 1% level are inversely related tofranchisee satisfaction. This means that franchiseesare less satisfied with the franchise relationship asagreements are perceived to be weighted more heavilyin favor of the franchisor, and when they have beenthreatened with a revocation of their franchise. Thecoercive sources of power explain 20.7% of thevariance in franchisee satisfaction.

The coercive and noncoercive sources of powertaken together explain 30.3% of the variance in

franchisee satisfaction. Although the noncoercivesources explain slightly more of the variance infranchisee satisfaction than the coercive sources, theproportion of explained variance is increased consid-erably by including both the noncoercive and coercivesources in the analysis. The above results suggest that,as hypothesized, franchisors can increase franchisees'level of satisfaction with the franchise relationship byrelying more on noncoercive sources and less oncoercive sources of power.

CONCLUSIONSThe results provide empirical evidence that a signifi-

cant relationship exists between the power of a channelmember and the sources of power available to him.In one kind of franchisor-franchisee channel of dis-tribution, franchisors were found to rely primarily oncoercive sources of power to achieve power over theirfranchisees. The coercive sources of power signifi-cantly related to perceived power were: control ofland, fairness of the contractual agreement, restrictionof the right to sell the franchise, and control of building.

The results also provide empirical evidence that theconsequences of exercising power in a channel ofdistribution depend on the sources of power exercised.Franchisors were found to be able to increase thefranchisees' satisfaction with the franchise relation-ship by relying more on noncoercive sources of power,such as providing higher quality assistance in the areasof equipment, site location, national advertising, "on-the-job" training, local advertising, pricing assistance,and product deletions and additions. The results alsoshowed that franchisors could increase franchiseesatisfaction by relying less on the coercive sourcesof power.

The results of this research suggest several addition-al research questions on the role of power in channelsof distribution: (1) Is the power of channel membersother than franchisors a function of the sources ofpower available to them? (2) What sources of powerdo members of other channels of distribution relyupon? (3) Do the consequences of exercising powerdepend on the sources of power that are exercisedin channels of distribution other than the franchisor-franchisee channel?

APPENDIX

RATING SCALESThis section presents the rating scales not discussed

in the body of the article.

Power (Y,)

The six points of the scale were: "I have almostcomplete responsibility for this," "I have much moreresponsibility for this than my franchisor does," "Ihave a little more responsibility for this than myfranchisor does," "My franchisor has a little more

POWER IN A CHANNEL OF DISTRIBUTION 193

responsibility for this than t do," "My franchisorhas much more responsibility for this than t do,""My franchisor has almost complete responsibilityfor this."

Noncoercive Sources (X,,X2,...,X,J

The five points of the scale were: "Poor Quality,""Below Average Quality," "Average Quality,""Above Average Quality," and "Excellent Quality."

Need for Legislation (X,g)

The five points of the scale were: "Very Unneces-sary," "Unnecessary," "Neither Necessary Nor Un-necessary," "Necessary," and "Very Necessary."

Right to Sell Franchise (X,^)

The five points of the scale were: "I may sell toanyone without my franchisor's approval," "I maysell to anyone who is approved by my franchisor,""I may sell to any approved person, but my franchisorhas the 'right of first refusal' to buy back the fran-chise," "I may sell only to my franchisor," and "Ido not have the right to sell my franchise."

Fairness of Agreement (X^g)

The six points of the scale were: "Weighed verymuch in favor of the franchisee," "Weighed mod-erately in favor of the franchisee," "Weighed slightlyin favor of the franchisee," "Weighed slightly in favorof the franchisor," "Weighed moderately in favorof the franchisor," and "Weighed very much in favorof the franchisor."

REFERENCES

1. Alderson, Wroe. "Survival and Adjustment in OrganizedBehavior Systems," in Reavis Cox and Wroe Alderson,eds.. Theory in Marketing. Homewood, III.: RichardD. Irwin, 1950, 65-87.

2. Andrews, Frank M., James N. Morgan, and John A.Sonquist. Multiple Classification Analysis. Ann Arbor:Institute for Social Research, The University of Michi-gan, 1967.

3. Beier, Frederick J. and Louis W. Stern. "Power inthe Channel of Distribution," in Louis W. Stern, ed..Distribution Channels: Behavioral Dimensions. Boston:Houghton Mifflin, 1969, 92-116.

4. Dahl, Robert A. "The Concept of Power," BehavioralScience, 2 (July 1957), 201-15.

5. El-Ansary, Adel I. and Louis W. Stern. "Power Mea-surement in the Distribution Channel," Journal of Mar-keting Research, 9 (February 1972), 47-52.

6. French, John R. P. and Bertram Raven. "The Basesof Social Power," in Darwin Cartwright, ed.. Studiesin Social Power. Ann Arbor: University of MichiganPress, 1959, 150-67.

7. Heskett, James L., Louis W. Stern, and Frederick J.Beier. "Bases and Uses of Power in InterorganizationRelations," in Louis P. Bucklin, ed.. Vertical MarketingSystems. Glenview, 111.: Scott, Foresman, 1970, 75-93.

8. Little, Robert W. "The Marketing Channel: Who ShouldLead This Extra-Corporate Organization?" Journal ofMarketing, 34 (January 1970), 31-8.

9. March, James G. "An Introduction to the Theory andMeasurement of Influence," American Political ScienceReview, 49 (June 1955), 433-51.

10. Ozanne, Urban B. and Shelby D. Hunt. The EconomicEffects of Franchising. Washington, D.C: U.S. Gov-ernment Printing Office, 1971.

11. Palamountain, Joseph C. The Politics of Distribution.Cambridge, Mass.: Harvard University Press, 1955.

12. Peters, William H. "Using MCA to Segment New CarMarkets," Journal of Marketing Research, 1 (August1970), 360-3.

13. Raven, Bertram H. "Social Influence and Power," inIvan D. Steiner and Martin Fishbein, eds.. CurrentStudies in Social Psychology. New York: Holt, Rinehartand Winston, 1965, 371-82.

14. Ridgeway, Valentine. "Administration of Manufac-turer-Dealer System," Administrative Science Quarter-ly, 1 (March 1957), 464-83.

15. Siegel, Harvey and Elaine Siegel, et al., v. ChickenDelight, Inc., et al., and Vance E. Shephard and BenZachary,etal., Intervening Plaintiffs, CCH 73,146 (D.C.V. Cal., April 1970); BNA ATRR No. 458, (April 21,1970), A-1, X-1.

16. Simon, Herbert. "Notes on the Observation and Mea-surement of Power," Journal of Politics, 15 (November1953), 500-18.

17. Stern, Louis W. "The Concept of Channel Control,"Journal of Retailing, 43 (Summer 1967), 14-20.

18. and J. L. Heskett. "Conflict Management inInterorganization Relations: A Conceptual Frame-work," in Louis W. Stern, ed.. Distribution Channels:Behavioral Dimensions. Boston: Houghton Mifflin,1969,92-116.

19. Wilemon, David L. "Power Negotiation Strategies inMarketing Channels,'' The Southern Journal of Business,7 (February 1972), 71-2.