Are Nonprofits Unfair Competitors for Businesses? An Analytical Approach
Author(s): Yong Liu and Charles B. Weinberg
Source: Journal of Public Policy & Marketing, Vol. 23, No. 1, Conceptualizing, Assessing, and
Managing Risk: Public Policy Perspectives (Spring, 2004), pp. 65-79
Published by: American Marketing Association
Stable URL: http://www.jstor.org/stable/30000473 .
Accessed: 07/06/2011 13:55
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at .
http://www.jstor.org/action/showPublisher?publisherCode=ama. .
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
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].
American Marketing Association is collaborating with JSTOR to digitize, preserve and extend access to
Journal of Public Policy & Marketing.
http://www.jstor.org
Are
An
Nonprofits Unfair
Competitorsfor
Analytical Approach
Businesses?
YongLiuand CharlesB. Weinberg
Thisstudy examinesduopolyprice competitionbetweena for-profitfirmand a nonprofitorganization.
It shows that the competitiveoutcome is predominantlythe consequence of theirdifferentobjective
functions. Thedamage to the for-profitcaused by the nonprofit'spolicy and regulatoryadvantagesis
only marginal.Moreover,the for-profitcan protect itself by acquiringStackelbergprice leadership.
The
competitionbetweenfor-profitfirmsandnonprofit
organizations has produced a fair amount of controversy and has attracted an increasing amount of
research (e.g., Rose-Ackerman1990; Schiff and Weisbrod
1993). A particulardebate focuses on unfaircompetition, a
claim that various policy and regulatory advantages that
nonprofits receive are the driving force that places forprofits in unfavorablemarketpositions. For example, nonprofits in the United States typically do not pay corporate
income tax and, in general, are exempt from state and local
propertytaxes as well as some sales taxes. Othernonprofit
advantagesinclude receiving lower postal rates. Small forprofit businesses and the industry groups that represent
them, especially the U.S. Small Business Administration,
have complainedin public forums and to regulatorybodies
(Weisbrod 1988, p. 107; see U.S. Small Business Administration 1983). They claim that the "decisive" advantages
enjoyed by the nonprofits enable them to charge lower
prices than private firms, many of which thus lose their
competitive edge and find it difficult to survive (e.g., Bennett and DiLorenzo 1989, p. 2).1
In response,the nonprofitsector points out thatthe unfair
competitioncriticismis mainly based on anecdotalinformation; its validity has not been examined by rigorous analyses. The sector also suggests that several benefits available
to the business sector (e.g., small business set-asides, loan
guarantees,managementassistance, tax credits and deductions, depreciationallowances, access to capital and business expertise) may be significantenough to renderthe criticism moot (Pires 1985).
The purposeof this article is to examine the significance
of regulatory advantagesthat nonprofits receive in affecting competitive outcomes in a marketin which for-profits
1As a business owner commented, "Unfaircompetition occurs when a
nonprofitorganization'stax-exempt status allows it to undercutcompetitors who pay taxes" (Murray2003).
YongLiu is Assistant Professorof Marketing,MartinJ. Whitman
School of Management,SyracuseUniversity(e-mail:[email protected]).
CharlesB. Weinbergholds the SMEVPresidents Professorshipin
Marketing,SauderSchool of Business, Universityof BritishColumbia (e-mail:[email protected]).
Vol.23 (1) Spring2004, 65-79
and nonprofitscoexist. We also explore situationsin which
for-profits can possibly avoid (or reduce) competitive
disadvantages.
Industry
Background
In the United States, the numberof nonprofitorganizations
has more than tripled from approximately310,000 in 1969
to nearly 1,000,000 today (Weisbrod 1998, p. 69). As a
result of this quick growth, competitionbetween for-profit
firms and nonprofitorganizationshas rapidlyintensified.
On the one hand, for-profitsare active in most industries
in which nonprofits historically have operated, such as
health care, child day care, family counseling, researchand
development, and arts and education (e.g., Pires 1985). For
example, Salamon and Anheier (1998) reportthat nonprofits and for-profitsaccount for 51% and 17%of U.S. hospitals, respectively.Of nursinghomes, 20% are nonprofitsand
75% are commercial.The artsand entertainmentindustryis
also marked by an extensive mixture of organizations.In
Boston, for example, some of the well-known performing
arts organizations have nonprofit status (e.g., American
RepertoryTheatre),and others are for-profit(e.g., Colonial
Theatre). Private, for-profit businesses are entering the
university-degree market. For example, University of
Phoenix reports that more than 75,000 students have
enrolled, "manyof whom are earningaccreditedbachelor's
degrees in fields such as business, nursing,and education,as
well as MBA degrees"(Farrington1999, p. 81).
On the otherhand, nonprofitscompete in several markets
that are typically regarded as commercial. The nonprofit
Mountain Equipment Co-op competes aggressively in the
outdoorclothing and equipmentretail marketin Canada.In
the audiovisual education industry, nonprofits (e.g., Minnesota Educational Computing Corporation,which is the
most well known) compete with privatefirms by producing
and distributingtheirown audiovisualmaterialsor computer
software; producingproductsfor others; distributingproducts created by other organizations; renting equipment,
facilities, and videos; trainingand consulting;and repairing
equipment(Bennett and DiLorenzo 1989). Retail organizations regularlycomplain about (what they consider unfair)
competitionfrom nonprofits.Table 1, based on a surveythat
asked business respondentsto indicate whether they faced
competition from nonprofits, shows extensive competition
in the six industriesstudied.Dependingon the industrystudJournalof PublicPolicy&Marketing 65
66 Nonprofit
forBusinesses
Competition
Table 1.
PerceivedTax-ExemptCompetitionby For-Profitsin SelectedIndustries
Percentageof RespondentsClaiminga
CertainNumberof Competitors
Projected
Respondents
No
Competitor
One
Competitor
More
ThanOne
Competitor
Audiovisual
356
36
7
57
Racquetsports
462
10
19
71
Researchand testing
215
16
2
82
28
57
6
39
Travel agent
7349
55
6
39
Veterinarian
18,444
39
20
41
Industry
Tour
Top ThreeTypesof
Tax-ExemptCompetitors
University or college: public
Government:state or local
University or college: private
YMCA/YWCA or YMHA/YWHA
club
Recreation/health/sports/fitness
Hospital
University or college: public
Government:state or local
Research organizations
Religious organizations
University or college: public
Business or professional association
Religious organization
University or college: public
Social/fraternalorganization
Humane or animal welfare organization
Government:state or local
University or college: public
Source:BennettandDiLorenzo(1989,Table2.2). OriginaldataarefromU.S. GeneralAccountingOffice(1987).
= YoungMen's(Women's)Christian
= YoungMen's(Women's)HebrewAssociation.
Notes:YMCA/YWCA
YMHA/YWHA
Association;
ied, between 40 and 80% of respondentssaid thatthey faced
more than one nonprofitcompetitorin their local market.2
TheLiterature
Relatively few studies have examined the competition
involving nonprofitorganizations.In the limited literature,
Rose-Ackerman (1990, 1996) suggests that ideology and
ideological commitmentpreselect the people who are likely
to own/operatenonprofitorganizations.In turn,this attracts
customerswho value this commitmentand preferto rely on
the nonprofitform because it signals reliabilityand credibility. Some researchfocuses on whetherdifferentinstitutional
forms imply differentmarketbehavior.For example, Weisbrod (1998) finds that private for-profit firms, churchrelatednonprofits,and othernonprofitsdiffer in their use of
price and waiting lists as alternativedistributionalmechanisms. Tuckman (1998) uses Porter's (1980) five-force
model to identify the conditions in which the commercialization of nonprofits is likely to occur. Tuckman suggests
that increasedcompetitive intensity between nonprofitsand
for-profitstends to push nonprofitsto use more heavily forprofit business techniques. Similar points of view can be
found in the social marketing literature (e.g., Andreasen
1994, 2002), in which researchershave observed "a general
model of intersector transfer of marketing concepts and
tools from the commercial to the nonprofit sector"
(Andreasen 2001, p. 4). Despite the growth of knowledge
2Inindustriesin whichnonprofitsand for-profitscoexist,on average
nonprofitsarelargerin termsof bothnumberof employeesandrevenue.
Thisis inconsistent
withthesituationof theeconomyas a whole,in which
arelarger(Rose-Ackerman
on averagefor-profits
1996).
aboutnonprofitorganizationsand the controversysurrounding nonprofitversus for-profitcompetition, the equilibrium
outcome and market structureof industries in which nonprofits and for-profitscoexist remainlargely unstudied.3
Existing researchon marketsin which organizationswith
different objective functions compete primarilyfocuses on
marketsthat comprise private firms and public or government enterprises, and these studies are mostly concerned
with social welfare issues. The existing studies are quite different from ours, which focuses on competitive strategies.
Furthermore,most of these studies make use of Cournottype games; that is, competitionoccurs as rivals set quantity
decisions. A naturalchoice in constructingthese games is to
assume thatproductsare homogeneous (see, e.g., Beato and
Mas-Colell 1984; Cremer, Marchand, and Thisse 1989;
Tirole 1988). We arguethatthe productsthat differenttypes
of organizationsprovide are presumablydifferent.Thus, we
allow for a varying degree of product substitutabilityand
examine a Bertrandpricinggame. Notably, our allowance of
productsubstitutabilityled to several unexpected results. In
addition, the existing studies on mixed markets have
adopted different game rules. We assume that the public
firm is the Stackelbergleader (Bis 1986; Rees 1984) or the
follower (Beato and Mas-Colell 1984) or that it moves
andequilibrium
of marketsin which
aboutthe survivability
3Questions
organizationswith differentobjectivefunctionscoexist have also been
raisedby, forexample,Pires(1985),in thediscussionof for-profits'
entry
intotraditionally
nonprofitarenas,andMahajanandVenkatesh(2000),in
firmsin theonlinebusiness
the discussionof customershare-maximizing
andregulatory
domain.Frommanagerial
pointsof view,analysesof these
marketswill providevaluableinsighton competitivestrategiesandpublic
policy.
Journalof PublicPolicy& Marketing 67
simultaneouslywith the private firm (Cremeret al. 1989).
Such assignment of game rules is, by and large, arbitrary
(see, e.g., Cremer,Marchand,and Thisse 1989, p. 283). In
this article,we examine all threepossible situationsandprovide justifications for each of them. Finally, the issues that
we examine, such as unfair competition, are unique to the
nonprofitand for-profitmarkets.
Using a duopoly model of price competition, we show
that whereasthe for-profitis indeed worse off when it competes with a nonprofit, its loss in market share and profitabilityis almost entirelythe consequenceof differentorganizational types (which we model as different objective
functions). The policy and regulatory advantages that the
nonprofitreceives (which we model as a reductionin marginal cost) play an insignificantrole. AcquiringStackelberg
price leadership may help the for-profit receive positive
profits in a highly competitive market(which we model as
high productsubstitutability)that would be unprofitablefor
it in a simultaneousgame. A less competitive marketdenies
the for-profitthis potentialbenefit.
The remainderof the article is organizedas follows: We
set up the model to examine competitiveoutcomes in a market in which a nonprofitand a for-profitcoexist, assuming
that the rivals move simultaneously.Next, we explain and
examine Stackelbergsituations.We discuss other objective
functions of the nonprofit,marketentry and exit, and some
empiricalevidence in supportof the theoreticalfindings.We
conclude with a discussion of limitations and further
research. Throughoutthe article, we summarize the main
substantive findings as "Results" and the more technical
ones as "Findings."
inWhichNonprofits
andForMarkets
ProfitsCoexist
Several importantfeatures distinguish nonprofitsfrom forprofits. First, typical nonprofitstend to help disadvantaged
people, provide social services, and supportsocially beneficial programs.Such social profit necessarily leads to nonfinancial objectives (Gallagher and Weinberg 1991). Second, nonprofits face a nondistributionconstraintenforced
by law. They cannot use the revenue from either operations
or other sources to compensate board members, trustees,
and other "owners"or "incorporators"
beyond an economic
salary.A corollaryof this constraintis that a nonprofitmust
use all its resources for purposes compatible with its nonfinancial objectives.4Togetherwith the nonfinancialobjectives, the nondistributionconstraint forces nonprofits to
focus on the distributionof their products as long as their
financial resources can support the activities. Third, the
socially beneficial natureof nonprofitsenables them to seek
supportfrom donors and government.However, faced with
declining governmentsupportand unableto increaseprivate
givings substantially,nonprofitsare increasinglyturningto
commercialactivitiesby selling productsor services for revenue (Dart and Zimmerman2000; Dees 1998; Schiff and
Weisbrod 1993).
4Seniormanagersin nonprofitorganizationsinfrequentlyact in ways that
appearto maximize their own personalwealth. Discovery of such behavior
typically resultsin discreditingboth the agency and the managersinvolved,
as was the case for the United Way in 1992 (Barnes 1994).
In a marketin which a nonprofitand a for-profitcompete,
the competition is asymmetricalas a result of the different
natureof the competitors.In our model, each of the duopolists providesone product.The productsare substitutes,but
the degree of substitutionvaries. We do not intendto model
public goods, and we restrictthe discussion to privatelyconsumed goods.
Competingon prices, both the for-profitandthe nonprofit
generaterevenueby selling to customers.The nonprofitcan
receive donations,the amountsof which are often linked to
its performancein serving clients. Although other types of
financial resources (e.g., government subsidy) are often
determinedby regulationsand are not necessarily the consequence of marketbehavior, they can be viewed as ways
for the nonprofitto reducefixed or marginalcosts, which we
discuss in a subsequentsection.
We first set up the demandfunctions. We denote the forprofit with the subscriptf and the nonprofitwith the subscript n. Following Shubik and Levitan (1980) and Raju,
Sethuraman,and Dhar (1995), we specify marketdemandas
follows:5
1 qi = 2 [1 pi + 0(pj pi)]
(1)
where i, j = f, n representsthe two rivals; qi is the demand
for producti; andpi indicatesprice. Parameter0 capturesthe
degree of cross-price sensitivity or productsubstitutability;
a higher 0 implies more intensive competition(0 > 0).
The demandsystem specified in Equation1 leads to wellbehaved isoprofit curves. We plot a set of these curves in
Figure 1, which is helpful duringthe discussion of the reaction functions and the Stackelberggames. We assume that
competitors have the same fixed costs F. Their marginal
costs are cf and cn, respectively (0 ><cf, cn < 1).
5Shubik and Levitan (1980) initially proposed demand functions with
this structure(i.e., a term of own-price effect and anotherterm capturing
price difference effects), which can be derived from consumerutility maximization. We refer interestedreaders to Shubik and Levitan (p. 89) for
details.
Figure 1.
Isoprofit Curves and Price Reaction Functions of
Firm 1
C
A
.37
nt > 0
2
.35
Firm
of
.nt= 0
.33
Price
B
.31
.35
.4
.45
Price of Firm1
.5
68 Nonprofit
forBusinesses
Competition
Different
Functions
Objective
The for-profitnaturallymaximizes profits.We can write the
objective function as
max cf= (pf- c)qf- F.
(2)
Nonprofits are not profit maximizers. In this article, we
begin with the assumption that the nonprofit maximizes
quantity sold. This is the nonprofitobjective function supported most frequently by existing studies. For example,
Steinberg (1986) examines the revealed objectives of nonprofits in five industries.His results show that public welfare, education,and artsnonprofitsare quantitymaximizers.
In a study of Red Cross blood-service units, Jacobs and
Wilder (1984) find evidence for the objective function in
which output is maximized subject to a break-even constraint.Gapinski (1984) shows that the Royal Shakespeare
Company, a nonprofit performing-arts organization in
Britain, produces more output and sets lower prices than
does a profit maximizer.Supportfor quantitymaximization
can also be found in, for example, the works of RoseAckerman (1987) and Weinberg (1980). Although we
believe that quantity maximization captures a significant
amount of marketreality, we examine other possible nonprofit objective functions in the "Discussion" section and
show that the main results derived from quantitymaximization are robust to a wide range of nonprofit objective
functions.6
As we discussed previously, nonprofits must consider
restrictions imposed by the availability of financial
resources. Their competitive behavior should reflect this
consideration.Therefore,a nondeficit budget constrainthas
been used extensively in nonprofitstudies (e.g., Netz 1999;
Rose-Ackerman1987; Weinberg 1980) and indeed is supportedby empiricalresearch(e.g., Jacobsand Wilder 1984).
Donations are an importantfinancial resource for many
nonprofit organizations. When modeling the donation
effect, we acknowledge that many privatedonors base their
donation decisions on the nonprofit's mission and how
effective the nonprofitis in achieving its goals (e.g., RoseAckerman 1996; Weinberg 1980). In the context of maximizing quantity,we can formulatedonationas an increasing
function in quantity served to the market.The most parsimonious way to do so is to use a linear response function:
(3)
Dn(qn) = tqn, t > 0,
where t is the response rate of donation.7
6In one of the few studies of its kind, Voss, Cable, and Voss (2000, p.
336) analyze the organizationalvalues of nonprofitprofessional theaters
and find, in brief, five value dimensions: (1) prosocial (i.e., "providing
wider access to and appreciationof the arts"),(2) artistic(i.e., artisticcreativity, innovation,and independence),(3) financial(i.e., financial stability
and security), (4) market (i.e., overall customer satisfaction), and (5)
achievement (i.e., publicly recognized excellence). The different orientations reflect the tension between the marketand the artistic mission of an
arts organizationand, in turn, the congruency of their activities with the
pursuitof audience maximization.Voss, Cable, and Voss do not indicate
the relative prevalenceof these views. The analyses we presentin this article, particularlythe generalizedframeworkin the "Discussion"section, are
consistent with most of the five value dimensions and would capturethe
values espoused by many nonprofittheaters.
7To the extent that donations are independentof qn, the nonprofitcan
always use these exogenous resources to offset the expense items in the
budgetconstraint.Thus, theirultimateeffect is to reducefixed costs for the
Thus, we can write the nonprofit's pricing problem as
follows:
1
max qn = - [1 - Pn + O(pf pn],
2
(4)
subjectto (pn - c)qn + tqn z F.
Market
Without
Equilibrium
Regulatory
Advantages
Recall that concern about unfaircompetitionfocuses on the
regulatoryadvantagesthat the nonprofitreceives, such as a
lower postal rate. The primaryconsequence of these advantages is a lower operationalcost (i.e., the marginal cost in
our model). To examine how decisive the advantagesare in
pushing the for-profit into unfavorable market situations,
and thus to address the validity of the unfair competition
criticism, we examine the benchmarksituationby assuming
that the regulatoryadvantagesdo not exist. Thus, we begin
with an analysis in which the nonprofit and the for-profit
have a common marginal cost c. We then compare the
results with the situationin which the nonprofitreceives the
advantages.We assume away donations for the moment to
enable us to focus on the main issue.
PriceReaction
Functions
andEquilibrium
We solve for the price reaction functionspi*(pj). Obtaining
pf (Pn) is straightforward:
1 + (1 + 0)c + Opn
pf(pn) (=)
2(1 + 0)
(5)
We solve the constrained maximization problem for the
nonprofit,which indicatesthat the optimal price is achieved
when the budget constraint is binding.8 We find that the
nonprofit'sreactionfunction is9
P*n(Pf)= {[1 + Opf+ (1 + 0)c]
(6)
-
[1 + Opf + (1 + O)c]2 - 4(1 + 0)(c + Ocp + 2F)}/
2(1 + 0).
Figure 1 illustrates the reaction functions (Equations 5
and 6). If Firm 1 is a for-profit,its reactioncurve will be line
BC, which representsprice responses that lead to optimal
profit levels. For a nonprofit,only the isoprofit curve representing zero profit is relevantbecause of the binding budget
constraint.Thus, if Firm 1 is a nonprofit,the reactioncurve
is AB, wherePoint B is the lowest point on the zero isoprofit
curve. For each price that Firm 2 charges, the nonprofit
nonprofit. We recognize that certain marketing expenses are necessary
when donations are sought. Because such expenses work in the opposite
direction of donation revenue and because revenue is usually greaterthan
expenses, we focus on the revenue side.
8This is consistent with the results in several existing studies in which,
because profitabilityis not the objective of a nonprofit, the authorshave
assumed that the budget constraintis binding so that the nonprofitalways
breakseven and the nondeficitconstraintbecomes (Pn- c)qn= F (e.g., Netz
1999; Rose-Ackerman1987).
9There are two solutions for the maximization problem (Equation 4).
One is shown in Equation6, andthe otheris the dualwith a plus sign before
the squareroot. Because the nonprofitwants greaterquantity,Equation6 is
optimal.
Journalof PublicPolicy& Marketing 69
potentiallycan set two price levels to remain at breakeven.
Because the left branchof the zero isoprofitcurve represents
the lower price and thus greaterquantity,the nonprofitwill
respondalong line AB. Although it is not illustratedin Figure 1, both the nonprofitand the for-profitreaction curves
show the desired property (i.e., become steeper as 0
increases),which indicates a more competitive market.
As in typical Bertrandgames, the for-profit's reaction
function follows the "strategic complement" pattern
(Bulow, Geanakoplos, and Klemperer 1985; Tirole 1988,
pp. 207-208): One party's increasing (decreasing) price
induces the rival to raise (lower) its price. However, the
nonprofit's reaction curve is downward sloped. Such a
strategicsubstitutepatternhas been mainly associated with
quantityresponses in the literatureand is quite unusualfor
price competition.lo It indicates that if the for-profit
increases its price, the nonprofitwill lower its price to gain
more customers. However, if the for-profit decreases its
price, the nonprofit'sbest reactionis to raise its own price to
remainat breakeven.Consistentwith the nonprofit'spricing
problem,this implies that the nonprofittargetslow price to
the possible limit. In the extreme, the nonprofit would
merely charge c if fixed costs equal zero. Nevertheless, the
nonprofitcannotprice too low because the budgetconstraint
needs to be met in the face of these costs.
Finding1:Unlikethe reactionfunctionof for-profits,that of
nonprofitsin a pricinggameis of the strategicsubstitutetype;thatis, it is downwardsloping.
If a price equilibriumexists, it will occur at the intersection between p*f(Pn) and (pf). We solve Equations5 and
pnequilibriumpricesp*fand p*n:
6 simultaneouslyto obtain
(7) p*f= 502
+ 100 + 4 + 15c02 + 14c0 + 4c + 4c03 - e-K
4(02 + 40 + 2)(1 + 0)
and
(8)
p*n =
30 + 2 + 2c2 + 5c + 2c - JK
2(02 + 40 + 2)
where K is a complicatedclosed-form function of 0, c, and
F. The Appendix provides the details of K and other complex functions.
We can show equilibriumquantitiesthatcorrespondto pZf
and p*nto be positive. Therefore,the price equilibriumsfall
into the appropriateregion describedby the demandsystem
of Equation 1. Note that different from for-profitcompetitions, the nonprofit'sfixed costs is an importantfactor that
influences equilibriumprices.
The equilibriumsolutions bear several desirable properties. First, both prices decrease as competition becomes
stronger. Second, the local monopoly case arises as 0
approacheszero. As 0 approachesinfinity, the marketfeatures perfect competition,and both equilibriumprices equal
marginalcost. Finally, as long as an equilibriumexists, the
nonprofitcharges a lower price than does the for-profit.The
10Thispairof incompatiblereactioncurves may have strongimplications
for tacit collusion involving nonprofits. Cooperationis accommodatedif
the rivals respondsimilarlyto each other's actions. Thus, collusive behavior seems to be unlikely in a market in which for-profits and nonprofits
coexist (for an example of price fixing by the Ivy League schools, see Netz
1999).
duopoly market ceases to exist as the nonprofitbegins to
charge a higher price than the for-profit (the proof is provided in the section "Comparison of Nash Equilibrium
Prices in the For-Profit Versus Duopoly Market"in the
Appendix).
A price equilibriumexists when neitherthe nonprofitnor
the for-profit earns negative profits.11In the section "The
Properties of Price Equilibriums" in the Appendix, we
examine in detail the propertiesof the equilibriumprices.
We summarizethe key findings as follows:
Althoughthe nonprofitis able to remainat breakeven,the
for-profit will receive positive profit only when F is less
than FI:
(9) F1 =
402 + 40 + 1 - 2c - 8c02 + 4c202 + 4c20 + C2
2(2102 + 903 + 160 + 4)
Thus, the duopoly marketexists in equilibriumonly when
fixed cost is less thanF1.
The degree of competitive intensity (0) also influences
the price equilibrium.If the competingproductsaretoo similar to each other (i.e., 0 is high), the equilibriummay not
exist, because either the for-profit does not earn positive
profit or the nonprofitcannot break even. As we detail in
"ThePropertiesof Price Equilibriums"in the Appendix,this
upperlimit of 0 equals 01.12The duopoly marketobtains in
equilibriumonly when 0 < 01.
As F and c increase, p*nincreases at a much faster rate
than does p*. This happens because the nonprofit'sbudget
constraintmakes it more sensitive to changes in cost factors.
The price differencebetween the nonprofitandthe for-profit
thus decreases as costs become higher. This patternis illustratedin Figure 2, Panel A.
Figure 2, Panel B, indicates that both p*fand p*ndecrease
as competition intensifies. However, p*ndecreases at a
slower rate than p* does, making the price difference continuously decrease. Thus, contrary to the case of cost
changes, the nonprofitis less sensitive to changes of competitive intensity.The explanationof this drawson the result
that a nonprofit tends to reduce its price to the degree of
breakeven. Unless the market is highly competitive (i.e.,
cross-price sensitivity of demand is high, which induces a
net loss for the for-profitand thereforeno equilibrium),the
changes in 0 do not shift much demand between the two
competitors.Therefore,the nonprofit'sprice can be reduced
by a small amountto remainat breakeven.The insensitivity
of the nonprofit to changes in competitive intensity when
two nonprofits compete is shown to a fuller extent in the
"Discussion"section.
Finding2: The nonprofit'sabilityto chargea lowerpricethan
thefor-profit
decreasesas costsincreaseand/orcompetitionintensifies.However,themarketequilibrium
breaksdownbeforethenonprofitis forcedto charge
a higherpricethanthefor-profit.
3:
Finding Comparedwith a for-profitrival, the nonprofit's
pricingbehavioris moresensitiveto changesin cost
factorsbut less sensitiveto changesin competitive
intensity.
11Otherwiseeither will stay out of the marketin a perfect equilibrium.
12The approximatelinear expression of 01 is 01 = 14.25 - 19.18c 151.99F, which indicates that the higher the costs, the lower is the maximum competitive intensitythat allows for a price equilibriumto exist.
70
for Businesses
NonprofitCompetition
Figure 2.
The Impact of Changes in Costs and Competitive
Intensity
A: Cost Changes
.4
Pf
.35
.3-
.25
.2
Pn
Pn
.15
F1
.15
0
.02
.06
.04f
.08
B: CompetitionIntensityChanges
.54
.52
.5
.48
.46
pf*
.42
.4
.38
.36
.32
.3
.28
.26
.24
81
0
.2 .4 .6 .8
1 1.2 1.4
e0
TheFor-Profit's
Loss
To addressthe issue of unfaircompetition,we examine the
extent to which the for-profitis worse off as a result of the
existence of a nonprofitcompetitor.We make two comparisons for this purpose.First, what is the impact on the focal
for-profitwhen it faces a nonprofitratherthan a for-profit
competitor in duopoly competition? Second, what is the
impact of a nonprofit entrant on a for-profit monopolist?
Note that at this point we assume that there are equal costs
for the duopoly rivals. Any gain or loss to the competitorsis
thus entirely the consequence of different objective functions, which have nothing to do with the advantages conveyed by public policy.
The for-profitduopoly result is well known. At equilibrium,each for-profitwould charge [1 + (1 + 0)c]/(2 + 0) and
would receive a profit of (1 + 0)(1 - c)2/[2(2 + 0)2] - F. If
the for-profitcompetes with a nonprofit,at equilibriumthe
for-profit would charge pf, as in Equation 7, and would
receive a profit of R2(c, F, 0), which has a complicated
closed-form expression.
The competitionwith a nonprofit(ratherthan anotherforprofit) has two effects on the focal for-profit.First, it makes
an otherwiseprofitablemarketunprofitable.Specifically, if
the competitoris anotherfor-profit,the focal for-profitearns
positive profit as long as F < (1 + 0)(1 - c)2/[2(2 + 0)2].
However, competitionwith a nonprofitis not profitablefor
a for-profitwhen F < F1, and it holds that F1 < (1 + 0)(1 c)2/[2 + (2 + 0)2]. Second, the for-profit's profit is much
lower when the competitor is nonprofit rather than forprofit.At parametervalues c = 1, 0 = .7, and F = .06, the forprofit's profit is .034 in a for-profitduopoly and .013 when
it competes with a nonprofit.This indicates a 62% decrease.
Otherparametervalues producedsimilar results. Therefore,
the for-profit is much worse off when facing a nonprofit
than another for-profit, even if the nonprofit does not
receive any policy advantages.
Now consider the case of a for-profit monopolist that
faces a linear demand curve qn = 1 - pn.13 The optimal
= (1 + c)/2, sell qn = (1 - c)/2, and
strategyis to chargepmn
receive a profit of rTn= (1 - c)2/4 - F. Two effects similar
to the case of a for-profitduopoly arise. First, the nonprofit
entrantcan make a marketthat is profitablefor a for-profit
monopolist unprofitable,because the for-profit earns positive monopoly profit as long as the fixed cost is less than
(1 - c)2/4, which is greaterthan F1. Second, the for-profit's
profit drops (quite drastically)when a nonprofitcompetitor
enters the market.At the same parametervalues c = 1, 0 =
.7, and F = .06, it holds that m = .143 and nr = .013,
which indicates a 91% decrease. Table 2 (specifically the
column "No Cost Advantage")presents similar results for
more parametervalues. We summarizethe resultas follows:
Result1:A for-profitcanbe muchworseoff whenthereexists
a nonprofitrival,even in the absenceof any policy
to thenonprofit.
advantages
TheRoleof Regulatory
Advantages
We now turnto the case in which regulatoryadvantagesare
available to the nonprofit.As we discussed previously, we
can conveniently model such advantages as a lower marginal cost.14Assuming that the nonprofit's marginalcost is
cn and the for-profit'smarginalcost remainsat c, we resolve
the price equilibrium:
(10)
pf = [502 + 100 + 4 + 11c02 + 12c0 + 4c + 3c03
+ Cn03 + 4cn02 + 2cnO - 80-i]/[4(02
+ 40 + 2)(1 + 0)],
and
(11) Pn =
30 + cn02 + 4cn0 + c02 + 2 + 2cn + cO - M
2(02 + 40 + 2)
The closed-form expression of M is presented in the
Appendix.
Compared with the situation of equal cost (c) for both
rivals, a new (and lower) marginalcost cn for the nonprofit
induces both of them to charge a lower price. Furthermore,
13Thisis the monopoly demandcurve that is consistent with the duopoly
demandfunctions in Equation 1.
14Thesituationin which these advantagesreduce fixed costs for the nonprofit is straightforward(no additional solving is needed) and generates
outcomes that resemble those of reduced marginal costs. Moreover, tax
exemptions have an effect similarto that of reduced costs.
Journalof PublicPolicy& Marketing 71
Table2.
The For-Profit'sProfitLoss WhenCompetingwith a Nonprofit
Profitof the FocalFor-ProfitWhenIt
Competeswith a NonprofitThat Has
BenchmarkCase
0
Profitof the Focal
For-Profitin the
BenchmarkCase
The focal for-profit
competes with another
for-profit
.5
.7
.9
.037
.034
.031
.019 (49%)
.013 (62%)
.009 (71%)
.016 (57%)
.010 (71%)
.006 (80%)
The focal for-profitis
a monopolist
.5
.7
.9
.143
.143
.143
.019 (87%)
.013 (91%)
.009 (94%)
.016 (88%)
.010 (93%)
.006 (96%)
No CostAdvantage
CostAdvantage(r)
arethepercentages
of decreasein profitfromthebenchvaluesusedforillustration
arec = .1, F = .06, andr = .8. Numbersin parentheses
Notes:Parameter
markcase(e.g.,49%= [.037- .019]/.037).
the nonprofitreduces its price to a greaterextent than does
the for-profit.As a result,the nonprofitcapturesnot only all
the new demandthatthe lower prices generatebut also some
demand that is served by the for-profit in the equal-cost
case. To quantify the impact of this cost difference on the
for-profit,we furtherexamine cn = rc. That is, the nonprofit
receives policy advantagesthat amountto a (1 - r) percentage reductionin marginalcost.
The result of r = .80 is summarizedin the "Cost Advantage (r)" column in Table 2. For example, in a moderately
competitive market(0 = .7) the focal for-profitwill earn a
profit of .010 when it competes with a nonprofit that
receives a 20% cost reduction(leading to cn = .80c). This is
71% less than what it would obtain if the competitor is
anotherfor-profit.Although this is a dramaticloss for the
for-profit,note that 62% is actually due to the difference in
objective functions and only 9% is due to costs, as we have
discussed. Comparedwith the situationof being a monopolist, the for-profitloses 93% in profits when there is a nonprofit rival. However, again, 91% of this drop is due to the
presence of a nonprofitcompetitor,and only 2% is due to
cost differences.
The message of this analysis is clear:It is the natureof the
competitorsthat mattersthe most, not the policy or regulatory advantagesthat the nonprofitreceives. The for-profitis
put into unfavorable market positions because its profitmaximizingbehavioris inherentlyvulnerableto the competition from a service-maximizing nonprofit. At most, the
regulatoryadvantagesplay a secondaryrole that is far from
decisive. If a for-profit suffers profit loss (and becomes
unable to survive) in the competition with a nonprofit, it
would make no substantialdifference even if the nonprofit
did not receive any favorabletreatments.
Result2: The regulatoryadvantagesreceivedby the nonprofit
areneitherdecisivenornecessaryto inducelossesfor
the for-profitandto enablethe nonprofitto chargea
lowerprice.Thedominantfactorthatdrivesthecompetitiveoutcomeis thenatureof thecompetitors,
especiallytheirobjectivefunctions.
PriceLeadership
Stackelberg
We have establishedthatthe for-profitoccupies an unfavorable market position when it competes with a nonprofit,
even if it has no regulatoryadvantage.Is there any strategy
that the for-profitcan adopt to reduce or avoid such unfavorable situations?In this section, we change the assumption that the rivals move simultaneouslyto the assumption
that either the for-profitor the nonprofitmay have Stackelberg price leadership.We show that the for-profitmay benefit from acquiringStackelbergprice leadershipbut a nonprofit cannot.
Stackelberg leadership assumes that a competitor can
credibly announceits price earlierand with foresight of the
rival's reaction.In the context of ourmodel, a for-profitmay
have the leadershipas a result of its better use of competitive strategiesthat are common in the business world (RoseAckerman 1990; Tuckman 1998), whereas a nonprofitmay
have this competitive advantageas a result of being more
influentialin termsof size and revenue (see Note 2). A particular example of nonprofit Stackelbergleadership is the
marketin which nonprofitshistoricallyhave been the product providerand for-profitshave been late entrants.In such
a market,the nonprofitmight be better informedof market
conditions and thus can compete as a price leader. For
example, such a scenario could unfold as colleges and universities increasingly contend with online alternatives,
many of which are for-profit businesses (e.g., Farrington
1999).
A Stackelbergleader has the chance to increase its profit
to a level higher than that obtainedin a simultaneousgame.
It does so by finding the position on the follower's reaction
curve that leads to the highest profit. Note that the leader's
price-reaction function of the simultaneous game is no
longer relevantin solving the Stackelberggame. Instead,the
Stackelberg leader examines the intersection of the follower's reaction curve with its own isoprofit curves (and
seeks the highest profit) (for a detaileddescriptionof Stackelberg games and their solutions, see Tirole 1998). In the
traditional context of for-profit competition, Stackelberg
leadership enables the price leader to raise price and
enhance its profits,regardlessof the intensityof the competition. The price follower will also set a higher price. Ultimately, only the consumers are worse off (Tirole 1988, p.
331).15
15Wecanshowthatthesetypicalresultsholdforthedemandsystemof
However,manyof them
Equation1 if the duopolistsarebothfor-profits.
ceaseto holdforthemodelof a marketin whicha nonprofit
anda for-profit
coexist.
72 Nonprofit
forBusinesses
Competition
However, the Stackelberg outcome in the market in
which a nonprofit and a for-profitcompete is quite different from the traditionalsituation in which two for-profits
compete. We provide the technical details of the Stackelberg solutions in the Appendix ("StackelbergLeadership
When a For-Profitand a Nonprofit Compete"),but we discuss the key findings and their intuition in the text. For the
ease of discussion, we refer to the price equilibriumin the
Stackelberggame as "Stackelbergequilibrium"and that in
the simultaneousgame as "Nash equilibrium."
Result3: The nonprofitStackelberg
is the sameas
equilibrium
the Nashequilibrium,
regardlessof the level of competitiveintensity.
TheFor-Profit
as the Stackelberg
Leader
When the for-profitacts as the price leader, it is impossible
to obtainclosed-form solutionsfor the game. As we show in
the Appendix, the best strategyof a for-profitacting as the
Stackelberg leader depends on the level of competitive
intensity (0).
When 0 is comparativelyhigh (i.e., 0 > 01), the Stackelberg equilibriumis superiorto the Nash equilibriumfor the
for-profit;it may use the Stackelbergleadershipposition to
improve its profits. We now summarize the explanation,
which is detailed in the Appendix.
Because of the strategicsubstitutenatureof the nonprofit
price-reactionfunction,the for-profitlowers its price to take
advantage of the Stackelberg leadership. This induces the
nonprofitto increaseits price to remainat breakeven.When
the competitive intensity is high (0 > 01), the price changes
result in enough shift of demand such that the for-profit
earns greaterprofits than it would in a simultaneousgame.
However, when the competitive intensity is low (i.e., 0 <
01), the for-profitdoes not anticipatesufficient demandshift
by charginga lower price. As a result, it cannotearn greater
profits than in the simultaneousgame. Thus, when 0 < 01,
Stackelbergleadershipdoes not benefit the for-profit.
For illustration purposes, Table 3 provides numerical
examples for these two patternsof the for-profitStackelberg
results. For the lower 0 in each pair of costs (e.g., 0 = 1), the
for-profit'sprofit is lower in the potential Stackelbergequilibriumthan it is in the Nash equilibrium.For the higher 0
(e.g., 0 = 5), the for-profit'sprofit may increase from a negative level in the simultaneousgame, which does not have a
duopoly Nash equilibrium,to a positive level in the Stackelberg game.
Recall that 01 separates the profitable and unprofitable
regions for the for-profitin a simultaneousgame. We found
that across all feasible values of 0, the profit level changes
signs at 01 between the simultaneousgame andthe for-profit
Stackelberggame. Specifically, if the competitive intensity
is low (i.e., 0 < 01) such that the for-profit earns positive
profits in the simultaneousgame (and thus a Nash equilib-
TheNonprofit
as the Stackelberg
Leader
When the nonprofit competes as the Stackelberg price
leader, the game is solved by substitutingthe for-profit's
reaction function into the nonprofit's objective function
(and the budget constraint).We found two prices that the
nonprofitmay adopt as follows:
(12)
5cO+ 2c02 + 2 + 2c + 30 N
2(02 + 40 + 2)
and
(13)
5cO+ 2c02 + 2 + 2c + 30 + 1k
NS (
+ 40 +
2)
2(02
where the superscriptNS indicates nonprofit Stackelberg.
The for-profit's responses to pNS and pNS
are derived from
P (pn). It can be shown that the nonprofit's quantity is
greater when it charges ps, which is thus the nonprofit
Stackelbergprice.
Note thatpnNSis
identicalto pT,the nonprofit'sNash equilibrium price. In other words, being able to announce its
price earlier and foresee the for-profit's reaction does not
enable the nonprofitto charge a lower price than that in the
Nash equilibrium.The explanationis as follows: A profitseeking Stackelbergleader usually moves to a differentisoprofit curve where profits are greater. However, the nonprofit's behavior is significantly different; because of the
(binding) budget constraint,it always remains on the same
isoprofitcurve where profitis zero. This constraint,which is
embedded with the nature of nonprofit organizations,
removes any advantages that Stackelberg leadership may
produce.
Table 3.
The For-Profit as a Stackelberg Price Leader and Its Profits
Marginal Cost (c)
Fixed Cost (F)
.4
.1
.01
.04
.01
.02
1
5
1
5
1
5
1
.330
.182
.345
.210
.557
.461
.566
Pn*
.118
.117
.181
.185
.429
.427
.462
Profit*
.043
.010
.020
-.003
.015
.001
.007
-.001
pfB
pnB
ProfitB
.000
.300
-.075
.059
.158
-.040
.000
.300
-.105
.197
.215
.003
.200
.500
-.120
.419
.458
-.003
.366
.541
-.034
.476
.482
.001
Competition (0)
pf*
5
.479
.469
Notes: pf* and pn* are the Nash equilibrium;pfBand pnBare the potentialfor-profitStackelbergequilibriumthat correspondsto Point B in Figure Al, Panel
B. Note that if the profit is negative, a perfect duopoly equilibriumdoes not exist.
Journalof PublicPolicy& Marketing 73
rium exists), the potential Stackelberg position makes the
for-profitincur a net loss (thus, the Stackelbergequilibrium
is the same as the Nash equilibrium).In contrast, a more
competitive market(i.e., 0 > 01) that is unprofitablein the
simultaneousgame can be profitableif the for-profitobtains
Stackelbergleadership.
Result4: ObtainingStackelbergprice leadershipin a more
competitivemarket(0 > 01) helpsthe for-profitsurvive in the competition
witha nonprofit.Withoutthis
leadership,the for-profitwill earn negativeprofit;
does not exist.In
thus,the duopolyNashequilibrium
a less competitive
market(0 < 01),a Nashequilibrium
is identicalto
exists,andthe Stackelberg
equilibrium
the Nashequilibrium.
Previous research has shown that in the competition
between for-profits, consumers are worse off if the firms
compete in a Stackelberggame ratherthan a simultaneous
game. This does not hold when the for-profitfaces a nonprofitrival. First,if both the Stackelbergequilibriumand the
Nash equilibriumexist, they arethe same. Second, if the forprofit is able to take advantageof the Stackelbergleadership
in a highly competitive market,only the Stackelbergequilibriumexists. Therefore,there is nothing to say aboutconsumersbeing betteror worse off between the two solutions.
A few studies on a mixed marketof private and public
enterprises also have examined the Stackelberg behavior.
Although the findings are mixed, welfare is typically
improved over the Nash equilibriumwhen the public firm
acts as Stackelbergleader (e.g., Vickers and Yarrow 1988).
Our Stackelbergresults are different; maybe more important,allowing for productsubstitutabilityenables us to identify the differentbehaviorof a for-profitStackelbergleader
in marketswith differentcompetitive intensity.
Discussion
To concentrateon the main issue of unfaircompetition,the
analysis so far has avoided an explicit analysis of the effect
of donations.As we discussed in the previous section, donations provide additionalrevenue for the nonprofit.When a
linear donationresponse function is used (Equation3), the
nonprofit'sbudget constraintbecomes [Pn- (c - t)]qn2 F.
The consequence of donations is thus a reduced marginal
cost for the nonprofit.The results regardingthe nonprofit's
advantagein marginalcost then apply. Specifically, both the
nonprofitand the for-profitneed to dropprices, and the forprofit loses demandto the nonprofit.
More generally, whenever 3Dn/aqn> 0, the net effect of
donationsis to bring down the marginalcost. However, the
effective marginal cost may no longer be constant. How
quickly and to what extent it decreases depend on the shape
of the donationresponse curve.16
In the remainderof this section, we provide analysis of
and discussion about the robustness of the results to nonprofit objective functions,marketentry and exit, a comparison of equilibriumbetween a market served by two non-
16Inan empirical study of the Royal ShakespeareCompany, Gapinski
(1984) shows that grants, mainly from the Arts Council of Great Britain,
led to lower prices and a largeraudience for the performances.
profits and a market served by two for-profits, and some
empiricalsupportfor the results.
of Resultsto OtherNonprofit
Robustness
Functions
Objective
Quantity maximization is the nonprofit objective function
thathas been widely adoptedand supportedin the literature.
It describes the behavior of many nonprofitorganizations.
Nevertheless, the results we derived in the previous section
are not restrictedto quantitymaximization.
On the basis of empirical evidence and following Steinberg (1986) and Lowry (1997), we specify a more general
objective function that maximizes the weighted average of
total budget and quantity. At a minimum, a large budget
may bring higher managerialsalaries and prestige for the
nonprofit.
maxU = 8(F + cq) + (1 - 8)q,
(14)
where 0 d 8 d 1. Dependingon the managers'preferencefor
these two components,andin some cases as a political compromise among membersof the boardof directors,different
nonprofits may operate under different values of 8. For
example, the tension in arts organizationsbetween mission
and marketcan be conceptualizedas a debate over the value
of 8. Voss, Cable, and Voss's (2000) distinctionamong five
organizationalvalue dimensions can be modeled as a low
value of 8 for a market-orientedorganizationand a high
value of 8 for an achievement-orientedone.
In Equation 14, the marginal utility of quantity equals
8c + (1 - 8), which is always positive because 0 d 8 d 1.
Thus, the family of objective functions with budget and
quantitymaximizationhas the same equilibriumsolutionsas
quantitymaximization.
Indeed, any objective function that takes the general format of
(15)
max U{pf, Pn,v(pf, pn)),
where v(.) is any arbitrarycomponentof utility, can be analyzed in essentially the same way as we do in this article.
The results based on quantity maximization still hold as
long as the partialderivatives satisfy (aU/~pn)+ (WU/av)x
0. A particularexample of this type of objective
(Ov/aPn)<
function is the maximizationof consumersurplus.
There are situationsin which the nonprofitwants to maintain a certain amount of revenue for purposes such as
improving product quality, expanding into new service
areas,or preparingfor unforeseenevents. The pricing problem can then be modified by adding a positive amount of
profits to the budget constraint.The effect is the same as
when the nonprofithas higher fixed costs (see Note 14).17
17Ifthe amountof (positive) profits that the nonprofitreceives becomes
large enough, the for-profit and the nonprofit may end up charging the
same equilibriumprices. These prices are the same as those in the typical
for-profit competition situations and obtains when the nonprofitoperates
on the particularisoprofit curve that intersectsp (pn) with its left branch
exactly at the point at which the typical for-profitequilibriumoccurs. As
the amountof nonprofitprofits decrease, both the nonprofit's and the forprofit's equilibriumprices drop, and the nonprofit's price becomes less
than that of the for-profit.In the one-periodgame, however, it is straightforwardto show that the budget constraintis binding.
74 Nonprofit
forBusinesses
Competition
Therefore, the results derived from quantity maximization appearto be robustto a wide range of nonprofitobjective functions. Other possible objective functions include
maximizationof prestige, employee income, total revenue,
or a specified quantity/qualitytrade-off.Except for the support of quantitymaximization,little empirical guidance on
nonprofit objectives exists. It is beyond the scope of this
article to discuss all possible alternativesin detail.
FixedCostsandMarketEntry
It is well known that fixed cost plays an importantrole in
determining market entry decisions. Holding the opportunity cost of entryat zero, the for-profitdoes not want to participatein marketcompetitionif the fixed cost is higherthan
F1. However, when the fixed cost is higherthanF1but lower
than F2, the nonprofitis able to remain at breakevenif the
duopoly marketexists. This suggests that F1 < F < F2 is a
region in which the nonprofitcan survive the duopoly competition but the for-profit cannot, which is essentially a
reserved marketfor the nonprofit.This finding contradicts
the perception that the goal of profit maximization necessarily leads to greaterprofit for the for-profitthan what a
quantitymaximizing nonprofitcan obtain. It is exactly this
difference in objective functions that leaves the for-profit
with a net loss, whereas the nonprofitis able to breakeven
for this range of F.
Finding4: Fixedcostsin therangeof (F1,F2)createa reserved
marketforthenonprofit.
Althoughthenonprofitcan
breakeven in a duopolymarket,the for-profitwill
lose moneyif it choosesto compete.
When F is less than F1, a Nash equilibriumobtains in the
duopoly market.If F exceeds F2, either the nonprofitor the
for-profit can be a monopolist, depending on which is the
first mover. The nonprofitmonopoly price equals { 1 + c 4[(1 + c)2 - 4(c + F)] }/2, which is lower than the for-profit
monopoly price of (1 + c)/2.
We can show that a thirdcritical value of F exists, which
we denote as F3. Above F3, the fixed cost becomes prohibitively high such that neitherthe for-profitnor the nonprofit
can serve the marketeven as a monopolist. As a result, the
market collapses. From the monopoly prices shown previously, we can derive F3 to be (1 - c)2/4.
Figure 3 illustrates the nonlinear relationship between
fixed costs and market structure.Note that in the case of
prohibitively high fixed costs for socially desirable products, the nonprofit is more likely to be the survivor for
Figure3.
MarketStructureRelatesto Fixed Costs
Nash equilibrium
exists
0
F1
No duopolyequilibrium
F2
F3
Asymmetrical Reserved First-mover Market
duopoly marketforthe monopoly collapses
market
nonprofit
F
exogenous reasons.The governmentor donors may become
involved to help the nonprofitovercome the entry barrier.
Games
Symmetrical
In this section, we extend the analyses of the for-profitversus nonprofitduopoly to a marketin which two nonprofits
compete. The primaryreason for this extension is that in
reality, nonprofitorganizationsnot only compete with forprofits but also are involved in competition with other nonprofits. Such competition occurredinitially as the nonprofits in the same marketcompeted for financialresourcessuch
as donation and grants. As more nonprofits came to exist,
they began to compete for customers as well. For example,
dependingon universities' marketpositioning, they actively
compete for academicallygifted students,minoritystudents,
studentathletes, and full-tuition students.18Thus, it is valuable to examine the competitive outcome when nonprofits
compete and compareit with the more establishedresult of
for-profitcompetition.
We use the superscriptsf and n to denote the for-profit
duopoly and the nonprofit duopoly in these symmetrical
games, respectively. Solving the profit and quantity maximization problems, we obtain pf = pf = pf [1 + (1 +
O)c]/(2+ 0), and pn = pn = pn = {(1 + c) - I[(1 + c)2 - 4(c +
2F)] }/2.
We note two propertiesof pn. First, the price equilibrium
depends on fixed cost F because of the budget constraint.
The goal of quantity maximization leads to marginal cost
pricing if F becomes zero. This is different from pf, which,
regardlessof F, is always greaterthanthe first-bestoutcome
of marginalcost pricing. Second, the intensity of competition (0) does not influence the nonprofitprice equilibriums.
We find this to be both interestingand unexpected because
0 does matterfor the nonprofitreaction function (Equation
6). Our explanation of this result draws on the symmetry
between the two nonprofits and the fact that both of them
target breakeven. The ultimate impact of a higher 0 is to
make the rivals more sensitive to each other's action. In the
context of traditionalfor-profitcompetition,this would lead
to lower prices and lower profits. However, the competing
nonprofitduopolists always remainat zero profit, no matter
how intensive the competitionis. Thus, 0 does not affect the
profit level of the nonprofits; it can only affect the price
equilibriums by shifting demand between the two rivals
according to Equation 1. Because the identical costs of the
duopolists imply that the rivals' prices are always equal at
equilibrium,the impactthroughthis path disappearsas well.
The equilibriumprice in the nonprofit-servedmarket is
lower than that in the for-profit-servedmarket. Thus, to
make socially beneficial products more accessible to consumers, nonprofitdominatedmarketsare preferred.Consistent with Rose-Ackerman(1996), socially mindedmanagers
would be attractedto such markets.
If we comparethe price equilibriumsof the asymmetrical
game (i.e., p*nand p*f)with those of the two symmetrical
for all
games (i.e., pn and pf), we find that pf> p*f> pn>
pn price
possible parametervalues. A for-profitchargesa lower
18It is importantto note thatthe universities(and nonprofitorganizations
in a broadercontext) emphasize how they are different from one another
when competingfor students.
Journalof PublicPolicy& Marketing 75
when it competes with a nonprofit than with a for-profit,
because the nonprofitcompetitorprices more aggressively.
However, a nonprofitcharges a higher price when it competes with anothernonprofitthanwith a for-profit.The intuition is that the nonprofithas more flexibility in setting its
price if the competitoris a profit-orientedfirm ratherthanan
equally low price-orientednonprofitorganization.
the nonprofitchargesa higher
Finding5: At Nashequilibrium,
priceif it competeswithanothernonprofitthanif it
competeswitha for-profit.
Because pf is the highest among all the equilibrium
prices, consumers are worse off if they are served by two
for-profitsthan by either two nonprofitsor by a mixture of
nonprofitand for-profit.In contrast,because the equilibrium
price in the two-nonprofitmarketfalls between those in the
for-profitversus nonprofit duopoly market (i.e., pf*> pn >
p*n),it is not transparentwhich of the two marketsprovides
greaterconsumer welfare (thus, the greatest among all the
threemarketstructures).If we measureconsumerwelfareby
net consumer surplus,which is defined as aggregatedconsumer utility and equals the area under the demand curve
but above the horizontalline of a given price level, we can
show that the for-profit'sprice in the for-profitversus nonprofit duopoly marketis too high, such that consumers are
still better served by a pair of nonprofitduopolists.19
Remarks
Concluding
On the basis of a duopoly model of price competition, we
examine the controversial issue of unfair competition
between for-profitsand nonprofits.As a result of different
objective functionsandthe nonprofit'sneed to satisfy a nondeficit budget constraint,the equilibriumresults are different from those found in traditionalfor-profitmarkets.Most
important,our key results contradictthe argumentfor unfair
competition.We show that, at least for the duopoly demand
system we investigate, the difference in objective functions
is responsiblefor a dominantmajorityof the disadvantagea
for-profitfaces when competing with a nonprofit.The regulatorybenefits to the nonprofitare far from decisively placing the for-profit in unfavorable market positions. As a
result, the unfair competition criticism may have exaggerated the effect of these benefits. We also find that obtaining
Stackelbergleadershipmay help the for-profitsurvive in the
competition.
A theoreticalresult that is directly testable is that whenever marketequilibriumobtains, the nonprofitcan and will
charge a lower price than the for-profit competitor (see
Finding 2). We now provide some survey findings to support this result. The data all come from maturemarketsthat
are likely to be in a stable state. First, Bennett and
DiLorenzo (1989, p. 145) examine the audiovisual education marketfor deaf people and find that nonprofitscharge
about20% less than commercialfirms do, for an averageof
$11.75 per minute comparedwith $14.51 per minute. Sec19Mathematically,the net consumer surplusprovided by organizationi
(when competing with organizationj) is
NCSPi=
(1+ Opj
)/(1+0) 1 - p + O(pj - p) dp,
pi
2
where pi and pj are the price levels.
ond, in 2000, we surveyed 17 day care centers in a major
city on the West Coast. The nonprofitscharged an average
of $2.69 per hour, and the for-profitschargedan averageof
$2.98 (note that we did not controlfor qualitydifferencesin
this survey). Third,Merliss and Lovelock (1980) reportthat
nonprofitperforming-artsorganizationsin Boston chargeda
lower price thandid for-profitorganizationsat both the high
end and the low end of the price range. Their sample
included organizations such as the American Repertory
Theatre,the ShubertTheatre,Boston Symphony Orchestra,
the Wilbur Theatre,and the Colonial Theatre.We repeated
the survey in September2003 and found that the average
price at the low end charged by nonprofitswas $19, compared with $31 by the for-profits;however, at the high end
there was only a minimal price difference, which averaged
approximately$70.20
Some researchers have attempted to explore whether
institutionalforms (e.g., nonprofitversus for-profit)necessarily imply different market behavior (e.g., Schlesinger
1998). Ourresultsindicatethat at least for pricingdecisions,
institutionsthat are differentin natureshould behave differently in the rational pursuit of their goals. By the same
token, if no substantialdifferenceis observedon nonprofits'
pricing practice, it is likely that either the managementis
inefficient or, accordingto Weisbrod(1988), the nonprofits
arejust "for-profitin disguise."
Our analysis provides supportfor the view that market
competitioncan be an effective mechanismfor drivingmanagerial efficiency for nonprofits (e.g., Pires 1995). As we
have shown, the nonprofitis sensitive to changes in costs.
Beyond certainupperboundaries,it cannotsurvivethe competition. Thus, to serve its goals better,the nonprofitneeds
to strive for lower costs by efficient production and
management.
The duopoly model we analyzeis staticin nature.The differences between for-profitsand nonprofitsmake it worthwhile to examine some dynamic aspects. Besides the issue
of long-term efficiency, our results imply that nonprofits
and for-profitsmay differ in innovative behaviorand in the
adoption of advancedtechnologies (e.g., the Internet).The
findings that nonprofits are more sensitive to changes in
fixed costs and that they do not have access to equity markets make it more difficult for nonprofitsto bear a significant startupcost in orderto be innovative. Moreover,compared with investors in equity markets,private donors and
government agencies appear to be less motivated to bear
risks of innovative behavior.21
Our main results, which we developed in detail for a
quantity-maximizingorganization,hold for a broadrange of
nonprofitobjective functions. Nevertheless, some nonprofits pursue other goals. For example, producer selfsatisfactionis an importantobjective for some arts organizations (Adizes 1975; Voss, Cable, and Voss 2000).
Empiricalstudies of both the prevalenceof differentobjective functions and their impact on competitive behavior are
20Thispoints to a possible behavioralfeature of the nonprofit:As it is
trying to maximizeusage, it may charge a higherprice to customerswhose
willingness to pay is greater.We leave this issue for furtherresearch.
21Gallagherand Weinberg (1991) suggest that public scrutiny of nonprofits and of government spending may be importantfactors underlying
such risk aversion.
76 Nonprofit
forBusinesses
Competition
needed. Our theoretical results imply several empirical
questions to be addressedby furtherresearch.For example,
it would be worthwhile to examine a particularindustry
(e.g., child day care) for the predictedpricing patternsthat
are relatedto variablessuch as organizationaltypes, market
structure,and donationlevels.
Moreover, we have focused on one type of nonbusiness
organizationonly: the nonprofit.Other organizations,such
as cooperatives,associations,governmentagencies (e.g., the
Tennessee Valley Authority), and crown corporations(in
England, Canada, and elsewhere), also pursue goals other
thanprofitmaximization(e.g., Blais and Dion 1990; Kwoka
2002). For example, several government-backedinsurance
organizations (e.g., Wisconsin State Life InsuranceFund,
Oregon Medical InsurancePool) play an importantrole of
risk mitigation in some of the insurancemarkets.The analytical approachdeveloped herein can be used to study their
competitivebehaviorand possibly lead to furtherstudies on
other marketsin which organizationswith different objective functions compete.22
In this article, we do not model productdecisions; thus,
productpositioningis exogenous. We recognize thatthe ideology of nonprofitsmay constitute an importantfactor driving them to provide products or services that profitorientedfirms do not provide (or underprovided).Although
price competition alone shows many interestingresults, it
would be beneficial to model more than one dimension of
competition.A possibility is to explore how nonprofitscompete with for-profits on both price and quality. As RoseAckerman(1996) speculates, they may end up serving different marketsegments.
Appendix
Closed-Form
Expressions
(Al)
K = 12c20+ 9c20- 8c - 18c02+ 902+ 4 + 120
+ 4c2 - 24c0 - 1603F- 80F02 - 96FO- 32F.
(A2)
M = 4c0 + C204 + 4
-
- 2c04Cn- 12cn02C
- 8cn + C204 - 32F + 4c2
4cn0c + c202
from Equation5. This translatesto
of NashEquilibrium
Pricesin the
Comparison
For-Profit
VersusNonprofit
Market
Duopoly
We used a backwarddeductionprocess to comparethe magnitudes of pn andp*. If we assume that pn< p*, we have
22Wehave assumedthat the nonprofitprovides its productby charginga
positive price. We thus exclude cases in which nonprofits provide free
products. In those cases, nonprofits may cross-subsidize their activities
with revenues from othersources, such as donations,governmentgrants,or
mission-relatedcommercialactivities (Schiff and Weisbrod 1993). Olson's
(1971) by-producttheory provides an appealingframeworkfor modeling
such situations.
Pn <
(A4)
1 + c + c0
2+0
In contrast,Equation6 indicatesthatanyvalid solutions
of p*and needto satisfy[1 + Op*+ (1 + 0)c] z [pnx 2(1 +
pn theexpressionunderthe square-root
0)], because
signneeds
to be nonnegative.Substituting
Equation5 for p*,we have
(A5) 2(l + 0)pn < 1 + 0
1 + (1 + 0)c + 0pn
2(1 + 0)
+ (l+0)c,
which results in
(A6)
pn <
2 + 3 + 50c + 302c + 2c = 1 + c + cO
2 +0
4(1+ 0)2 -02
Inequality A6 is exactly the same as inequality A4. By
definition, EquationA6 holds as long as there is an equilibrium. Thus, it must be the case that p*n< P*fat any equilibrium. Marketequilibriumbreaksdown before the nonprofit
begins to charge a higher price than does the for-profit.
TheProperties
of PriceEquilibriums
We examine the existence of price equilibriums and their
relative sensitivities to changes in costs and competitive
intensity. First, we simultaneously solve two inequalities
(K > 0 andpn < p) to ensurethatpn andp* are valid expressions. This results in a possible upper bound of F above
which no equilibriumexists:
(A7)
Fmax =
(1 - c)2(1 + 0)
2(2 + 0)2
Second, as long as the equilibriumobtains, the nonprofit
remains at breakeven. For the for-profit, we substitute the
equilibriumprices back into Equation 2 and solve ntf(F,c,
0) = 0. We then derive two critical values of F:
(A8)
F1 = (402 + 40 + 1 2c + 8c0 8c02 +
4c202
+ 4c20 + c2)/2(2102 + 903 + 160 + 4),
+ 902 + 10c02 - 10C03Cn- 28cn02 + 120
- 28cn0- 1603F- 80F02+ 20C202+ 8c203
+ 16C20+ 2c203 - 6cn03 + 6c03 - 96F0.
1 + (1 + 0)c + Op*
2(1 + 0)
(Pn)<
(A3)
and
F2 =
(A9)
(1 - c)2(1 + 0)
2(2 + 0)2
It holds that F1 < F2 (note that F2 = Fmax).The for-profitis
able to earn positive profit when F is less than Fi, and rtf
decreases as F increases. However, when F falls within (F1,
F2), 7tf becomes negative. Thus, a marketexists in equilibrium only when fixed cost is lower than Fi.
We can similarly obtain the effect of competitive intensity (0) on the prices. First, correspondingto EquationA7,
0 must be in the following range for a price equilibriumto
exist:
(A10)
6min =
max
-8F + (1 - c)2 -
Ji
4F
-8F + (1 - c)2 + Ji
4F
Journalof PublicPolicy& Marketing 77
where B = -8F + 16Fc - 8Fc2 + 1 - 4c + 6c2 - 4c3 + c4. We
can show that 0minis always negative and thus that the
effective rangeof 0 becomes (0, Omax).
Second, anothercritical value of 0, which we denote as 01, exists between 0 and
When 0 falls within (0, 01), the for-profitreceives posOmax.
itive profits at equilibrium.When 0 falls within (01, Omax),
the for-profitcannotcover fixed costs and thus loses money.
The duopoly equilibriumobtainsonly when 0 < 01. Because
analytical solutions are not possible, we examine the properties of 01 with numericalprocesses. We calculatethe exact
values of 01 at a dense grid of values of c and F. We then use
these values in a linear regressionwith 01 as the dependent
variable and c and F as independentvariables. The regression results in 01 = 14.25 - 19.18c - 151.99F, with a satisfying R2 of .70.
anda
Whena For-Profit
Leadership
Stackelberg
Compete
Nonprofit
In this section, we provide detailed analysis for the Stackelberg games. Figure Al illustrates the economic structure
when eitherthe nonprofitor the for-profittakes the Stackelberg price leadership.The isoprofitcurves, which are shown
in Figure 1 to be well behaved, are essential in derivingthe
Stackelbergsolutions.
Stackelberg
Nonprofit
Recall that the Nash equilibriumand the nonprofitStackelberg equilibriumare identical.FigureAl, Panel A, provides
the explanationfor this result. As the Stackelbergleader, a
competitorhas the opportunityto select any point along the
follower's reactioncurve, which containsthe Nash solution.
Although a profit-seeking price leader usually moves to a
different isoprofit curve where profits are greater,the nonprofit remainson a single isoprofitcurve, 7n = 0, as a result
of the (binding)budget constraint.Thus, insteadof having a
continuous choice among the points along p*(Pn),the non(shown as
profit only makes a discrete choice between pn2NS
Point A) and pn1NS
(which is the Nash solution). As we dislead to higher
andthe correspondingpf1NS
cuss in the text, pn1NS
quantityfor the nonprofit.
For-Profit
Stackelberg
Figure Al. Price Equilibriums When One Organization Acts
as Stackelberg Price Leader
A: NonprofitStackelberg
Pf
Pf*(Pn)
As we note in the text, closed-form solutions for the forprofit Stackelberggame cannot be found. We used Figure
Al, Panel B, to derive the for-profit Stackelberg results.
Recall from Figure 1 that pn(pf) is the price reactioncurve
of the nonprofit,and Pf(Pn) is that of the for-profit.If the
Nash equilibrium (i.e., pn and pf) exists, it obtains at the
intersectionbetween the two curves.
Figure Al, Panel B, shows two types of for-profit isoprofit curves, which represent all possible situations. The
first type is and no, and the second type is nft. Note that
the shape of these isoprofit curves directly relates to the
intensity of competition(i.e., the magnitudeof 0). When nf
is the isoprofit curve, the market is less competitive than
when 4tcand are the isoprofit curves. This is because for
a given amountof change in Pn,the for-profitmust make a
greaterchange in pf to maintainthe same profit level if
and
are the isoprofit curves rather than 7t'. In other
words, as competition intensifies, the for-profit needs to
change pf to a greaterextent to maintaina certainprofitlevel
in responseto a given change of Pn.As a result, I and r2 are
graphically"flatter"than is rf.
If the isoprofit curves follow the curves illustratedby
and t24,the for-profit Stackelbergleader can move to isoprofit curves that bring greaterprofit. The highest level it
can receive while maintaining Stackelberg equilibrium is
This obtainsat Point B, the endpointof pn(pf). Note that
the for-profitenhances its profit by lowering the price from
p). This is a resultof the unusualstrategicsubstitutespattern
that p (pf) follows and is in contrast to the established
results for typical for-profit competitions (in which the
Stackelbergprice is higher than the Nash price). The forprofit's price at Point B is the lowest it may charge; otherwise, the nonprofitcannot survive and the duopoly market
ceases to exist.
If the isoprofit curves are similarto 7tr,the Nash equilibrium is achieved at a tangentpoint on pn(pf). Specifically,
the point at which isoprofit curve r[ touches the reaction
curve Pn(pf) is the Nash equilibrium.Any other points on
pn(pf) would lead to lower profits for the for-profit.As a
result, the for-profitStackelbergequilibriumis the same as
4t
A
4f
Pf*
in= 0
4t
4n
Pr*
Pn
B: For-ProfitStackelberg
Prn(Pf)
pf
nf1
nf2
p*(pn)
nf
Pf*
B
pn*
pn
Notes: (pn, p*) are the price equilibriums in the simultaneous game.
Depending on the intensity of competition (0), for-profit Stackelberg may shift the equilibriumsto Point B for large 0 or keep them
at (pn,pF)for small 0. NonprofitStackelbergdoes not shift the equilibriumsfrom (pn,pf).
4
4t
42.
78 Nonprofit
forBusinesses
Competition
the Nashequilibrium.
The intuitionbehindthis resultis as
follows: In a less competitive market, little demand is
gained from competition by lowering price. In the simultaneous game, the nonprofit's intention to set price to the lowest extent induces both rivals to preempt most of the switchable demand. Thus, the for-profit's potential gain from
lowering the price is minimal; it actually does not compensate for the loss in profit margin. Even though the for-profit
has the opportunity to do so as the Stackelberg leader, it will
not receive greater profits than in the Nash equilibrium.
Recall that the competitive intensity is captured by the
magnitude of 0. It is important to note that there is no
absolute value of 0 that separates a less competitive market
from a more competitive market. A particular value of 0 can
imply a less competitive market if the costs (F and c) are low
or a more competitive market if the costs are high. The same
0 with lower costs does not force rivals (especially the nonprofit) to compete as aggressively as they would in the case
of high costs. The main reason is that lower costs imply a
potentially larger profit margin, and it is relatively easy for
the nonprofit to remain at breakeven (recall that we have
shown that 01 is monotonically decreasing in F and c).
References
Adizes, Ichak (1975), "The Cost of Being a Artist," California
ManagementReview, 27 (4), 80-84.
Andreasen,Alan R. (1994), "Social Marketing:Its Definition and
Domain," Journal of Public Policy & Marketing, 13 (Spring),
108-114.
-,
(2001), "IntersectorTransferof MarketingKnowledge,"in
Handbook of Marketingand Society, Paul N. Bloom and Gregory T. Gundlach,eds. ThousandOaks, CA: Sage Publications,
80-104.
-
(2002), "MarketingSocial Marketingin the Social Change
Marketplace," Journal of Public Policy & Marketing, 21
(Spring), 3-13.
Barnes, N. (1994), "Cause-Related Marketing Revisited: The
Effects of The United Way Scandal," American Business
Review, 12 (2), 95-99.
Beato, P. andA. Mas-Colell (1984), "TheMarginalCost Pricingas
a Regulation Mechanism in Mixed Markets,"in The Performance of Public Enterprises, Maurice Marchand, Pierre
Pestieau, and Henry Tulkens, eds. Amsterdam:North-Holland,
81-100.
Bennett, James T. and Thomas J. DiLorenzo (1989), Unfair Competition: The Profits of Nonprofits. Lanham, MD: Hamilton
Press.
Blais, Andrdand StdphanieDion (1990), "AreBureaucratsBudget
Maximizers?The Niskanen Model and Its Critics,"Polity, 22
(4), 655-74.
Bis, Dieter (1986), Public Enterprise Economics. Amsterdam:
North-Holland.
Bulow, Jeremy I., John D. Geanakoplos,and Paul D. Klemperer
(1985), "MultimarketOligopoly: StrategicSubstitutesand Complements,"Journal of Political Economy,93 (3), 488-511.
Cremer, Helmuth, Maurice Marchand, and Jacques-Franqois
Thisse (1989), "The Public Firm as an Instrumentfor Regulating an Oligopolistic Market,"OxfordEconomic Papers, 41 (2),
283-301.
Dart, Raymond and Brenda Zimmerman(2000), "After Government Cuts: Insights from Two Ontario 'EnterprisingNonprofits,"' in The Nonprofit Sector in Canada, K. Banting, ed.
Kingston, ON: Queen's University School of Policy Studies,
107-148.
Dees, J. Gregory(1998), "EnterprisingNonprofits,"HarvardBusiness Review, 76 (1), 55-67.
Farrington,Gregory C. (1999), "The New Technologies and the
Future of Residential UndergraduateEducation," in Dancing
with the Devil: InformationTechnologyand the New Competition in Higher Education,RichardN. Katz et al., eds. San Francisco: Jossey-Bass, 73-94.
Gallagher,Katherineand CharlesWeinberg (1991), "Copingwith
Success: New Challenges for Nonprofit Marketing," Sloan
ManagementReview, 33 (Fall), 27-42.
Gapinski,JamesH. (1984), "TheEconomics of PerformingShakespeare,"AmericanEconomic Review, 74 (3), 458-66.
Jacobs, Philip and Ronald P. Wilder (1984), "PricingBehavior of
Nonprofit Agencies," Journal of Health Economics, 3 (1),
49-61.
Kwoka, John E. (2002), "GovernanceAlternatives and Pricing in
the U.S. Electric Power Industry,"Journal of Law, Economics,
and Organization,18 (1), 278-94.
Lowry, Robert C. (1997), "The Private Production of Public
Goods: OrganizationalMaintenance,Manager'sObjectives,and
Collective Goals," American Political Science Review, 91 (2),
308-323.
Mahajan, Vijay and R. Venkatesh (2000), "MarketingModeling
for E-Business," InternationalJournal Research in Marketing,
17 (2-3), 215-25.
Merliss, Penny Pittmann and Christopher H. Lovelock (1980),
"American RepertoryTheatre,"Case No. 9-580-133, Harvard
University.
Murray, Thomas D. (2003), "Non-Profits Have Unfair Advantage," The Ithaca Journal, (accessed March9, 2004), [available
at http://www.theithacajournal.com/news/stories/20030214/
opinion/983312.html].
Netz, JanetS. (1999), "Non-Profitsand Price-Fixing:The Case of
the Ivy League,"workingpaper,Departmentof Economics, Purdue University.
Olson, Mancur (1971), The Logic of Collective Action: Public
Goods and the Theory of Groups. Cambridge, MA: Harvard
University Press.
Pires, Sheila (1985), CompetitionBetween the Nonprofitand ForProfit Sectors:A Special Report.Washington,DC: The National
Assembly of National Voluntary Health and Social Welfare
Organizations.
Porter, Michael E. (1980), CompetitiveStrategy: Techniquesfor
Analyzing Industries and Competitors. New York: The Free
Press.
Raju, JagmohanS., Raj Sethuraman,and Sanjay K. Dhar (1995),
"The Introductionand Performanceof Store Brands,"Management Science, 41 (6), 957-78.
Rees, Ray (1984), Public Enterprise Economics, 2d ed. London:
Wenfeld and Nicholson.
Rose-Ackerman, Susan (1987), "Ideals Versus Dollars: Donors,
CharityManagers,and GovernmentGrants,"Journal of Political Economy,95 (4), 810-23.
(1990), "Competition Between Non-Profits and ForProfits: Entryand Growth,"Voluntas,1 (1), 13-25.
Journalof PublicPolicy& Marketing 79
-
(1996), "Altruism, Nonprofits, and Economic Theory,"
Journal of Economic Literature,34 (June),701-728.
Salamon, Lester M. and Helmut K. Anheier (1998), "The Third
Route: Government-NonprofitCollaborationin Germany and
the United States,"in Private Action and the Public Good, W.
Powell and E. Clemens, eds. New Haven, CT: Yale University
Press, 151-62.
Schiff, Jeraldand BurtonWeisbrod(1993), "CompetitionBetween
For-Profit and Nonprofit Organizationsin Commercial Markets," in The Nonprofit Sector in the Mixed Economy, Avner
Ben-Ner and Benedetto Gui, eds. Ann Arbor: University of
MichiganPress, 127-47.
Schlesinger, Mark (1998), "Mismeasuringthe Consequences of
Ownership: External Influences and the ComparativePerformance of Public, For-Profit, and Private Nonprofit Organizations,"in PrivateAction and the Public Good, W. Powell and E.
Clemens, eds. New Haven, CT: Yale University Press, 85-113.
Shubik, Martinand RichardE. Levitan (1980), Market Structure
and Behavior. Cambridge,MA: HarvardUniversity Press.
Steinberg,Richard(1986), "The Revealed Objective Functions of
Nonprofit Firms," RAND Journal of Economics, 17 (Winter),
508-526.
Tirole, Jean (1998), The Theoryof Industrial Organization.Cambridge:MassachusettsInstituteof Technology Press.
Tuckman, Howard P. (1998), "Competition,Commercialization,
and the Evolutionof NonprofitOrganizationalStructures,"in To
Profit or Not to Profit, Burton Weisbrod, ed. Cambridge,UK:
CambridgeUniversityPress, 25-46.
U.S. General Accounting Office (1987), Competition Between
Taxable Businesses and Tax-ExemptOrganizations.Washington, DC: U.S. GeneralAccountingOffice.
U.S. Small Business Administration(1983), Unfair Competition
by NonprofitOrganizationswith SmallBusiness:An Issuefor the
1980s. Washington,DC: U.S. Small Business Administration.
Vickers, John and George Yarrow(1988), Privatization:An EconomicAnalysis. Cambridge:MassachusettsInstituteof Technology Press.
Voss, Glenn B., Daniel M. Cable, andZannie GiraudVoss (2000),
"LinkingOrganizationalValues to Relationshipswith External
Constituents: A Study of Nonprofit Professional Theatres,"
OrganizationScience, 11 (3), 330-47.
Weinberg, Charles (1980), "MarketingMix Decisions for Nonprofit Organizations:An Analytical Approach,"Research in
Marketing,3, 261-69.
Weisbrod, Burton (1988), The Nonprofit Economy. Cambridge,
MA: HarvardUniversityPress.
(1998), "InstitutionalForm and OrganizationalBehavior,"
in Private Action and the Public Good, W. Powell and E.
Clemens, eds. New Haven, CT: Yale UniversityPress, 69-84.