Munich Personal RePEc Archive
Stakeholder Engagement in Land Development Decisions: A Waste of Effort?
Janmaat, Johannus A.
University of British Columbia Okanagan
6 December 2007
Online at https://mpra.ub.uni-muenchen.de/6147/
MPRA Paper No. 6147, posted 07 Dec 2007 00:25 UTC
Deisions: A Waste of Eort?
John Janmaat
∗
Eonomis, IK Barber Shoolof Arts and Sienes
University of British ColumbiaOkanagan
john.janmaatub.a
Deember 6,2007
Abstrat
Currently,managementdevolutionandengagementofloalstakehold-
ers - expeted to have better information - is seen as key to eetive
environmentalmanagement. Often,the abseneoflear property rights
and/orsupportingmarketinstitutionsleavesmanagement deisionstoa
politial proess. Whereundevelopedlandprovidesapubligood,when
to halt further development is modelled as a repeated lobbyingontest
betweenindustryand households. Lobbying eortaets the ontinua-
tionprobability. Dependingonhowstakeholdersareengaged,theremay
belittle impatonnaloutomes,or alobbyingwaranbe stimulated.
Overallwelfareisseldomenhaned.
Keywords: Stakeholderengagement,lobbyingontest,publigood,water
onservation,landdevelopment
JEL:Q24,Q25,R14,R52
1 Introdution
Powerto thepeople isamantrathat hasbeomeanimportantthemein the
environmentalmovement. Arangeofbooks andartiles,([33,36℄forexample)
arguethatloal environmentalstewardshipisthemeansbywhihourenviron-
mentalproblemsmustbeaddressed. Bothgovernmentalandnon-governmental
organizationsareoftenstronglypromotingsuhapproahes,whileritialanal-
ysis of these approahesare sare, and are so far at best inonlusiveabout
∗
FundingforthisresearhwasprovidedbytheCanadianSoialSienesandHumanities
Researh Counil. I am grateful to omments provided at the 2007 CanadianEonomis
Assoiation meetings and the 2007 European Assoiation of Environmental and Resoure
Eonomismeetings. Remainingerrorsandoversightsaremine.
island useplanning. Often,planninginvolvesdeidingwhetherornotaparel
ofopenspaeisgoingtobedeveloped. Developmentandiretlyonsumeare-
sourethatproduesaloalpubligood,suhasopenspae,ormoreindiretly,
suhasleadingtohangedowsinloalstreams. Inthispaperasituationloser
tothelatterismodelled. Aregulatorretainsnalauthorityoverwhetherornot
toallowfurtherdevelopment,whileapro-andanti-developmentlobbyattempt
toinuenewhendevelopmentwillbehalted.
Althoughonsumption ofopen spaebydevelopmentisprobablythe most
publiizedimpatofsprawl,itisnotobviousthat substitutionofalandsaped
suburbanyardforanagriulturaleldis destrutionofapubligood. Inon-
trast,wheredevelopmentonsumeswaterthatwouldotherwisemaintainstream
ows, the destrution of the publi good is moreapparent. We fous on this
type of situation here. Some eorts have been made to value instream ows
(examplesinlude [9, 12, 11, 28℄), oftenusing ontingentvaluation. A general
result of all this literature is that below a ertain level, redutions in stream
owreduethepubligoodvalueofthestream. Thistypeofpubligoodisthe
fousoftheurrentanalysis.
Anotherbranhofliteratureexaminesmehanismsthatanbeusedtoseure
instreamows([17,6,32,16,8℄amongothers),while[25℄doumentsafewases
whereinentivebasedapproaheshavebeenatuallyapplied. Game theoreti
approahesinthisliteraturehavelargelybeenrestritedtobargainingmodels,
whereatleastonedimensionofthebargaininginludesinstreamows[1,38,7℄.
Dynamisaregenerallyrestritedtoelementsofthephysial proesses. These
models are mostly built to inform poliy makers, rather than analyzing the
proessitself.
Although enabling orforing stakeholders to diretly engage eah other is
relativelynovel,stakeholdershavetraditionallybeeninvolvedinwaternegotia-
tionproesses.Theireetivenessatpromotingtheirinterestsisrelatedtotheir
eort, suintly ommentedon by [29, p343℄ ... only those with suiently
onentratedostsorbenets,whoattendhearingsandommitteemeetingsor
makelargeampaignontributionswillbeheard. AsLoomisargues,valuation
studiesareonewaytomeasuretheimpatsonthosenotativeintheproess.
However,manyarguethat engagementis preferableto valuationstudies. Par-
tiipation should therefore bemade less ostly, orfailure to partiipatemade
moreostly. `Stakeholderengagement','partiipatorymanagement,'andother
suhapproahesareperhapsbestinterpretedaseortstohangetheostsand
benetsofinvolvementinthedeisionproess. Theresultsoftheratherlimited
analysisofsuhapproahesismixed[3,26,42,23,30,39℄. Partiipantsexpress
agreaterappreiationfor others' situation,and suggesttheyare morewilling
to ooperate. However, thereis littleevidene of behaviourhange, and some
suggestionthatpeopleengageintheproesstodelayregulatoryhange,rather
than to partiipate in shaping that hange. Inthe model below, twoaspets
ofthe partiipationproessare onsidered,the eetiveness oflobbying eort
andtheresponsivenessoftheregulatortothateort. Greatereetivenessand
responsiveness are akin to inreasing the potential benet of partiipation to
Therelationshipbetweenhouseholdsanddevelopershasbeenexplored.Fis-
hel[13,14℄arguesthatloalpolitisisheavilyinuenedby'homevoters'who
partiipateoutoffearthatthevalueoftheirmostimportantasset,theirhome,
may derease. Hendersonand Beker[19℄reviews anumber ofmodelsof ity
development, and deides that the most appropriate has ities rst built by
developmentrmsand then turnedoverto agovernment. Managingdevelop-
mentto maximizeitizenwelfare onlyhappensafterthere areitizenspresent.
Lubellet al.[31℄ disussesseveral models, partiularly theontrast betweena
propertyrightsmodel-wheresaritydrivesademandforpropertyrightsover
that whih is sare - againstinterest groupmodels - the'Growth Mahine' -
and the politial market - where interests try to buy the regulator. Suess
inahievingonservationobjetivesdependsonboththepowerof theinterest
groupsand the institutional form. As evidene, onservation is generally less
where developersare themost powerful,while it is greatestwhere population
pressuresarehigher.
Muh researh has looked at the role of lobbying in government deision-
making. Inwhatfollowswefousonapartiularstream ofthisliterature, the
rentseekingontestmodelstartedbyTullok[40,41℄. Tullokisreditedwith
being the rstto oneptualize lobbyingas investing to aet the probability
of apturing aprize. In thestati ase with Tullok's speiation, theNash
equilibriumalwaysinvolveswastefulspendingonlobbying. Linster[27℄showed
thatwhentheontestisinnitelyrepeated,thisneednotbetrue. Dijkstra[10℄
found that interest groups may prefer regulation over a nanial instrument
if lobbying over use of revenues exhausts potential payouts. Graihen et al.
[15℄ showed that it may be optimal for rms to improvetheir environmental
behaviourifit reduesthelikelihoodthat anenvironmentallobbyanbeome
asubstitute. Ironially, environmental and onsumerlobbies may be working
in opposite diretions in suh situations. This paper implements a Tullok
stylelobbyinggameinadynamilanddevelopmentontext,wheredevelopment
onsumesapubligood,akintoinstreamow.
2 Model
Weonsideramodelwhereommunitymembers-households-andrmsboth
anpartiipatein anegotiationorlobbyingproess. Partiipationis ostlyto
both households and rms. Although the terms 'lobbying' and 'negotiation'
have almost ontraditory normative onnotations, for our purposes the key
issue is that engagement is ostly, but engagement does inuene outomes.
Thetermswillthereforebeusedsomewhatinterhangeably. Thenegotiationis
overwhetherornottoallowmoreofasareresoureessentialforommunity
growth - land or instream water - to be developed. Undeveloped, the sare
resoure produes a publi good. When used, this developed land provides
aommodationfor membersof the ommunity. Thus, development inreases
thesizeoftheommunity,andtherebythemarginalpubligoodvalueoffurther
Tobespei,let
w
bethetotalamountoftheresourethatanbedevel-oped,
z t
theamountleftundevelopedatt
,andq t
(q it
foreahrm)theamountthatisdevelopedin
t
. Thepubligoodandpopulationarebothnormalizedtothesameunits as
z t
,sothatatthebeginningofaperiod, populationisw − z t
.Developmentoursatthebeginningoftheperiod,afterwhihthepubligood
is enjoyed. Thus, thepopulation in period
t
, afterq t
units of land are devel-oped,is
w − (z t − q t )
. This isthepopulationableto enjoythe publigood inthatperiod, andwillbelabelled
m t
whenthisisonvenient. Attheendof theperiod, rmshoose an amount
x it
to devote to the negotiation proess, and householdshooseanamounty jt
. Theregulator'sdeisionproessisprobabilis- ti,asinaTullokgame,withtheprobabilityofahaltinfurtherdevelopmentgivenby
π(x t , y t )
,withx t
andy t
beingthetotalnegotiationeortforrmsand households. It is assumed thatπ 1 < 0
andπ 2 > 0
, where subsripts indiatepartialderivativearguments. Onedevelopmentoftheresourestops,itnever
resumes.
Thereare
n
identialrms,withperiodprotgivenbyp t (m t )q it − c t (q it ) − x it
(1)where
p t ()
isthepriethataunitoftheresoureommandsandc t (q t )
istheosttothermofdeveloping
q t
units. Thedemandurvesatisesp ′ t < 0
andp ′′ t ≥ 0
,whiletheostfuntion satises
c ′ t > 0
andc ′′ t > 0
. Theresoureisdurable, inthateah onsumeronlyonsumesoneunit,whih providesindenite servie.
Demandthereforedependsontotaldevelopment(population,
m t = w−(z t −q t )
),not period development (immigration,
q t
). This is onsistent with land, orannual rights to a nite renewable resoure like water. For eah household,
periodutilityisgivenby
u(z t , Y i − y it )
(2)where
Y i
isanexogenouslygiveninomeforhouseholdi
. Standardassumptions aresatised, withu 1 > 0
,u 11 < 0
,u 2 > 0
andu 22 < 0
. Note that this is theutilityofaresidentwhoownsahouse,withthepriepaidforthehousetreatedas
asunkost. Forsimpliity,thedemandforhousesisassumedtobeindependent
oftheleveloftheloalpubligoodsuppliedbyundevelopedland. Houseprie
and the utility of residents is therefore independent - we are onerned with
utility of residents after purhasing a house. A visual representation of the
gameisshownin gure1.
Determiningtheoptimaldevelopmentpathandstoppingpointrequiresa-
ounting for rm prots and household utility. If prots of the development
industryleavetheommunity,sothataountingforthemonsistsofignoring
them,thentheoptimalommunitysizeandpubligoodlevelarefound,forany
initial
z t
,bysolving[β/(1 − β )] max
q t
{u(z t − q t , Y i )m t }
(3)the present value of an annuity generating
u(z t − q t , Y i )m t
aggregate utilityindenitely, with disount fator
β
. The rst order ondition isu 1 = u/m t
,F
H
F
N
0 ≤ x t ≤ R t − C t
x t
0 ≤ q t ≤ z t
q t
y t
0 ≤ y t ≤ Y i
1 − π t
π t
" T
X
t=0
β t ( R t − C t − x t ) ,
T
X
t=0
β t U t + β T +1 1 − β U T
#
Figure 1: Game Tree. Firm payo when the game ends is the present value
ofprotsfrom housesales,lesslobbyingost, tothe pointwheredevelopment
ends. Forhouseholds, payo is the present valueof the utility stream earned
whiledevelopmentours,andthepresentvalueofthepermanentlevelofutility
earnedafterdevelopmentstops.
ormarginal utility equals averageutility. Thus, not surprisingly, theoptimal
ommunitysize is that whih maximizes average utility. The dynami nature
oftheproblem isirrelevant, asthe driverforan optimalpathovertime is the
ost funtion. This is borne by the rms, and therefore ignored in this ase.
Theutilitymaximizing solutionis tosimply set theommunityat itsoptimal
sizeimmediately. Theonlydependeneontheurrentlevelofdevelopment,
z t
,isthat if
z t
is smallerthantheoptimal undevelopedresourelevel, nofurther developmentours.The polar opposite is to onsider only rm prots and ignore household
utility. We are then foused only on the rm, whih seeks to maximize the
presentvalueofprotsfromdevelopment. Intheformofareursiverelationship,
thevalueofarmis
V t (z t ) = max
q {p t (w − (z t − q t ))q it − c t (q it ) + (1 − π)βV t +1 (z t − q t )}
(4)where
π
isanexogenousprobabilitythatfurtherdevelopmentwillnotbepermit- tedafter theurrentperiod. Inperiodt
,theequilibrium (assumingsymmetri rms)onditionis(q t /n)p ′ t + p t − c ′ t = (1 − π)βV t ′ +1
(5)Firms equate the marginal benet of urrent development to the disounted
expeted marginal benet of delay. Inreasing the numberof rms generates
moredevelopmentin
t
(reallthatp ′ t < 0
).utilityandprots. Ifweletrmsbeownedbyhouseholds,and,in thespiritof
theHartwikrule[18℄,protsanbeinvestedtogenerateapayoinperpetuity
(nodepreiation),athirdoptimization anbeperformed. FollowingBellman's
prinipleofoptimality[4℄,andassumingthatallprotsaredistributed equally
amonghouseholds, the reursiveequation that haraterizesthe optimalpath
is
U (k t , z t ) = max
q t
{u(z t − q t , Y i + (1 − β)k t /m t )m t + βU (k t +1 , z t +1 )}
(6)where
z t+1 = z t − q t
,andk t+1 = p t (m t )q t − c t (q t ) + k t
. Thenewstatevariable,k t
, is the aumulated prot from development ativities, whih generates a return at(1 − β)
that is sharedequally among the populationm t
. The rstorderonditionis
u t − u 1,t m t − (1 − β)(k t /m t )u 2,t = β(p ′ t q t + p t − c ′ t )U 1,t+1 − βU 2,t+1
(7)where with two subsripts, the rst indiates argument of derivative and the
seond the time period at whih the arguments are evaluated. The marginal
benetofinreasingdevelopmentinperiod
t
istheutilityearnedbytheadditiontothepopulation,lessthepubligoodlossandapital dilutionimpats. This
issetequaltothepresentvalueofthemarginalbenetofdelay,whihinludes
theimpatsof thehigherprieand protnextperiodand thegreater levelof
thepubligood.
Introduinglobbyingornegotiationdemands that householdsandrmsbe
onsidered simultaneously. Further, if either type of agent is to lobby, there
must be a benet to doing so. Here, this is the impat on the probability
development stops after the urrent period,
π(x t , y t )
, wherex t = P
x it
andy t = P
y jt
arethetotallobbyingeortsof thehouseholdsandrms. Thetworeursiveequationsthatharaterizethissystemare
V (z t ) = max
q it ,x it
{p t (w − (z t − q t ))q it − c t (q it ) − x it
+(1 − π(x t , y t ))βV (z t − q t )}
(8)U (z t ) = max
y jt
{u(z t − q t , Y j − y jt )
+π(x t , y t )β 2 /(1 − β)u(z t − q t , Y j ) + (1 − π(x t , y t ))βU (z t − q t )
(9)Twokeypointsarehighlightedinthesereursiverelationships. First,ifdevelop-
mentstops,thenthermearnsnofurtherprots. Seond,forthehousehold,if
developmentstops,householdsearnthepresentvalueof
u(z t − q t , Y j )
,aonse-queneofthefatthatthepubligoodlevelremainsat
z t − q t
forever,andthatnofurtherspendingonnegotiationisrequired. Theannuityfatoris
β/(1 − β)
.Some basi results of this formulation an be explored by examining the
rstorderonditions. Therstorder onditionsfor
x it
andy jt
denethebestresponsefuntionsof thetwotypesofplayers. Theseare
−1 − π 1,t βV t+1 = 0
(10)−u 2 ,t + π 2 ,t β [β/(1 − β)u t − U t +1 ] = 0
(11)the marginallobbing ost,
1
. Forhouseholds, urrent marginalutility loss is equated to marginal probability saled net benet if development stops, thedierene between
u t = u(z t − q t , Y i )
andU (z t − q t )
. A lobbying warourswhentheresultant
x t
andy t
valuesarelargerthanthehangeinwelfarerelativetotheequilibriumthatwouldourwithoutlobbying.
For bothtypesof agents,twofores interatto inuene whether andhow
muh itisworthengagingin negotiations. Forrms,these arethesizeof
V t+1
and the marginal impat on the probability. If the rm value is large, then
thevalueof negotiating to allowfurther developmentis high. Likewise, if the
impatontheprobabilityislarge, thenthevalueofnegotiating ishigh. With
housingadurablegood, itfollowsthat
V t ′ +1 < 0
. Asmorehousingissold, theprieofhousing, and thereforefuture prots,falls. This impliesthat industry
lobbyingeortwilldelineovertime,asthestokofhousinggrows(asthestok
ofthepubligoodshrinks).
Forhouseholds,themarginalimpaton
π
interatswiththeexpetedutilityimpatof future development,
β/(1 − β)u t − U t +1
. A greaterprobability im- pat,the moreit isworthspending. Likewise,the greatertheexpetedutilityimpat- due tofuture redutionsin the publigood, themoreeortput into
negotiations. Therefore, if the publigood has diminishingmarginalvalue to
households,thenhouseholdlobbingeortwillinreaseovertime,astheamount
ofremainingpubligoodshrinks. Takentogether,thisimpliesthatasthegame
proeeds, industry lobbying delines and householdlobbying inreases. Thus,
theprobabilitythatthegameendsaftertheurrentperiodinreaseswithtime.
Theendofperiod lobbyinggameNashequilibrium generatestwofuntions
x ∗ t (q t )
andy t ∗ (q t )
. Assumingperfet information, these funtions areinluded inthermshoieofq t
atthebeginningoftheperiod. Therstorderonditionforthisoptimizationleadstotheequilibrium ondition
p ′ t q t /n + p − c ′ t
| {z }
− x ∗′ t /n
| {z }
= (1 − π)βV t+1 ′
| {z }
+ (π 1 x ∗′ t + π 2 y t ∗′ )βV t+1
| {z }
M N R M L M F C M LI
(12)
withsymmetryassumed. Thefouromponentsofthisrelationareherelabelled
marginalnet revenue (
M N R
), marginallobbying (M L
), marginalfuture ost(
M F C
)andmarginallobbyingimpat(M LI
). Withaprobabilityπ
thatdevel-opmentwillendaftertheurrentperiod,andnolobbyinggame,
q t
ourswhereM N R = M F C
. Howthisoutomeisshiftedbythelobbyinggamedependsonhow
M L
andM LI
interat. Thisdependsonthebehaviourofx ∗ t
andy ∗ t
,andhowthatinteratswith
π
,whereπ 1 < 0
andπ 2 > 0
.Sineboth
x ∗ t
andy ∗ t
anbeeither dereasingorinreasing,there arefourpossibleombinations(see table 1). Absent the
M LI
eets, whenx ∗′ t > 0
,q t
isredued, as rmsredue developmentto redue the resultinglobbyingost.
When
M LI
ispositive,whihoursforasuientlylargeandpositivey t ∗′
,thiseetisenhaned. Development,
q t
falls asrmsareseekingto avoidengaginginalobbyingwar,rms'fold'. Forallremainingvaluesof
y ∗′ t
,M LI
isnegative,and
q t
isinreased. Forlargeandnegativevaluesofy t ∗′
,theinreaseinq t
ausesTable1: NegotiatingGame Equilibria. Households(H) hoose
y
andrms(F)hoose
x
. Inequilibria,either householdsorrmsfold,orthesystemmovesin thediretionofalobbyingwar. Withintwoells,theoutomeisindeterminateasitdependsonthepreiserelationships,notsimplythesigns.
y t ∗′ ≪ 0 y ∗′ t < 0 y t ∗′ > 0 y ∗′ t ≫ 0 x ∗′ t < 0 q t ↑
,H foldq t ↓
,towarq t ↓
,indet.q t ↓
,Ffoldx ∗′ t > 0 q t ↑
,H foldq t ↑
,indetq t ↑
,towarq t ↓
,Ffoldalargerdropin
y t ∗
thantheinreaseinx ∗ t
,essentiallyhouseholds'fold'. Wheny t ∗′
issmallinabsolutevalue,inreasingq t
maymovethesysteminthediretionof alobbing war. Thisis ertain for
y ∗′ t > 0
, and analso be saidto our ify t ∗′ < 0
butx ∗′ t + y t ∗′ > 0
. Thehoieofq t
induesaninreaseintotallobbyingeort. Notethatsinethesearemarginaleets,whetherrmsandhouseholds
areatuallyengagedinawastefullobbing warannotbedeterminedfrom the
signsalone.
When
x ∗′ t < 0
, and noM LI
, then rms inrease urrent development in order to redue lobbying expenditures. WithM LI
negative, this eet is en-haned. Thisourswhen
y ∗′ t
is negativeandlargeenough(in absolutevalue).Oneagain, householdsfold. Likewise, when
y ∗′ t
is largeandpositive,q t
is re-dued. Lobbyingeortbyrmsdoesinrease,butthedelinein
q t
issuientlysatisfyingtothehouseholdsthattotallobbyingfalls,andweagainsaythatrms
fold. With intermediate valuesfor
y ∗′ t
,q t
falls and industrylobbing inreases.When
y t ∗′ < 0
, householdlobbing also inreases, andthe systemmovesin thediretionof alobbyingwater. With
y ∗′ t > 0
, whetherornotthesystemmovestowardsalobbyingwardependsonwhether
x ∗′ t + y t ∗′ < 0
.A key poliy question is whether government has a role. At present, the
roleis often seenasfailitating negotiationby bringing stakeholders together.
Itisunlearwhethersuh failitationaets
π(x t , y t )
. Itislikelytoredue theostof partiipationfor thestakeholders,whih, allother thingsequal, would
inreasestakeholderresouresdevotedtonegotiation. Insofarasthisinrease
inresouresdoesnothangetheoutome,suh eortsaredoublywasteful. At
therstlevel,governmentresouresaredevotedtoaproessthataomplishes
nothing. Atthe seond level,these governmentresouresleveragestakeholder
resoures,whih againaomplishnothing.
Thus, this suggests that government should not be devoting resoures to
failitatingthenegotiationproess. Rather,resouresshouldbediretedatun-
derstandingtherelativestrengthintheurrentproess,andlookingforwaysto
shiftthatbalanewhihinreasesthelikelihoodofinreasingaggregatewelfare.
The objetivetherefore is to hange
π(x t , y t )
. This may involvefunding tar-getedat strengtheningommunitygroups-thehousehold lobby in themodel.
Funding should diretlyaddresswherethe weakness isin the householdinu-
eneon
π(x t , y t )
-overomingthefreeriderproblem,supportingloalresearh,providingexpertise,et. Unfortunately,itisexpetedthatanysuhshiftswould
3 Numerial Example
Anumerialexampleisdevelopedin R[34℄. Funtiondenitionsare
u t (y jt , z t ) = Az t α (Y j − y jt ) β
(13)π(x t , y t ) = a y y t + b x
(a x x t + b x ) + (a y y t + b y )
(14)p t (q t , z t ) = B(w − (z t − q t )) γ
(15)C(q it ) = Cn δ − 1 q δ it
(16)with all parameters exept
γ
assumed positive, and for diminishing marginal utility,0 < α, β < 1
. Equation 14 is a ontest suess funtion frequentlyusedfor lobbyinggames, militaryontests, and similarsituations. It isgener-
allyattributedto Tullok[40,41℄,withtwodierentaxiomizationpresentedin
Kooremanand Shoonbeek[24℄ andin Skaperdas[37℄. Theostfuntion de-
nition,equation16,hasasalingfatordependenton
n
. Withthisadjustment, theaggregate osturveremains onstant asthe numberof rmsis hanged.Thus, hangesinostsarenotdrivingresults.
The ontest suess funtion used is the ratio form. As pointed out by
Hirshleifer [20℄, this hasthe property that beause of the high marginal gain
from lobbying when total lobbying is near zero, a Nash equilibrium without
somelevel of wasteful lobbying annot exist. In the situation modelled here,
theagentthat has'theearoftheKing'likelygetstherewayifthereisnobody
elsewhispering. Assuh,theratioform istakento beappropriate.
Identiationofequilibriawasestimatedthroughabakwardindution im-
plementationof numerialdynami programming(See forexample 2). Ave-
torof 50disrete valueswasusedfor eah statevariable, generatingthe value
funtion(s)at 50 spei points. Quadrati interpolationwasusedto approx-
imatevaluefuntion levels betweenthe spei points. As for thetheoretial
development,symmetry wasimposed ontherstorder onditionsto establish
equilibriumrelations. ToloatetheNashequilibrium,asearhwasthenusedto
identifyapairofnegotiationexpendituresthatmutually zeroedtherespetive
relations. Thederivatives
x ∗′ t
andy t ∗′
werenumeriallyestimatedbyperturbingthevalueof
q t
that dened theNE. Thethus alulatedderivativeswere part ofthezerorelation thatdened theequilibriumq t
level. Thiswasiteratedonuntiltheeulideandistane betweenvaluefuntion vetorsforsuessiveperi-
odswaslessthan
10 − 10
orthenumberofiterationsexeeded100. Fortheaseomparisons, the onvergene threshold was
10 − 10
, with a maximum numberofiterationsat40. Inalargemajorityofases,theonvergene thresholdwas
attained. However,someasesenteredstableosillationsinvaluefuntionlevel.
Thesewereterminated afterthemaximumnumberofiterations.
The analytial development foused on asingle period, examining the in-
entivesgoverningthe amount ofdevelopmentin aperiod, and theamountof
0 10 20 30 40 50
0 5 10 15 20
Optimal Resource Use
z t
q t
Utility Profit Hartwick q t Hartwick k t = 0 Hartwick k t
0 500 1000 1500 2000 k t
Figure2: Optimalresoureuse,forutilitymaximizationignoringprot,prot
maximization ignoring utility, and Hartwik style utility optimization, where
protsareinvestedintoaapitalstokthatontributestohouseholdinome. All
urvesexeptthatforHartwik
k t = 0
ase(grey)plotonleftaxis. Numerialonvergeneproblemsresponsiblefor'bumpiness'.
lobbyingeort. Thenumerialresults areillustrated by linking a sequeneof
periods together, and examining the development paths, as well as expeted
totaldevelopment, totallobbying, and time till developmentstops, asseenat
thestartofthegame. Figure2showsthestatedependentresoureusefuntion
forthreealternativeoptimalityonditions. Ifrmprotsareignoredandonly
household utility is maximized, then no development ours if
z t
is lessthan8.3333,andfor
z t > 8.3333
,itisoptimaltohooseq t = z t −8.3333
. Inontrast,ifhouseholdutilityisignoredandthepresentvalueoftheprotofamonopoly
ownermaximized,alloftheresoureisused. However,spreadingostsleadsto
amoregradualdevelopmentpathovertime. Priortoanydevelopment,
z t = w
,where,inthisase,
w = 50
.AthirdoptimizationistheHartwikstylease,wherereturnsfrominvested
protareaddedtohouseholdinome. When
z t
issmall,thereismoredevelop-mentthanwithamonopolist,asthatdevelopmentinreasesommunitypopula-
tionandaggregateutility. However,developmentfalls morerapidlyhittingthe
z t
axisoinidentally loseto the point where aggregateutility is maximized.As
z t
falls,k t
inreases, inreasing household inome, and themarginal valueofthe(normal) publi good. Relativetothe pathwhere apitalaumulation
beginswhen
z t = w
,that forapitalaumulationbeginninglaterisuniformlyhigher,anddevelopmentishaltedwithasmaller
z t
value. Withoutanyapital,Although notexplored in detail, theoptimalityquestion highlightsan im-
portant aspet of deentralization. Even if full ownership of the resoure is
providedto aommunity, thereisnoguaranteethat theenvironmentwill fare
better. Ifloalsareabletoapturethebenetsofdevelopment,theymayhoose
moredevelopment. Twofeatures of the urrent model would leadto suh an
eet. First,ifthebenetsofdevelopmentarelarge,relativeto theloal value
ofthepubli good, then greaterdevelopmentis likely. Seond, ifloalinome
is low, making the marginalvalueof the inome ontribution of development
large,then greater developmentis likely. If theresourein questionhas large
aggregatevaluebeyondtheloalommunity,deentralizingontroltotheloal
levelmaybeworseforaggregatewelfarethanmaintainingentralontrol.
Foromparisonpurposes,protmaximizingresoureuseandrmvalueare
shown in gure3. Not surprisingly, rmprot is inreasingin the amountof
resoureremaining. Theimpat of hangingthe probabilityis also onsistent
withexpetations, inthat thegreatertheprobabilitythat theresourewill be
availableforusenextperiod,themoreoftheresourermsarewillingtoleave
tonextperiod. What ismoreinterestingis thefat that thevaluefuntion is
notonave. Rather,whenthereismorethanonerm,itispieewiseonave.
Thisisaonsequeneoftheinterationbetweenthedisretetimenatureofthe
optimization and the prisoner's dilemmagame that the rmsare engaged in.
Disrete time implies that ations takeplae at preise points in time. Here,
a spei quantity of resoure is used in eah period, with all the remaining
resoureused upin some nal period. Onetherefore does nothaveasmooth
onsumption path over time, but a sequene of distint quantities. This is
responsibleforthekinksor disontinuitiesintheurves.
Theprisoner'sdilemmabetweenimperfetlyompetitiverms,inthisase,
hasthermsusingmorethantheoptimalamountoftheresoure,astheyaim
to apture it before their rival does. In panel (b), this manifests itself in the
fatthatthepointatwhihtheindustrynolongeronsumesalloftheresoure
intherstperiodoursforalarger
z t
thegreaterthenumberofrms. Thisismarkedbythepointwheretheresoureuseurversthasakink. Fortherst
segmentafterthekink,theremainderoftheresoureisonsumedintheseond
period. However,onernsaboutbeing'beat'totheresoureintherstperiod
are no longer drivers. As
z t
is inreased, a point is reahed where the rmsnolonger use all the remainingresoure in the seond period. This produes
a downward shift in the optimal resoure use, a onsequene of the drop in
the slope of the value funtion. The quadrati interpolation used smoothed
thekink in thevaluefuntion thatorrespondsto thehangein resoureuse,
distorting the graphsomewhat. In panel (a), for the
π = 0.0
ase, the valuefuntion didnotonverge,but ratherentered aylialpattern fortwoof the
50
z t
levels,representedbythesawtoothjustbelowz t = 40
. Astheprobability thatdevelopmentwill takeplaenextperiod inreases,panel(a), thedistanebetweentheslopehangesandbetweenthedisontinuitiesdelines,andthesize
ofthedownwardstepsinreases.
Figure 4illustrates the impat of hangingthe eetiveness of negotiating
q t 0 5 10 15 20 25
a) Use by Probability, n = 2
π = 0.0 π = 0.2 π = 0.5 π = 0.8 π = 1.0
b) Use by Concentration, π = 0.5
n = 1 n = 2 n = 4
0 10 20 30 40 50
0 200 400 600 800
z t
V t ( z t )
c) Value by Probability, n = 2
π = 0.0 π = 0.2 π = 0.5 π = 0.8 π = 1.0
z t
0 10 20 30 40 50
d) Value by Concentration, π = 0.5
n = 1 n = 2 n = 4
Figure3: Resoureuseandrmvalueasfuntionsofresoureremainingunused.
Panels(a)and(b)showtheamountoftheresoureusedintheurrentperiod,
giventhe amount remaining. Panels() and (d) report thevalueof the value
funtion,whih isthepresentvalueoftheoptimal pathfromthe urrentdate
forward. Inpanels (a) and (), grey segment for
π = 1.0
indiates alternateresoureuseandvaluefuntionforvaluefuntionyle. Inpanel(d),greylines
marktotalindustryvalue,whileblaklinesmarkindividualrmvalue.
eort(
a x
anda y
)andtheresponsivenessoftheregulator(b x
andb y
). Inreasingtheeetiveness orrespondsto making iteasier forstakeholders to engage in
negotiation,astheeterusparibus impatofmorenegotiationeortisgreater.
Inreasingtheregulatorresponsiveness(reduing
b x
andb y
)reduesthestrengthoftheregulator'sbias. Resultsareallbasedonaninitialpubligoodlevel
z 0 = w
. Thegraphsshowtheexpetednumberofperiodsduringwhihdevelopment oftheresouretakesplae,expetedpubligoodlevelwhendevelopmentends,andpresentvalueofindustryandhouseholdexpendituresonnegotiationeort.
Allresultsarerelativetothemonopolistasewith
a x = a y = 1
andb x = b y = 1
.Panelsa,,andeshowtheimpatofhangingtheeetivenessofnegotiation
eort. Inreasing thesize of
a i
inreasestheimpatofadditionalexpenditures onnegotiation eort foragent typei
. Inreasing the eetiveness of industrylobbying inreases the duration of development ativity and redues the ex-
petednal publigoodlevel,foreahindustryonentrationlevel. Inreasing
rmnumberstendstoreduethedurationofdevelopment,whileinreasingthe
expetednalpubligoodlevel. Twofreeridingeetsontributetothis. First,
withmorerms,eahseekstofreerideontheotherslobbyingeort. Seond,the
openaessnatureoftheresourerightsresultsininreasingdevelopmenteah
period. Takentogether,householdlobbyinginreasesmorerapidlyin response
tofasterdevelopment,andrmsareunable tooordinateonounteratingit.
Negotiationeortsfollowamoreinterestingpattern. When
n = 4
,householdeort is greatest when the eetiveness levels are approximatelyequal. This
eortis suientto overpowertheindustrylobbyingeort,protetingafairly
high level of the publi good. Inreasing the eetiveness of the household
lobbying inreases the expeted nal publi good level and redues the Nash
equilibriumnegotiationeorts. Likewise,reduinghouseholdnegotiationeort
eetiveness redues the NE negotiation eorts, andresults in equilibria with
lowernalpubligoodlevels.Inthemostextremease,thereisalmostnopubli
goodleft, but rmstakealongtimeto developalltheresoure. When
n = 2
,the greatest household expenditure on lobbying takes plae when households
aresomewhatmoreeetivethanrmsintheirlobbying. Otherwise,theeets
are similar. When
n = 1
, there is evidene of a lobbying war. Whena y
issomewhat larger than
a x
then both industry and household lobbying eortsare high. However,there is notmuh dierene in the outome than what is
observedwithmorethanonerm.
Whenhangesintheresponsivenessoftheregulatorareonsidered,expeted
nalpubligoodlevelsaregenerallyinreasingwiththenumberofrms,while
durationtends to deline. Forany partiular numberof rms, both duration
and expeted nal publigood level are not very responsive to hanges in
b x
and
b y
. Thisissomewhatsurprising,sinetheregulator'sdefault(absentlobby- ing)probabilityofendingfurtherdevelopmentrangesfromapproximately0.01(
0.1/10.1
)toapproximately0.99(10.0/10.1
). Theindustryandthehouseholdsrespond in suh a way that the bias is oset. Notie that as the bias shifts
to favourone type of player - inreasing
b x
favours rms - negotiation eort from that type of playerfalls. Thisours forboth playertypes, but is morepronounedfor thehouseholds. The freeriding eet on industrynegotiation
T z
x y
0.2 1.0 5.0
a) Effectiveness, n = 1
a y 0.2 1.0 5.0
c) Effectiveness, n = 2
a y
0.2 1.0 5.0
0.2 1.0 5.0
e) Effectiveness, n = 4
a x
a y
T z
x y
0.1 1.0 10.0
b) Responsiveness, n = 1
b y 0.1 1.0 10.0
d) Responsiveness, n = 2
b y
0.1 1.0 10.0
0.1 1.0 10.0
f) Responsiveness, n = 4
b x
b y
Figure4: Expetedduration(
T
),expetednalpubligoodlevel(z
),expetedPV of rm (
x
) and household (y
) negotiation eort, for various expenditure eetivenessand regulatorresponsivenessandz 0 = w
. Radius ofirlesetionmeasuresvariablerelativeto monopolyasewith
a x = a y = 1
andb x = b y = 1
.Graydashedirleinpanels()through(f)enablereferenetomonopolyase.
14
Eorts to inrease stakeholder engagement are analogous to hanging ef-
fetiveness andresponsiveness. Changingresponsivenesshasalmost noimpat
onexpeted time to developmentessation,nor onexpeted remainingpubli
good. Responsiveness is analogous to a regulator ontributionto lobbyingef-
fort,substitutingforhouseholdorrmlobbyingeortandleavingtheoutome
unhanged. It doesnot hange the marginalvalue of lobbying, and therefore
doesnotenouragemuh ofahange. Inontrast, hangingeetivenessdoes
hangethemarginalimpat. Whatstandsouthereisthat symmetrihanges
in eetiveness,suh asaregulatoropeningtheproessin anequalwaytoall
stakeholders,doesalmostnothingtohangetheoutome. Theoutomeisonly
hangedifengagementfavoursoneparty. However,eventhenitmayresultina
largeinreasein expetedlobbyingandamoveawayfrom theeientlevelof
thepubligood. Inreasingstakeholderengagementmayhavelittleimpaton
outomes, and mighteven makethings worse. Thus, whilepriniples of good
governanemaybeonsistentwithinreasingpartiipation,itisfarfromertain
thatitwillimprovetheeonomieienyoftheoutomes.
Figure5showstheeetofhangingtheutilityfuntionparametersandone
demandfuntion parameter. Whenhangingtheutilityfuntion,theelastiity
parameters were restrited to satisfy
α + β = 1
. In panels (a) through (),theexpeted nal publigood level inreasesasboth
α
andY i
are inreased,while duration falls. This is aonsequene of the fat that inreasing
α
andinreasing
Y i
eah inrease the marginalutility of the publi good relative toresidual inome. Thus, the marginal benet of lobbying inreases, leading to
aninreaseinthisativitybythehouseholds. Inreasingrmnumbersredues
lobbying or negotiation eort by the rms. However, the response patterns
dierforeahindustryonentrationlevel. Whendevelopmentismonopolized,
thermrespondstoinreasedhouseholdlobbingbyinreasingitsownlobbying
eort. Forthelower
α
andY i
levels,thiseetisalsopresentfortheoligopolisti industry. However,whenα = 0.5
andY i = 50
,industrylobbyingeorthasfallenrelativetotheaseswhere
α = 0.25
andY i = 50
orα = 0.5
andY i = 10
. Forallasesinpanel (b),household lobbying eortisgreaterthanin panel(a). Free
riding by therms inreasesthe marginalprodutivity of household lobbying
eort,whih for large
α
andY i
is suessful in further driving down industrylobbying. Although dominated bythe free riding eet,in panel () industry
lobbyingagain falls for the highest
α
andY i
levels. For lowα
andY i
levels,householdsinreaselobbying, relativeto the
n = 2
ase. However,the abilityofhouseholdlobbyingtodrivedownindustrylobbyingisstrongenoughwhen
α
and
Y i
arelargethathouseholdlobbyingatuallyfallsrelativetothen = 2
ase.Whenthepubligoodisvaluableenoughtothehouseholds,andhouseholdshave
enoughinome, rms essentially apitulate in the lobbying game, whih then
reduesthehousehold'sneedtolobby. Notsurprisingly,wealthyneighbourhoods
gettheirwayatlowerostthanlesswealthyneighbourhoods.
Aninterationbetweenthefreeridingeetonlobbyingandthefreeriding
eetondevelopmentisalso evidentin panel(d). When
γ
issmall,inreasingn
redues durationandinreasesthenalpubli good level. This isdrivenbyinreasedlobbyingfromhouseholds,leadingtoashiftinontinuationprobabili-
tiesfavouringthehouseholds. Thus,theexpetednalpubligoodlevelishere
inreasing in the number of rms. When
γ = −0.5
, rms essentially give up onanydevelopmentafter therstperiod. Consequently,theydonotinvest innegotiationeort,makingitunneessaryforhouseholdstoinvestmuh. Redu-
ingrmonentrationnowredues thenal publigood level,aswithgreater
onentration,rmsreduerstperioddevelopmenttoinreaserentsaptured.
4 Disussion
Mostalloationdeisionsaroundniteresouresarenotmadebyanomnisient
regulator. In general, the regulator is inuened by the aeted parties, the
stakeholders,in some way. At one extreme, stakeholders may diretly or in-
diretly tryto bribetheregulator. Theunfavourableonnotation to theterm
lobbyingseemsto reetthissomewhatshadysideoftryingtoinueneareg-
ulator'sdeision. Atpresent,inresponse toawidelyhelddisillusionmentwith
theonventionalproess,manyareenouragingstakeholderstodiretlyengage
eah other. This proess may appear dierent, but fundamentally it still re-
quires stakeholders to expend resouresin an eort to inuene the resoure
managementdeisionintheirfavour. Now,ratherthandiretlybuyingfavours,
theyhavetoshowthattheyareooperatingwithotherstakeholdersinresolving
theresouresharingonit.
Hereinwehaveonsideredasimplemodelofaresouremanagementprob-
lem,wherethemanagementdeisiononsistsofhaltingallfurtherdevelopment
of theresoure. Theresoureis bothnite and durable, onsistentwith land
orrightstoanannualalloationofwater. Bothpro-andanti-developmentlob-
bies attempt to inuene the deision of aregulator with nal authorityover
whether further developmentours. Fouroutomesare identied, aseswith
limitedlobbying,those where thepro-developmentlobbyoverpowerstheanti-
developmentlobby,theonverse,andalobbyingwar.Thislatteraseisperhaps
themostdisturbing,asitsuggeststhatlobbying,whetherastraditionallyenvi-
sioned,ornowas'stakeholderengagement'mayhavelittleimpatonthenal
outome. Assuh,lobbyingissimplyawasteofresoures,engagedin beause
partiipantsare involvedin aprisoner's dilemma. Inpartiular, iflobbyingis
madeeasier forstakeholders to engage in by governmentfailitating meetings
and other forms of interations, losses may be exaerbated. In therst ase,
governmentspendsmoneytolowertheostofinvolvementintheproess. The
redutioninthemarginalostofinueningtheregulatorinduesstakeholders
to spend moreonthis ativity. In theend, theoutome maybeno dierent,
butoverallostshaveinreased.
These resultsalso suggestthat ompensation paymentsfor habitat prote-
tionmaybesensitivetoindustrystruture. Thepreseneorabseneofompen-
sationhas beenshownto aet therateof development, withunompensated
takings hastening development and full ompensation slowing it [5, 21, 35℄.
T z
x y
0.1 0.2 0.5
2 10 50
a) ∆α and ∆ Y i , n = 1
α Y i
0.1 0.2 0.5
2 10 50
b) ∆α and ∆ Y i , n = 2
α Y i
0.1 0.2 0.5
2 10 50
c) ∆α and ∆ Y i , n = 4
α Y i
T z
x y
−0.10 −0.25 −0.50
1 2 4
d) ∆γ
γ
n
Figure5: Variationsin utilityfuntion parametersand demandelastiity. For
panels(a) to (),
α
andY i
respetively measureutility elastiity ofthepubli good and residual inome. In panel (d),γ
measures the prie elastiity ofdemand. Quarter irles have been lipped to prevent overlap with adjaent
graphobjets.
good value [22℄. Although ompensation is not the fous of our model, our
resultsdospeakto this issue. Inpartiular, ifthepurpose ofompensationis
to pay rmsthe present valueof the assetbeingtaken, our resultsshow that
industry struture an play an important role in determining the size of the
presentvalue-thelessonentratedtheindustry,thelowerthepayment. Thus,
althoughthemonopolistmaybethefriendoftheonservationist,intermsofthe
rateofdevelopment,themonopolistdoesnotlettheonservationistoheaply
ifdevelopmentistobehalted.
The ineienies highlighted here are a onsequene of the lak of lear
propertyrights. In thease of land, ownersdo nothavean exlusiverightto
deidehowto usetheirproperty. Therightto apartiularuse-development-
mustbeatedupon before theregulatordeidesthat enoughsuhrightshave
been atedon. With land, suh 'takings' themselvestypiallyleadto aset of
legalbattlesaboutwhetherornotthepropertyownerdidown adevelopment
right-requiringompensation-ornot. Forwater,inontrast,rightsaretypi-
allyusufrutory,andonlyalloatedwhenaneedisdemonstrated. Thus,water
rightsarenotowneduntil developmentthatanuseithastakenplae. When
property rights to waterare vested with the user, then arrangementssuh as
leasingorpurhaseanprotetpublivaluesfrominstreamows. Inbothases,
thekeypropertyrightisessentiallyopenaess. Clarityofthelegalrighttode-
veloppropertyortoaesswater,independentofwhenthoseusesour,would
reduetheinentiveproblems analyzedin thispaper.
The roleof rmsize also bears somereetion. As modelled, there areno
eonomies of sale related to development osts aptured by the rms. This
is likelya reasonablereetionof the onstrution industry, where thetrades
(eletriians,plumbers,arpenters)aregenerallyharaterizedbyalargenum-
ber of small rms. The benets of size in the model ome from a redution
infreeriding inentivesbothin lobbyingandin thenal market. Theurrent
trend towards relatively large developments, with multiple year development
plans,maybeameansofpre-emptingthelobbyingproess. Firmsareableto
seuretheirdevelopmentrightsbeforetheoupantsarrive,whomaydesireless
development. An interesting empirial analysis would beto searh for arela-
tionship betweendevelopmentprojetsize (units, area, oryears tobuild) and
perunit publi spae. Iflargerprojetsarepre-empting thelobbyingproess,
thenpublispaeshouldbesmaller.
5 Conlusion
Astylizedlanddevelopmentindustrywhihonsumesapubligoodgenerating
resoure(openspae,surfaewater)ismodelledinteratingwithommunityo-
upantsinalobbyinggame. Householdsandtheindustryarelobbyingtoaet
whether further development is allowed. Lobbying eortinvested depends on
therelativepoweroftheagents,withsomeombinationsresultingin alobby-
ingwar. Inreasingtheeetiveness oflobbying,akintoreduingthemarginal
outomewhenbothinterestsareaetedsymmetrially. Thissuggeststhatthe
urrentlypopularstakeholderengagementeorts may dolittle to hange out-
omes. Iftheseinitiativesareostlytoimplement,andiftheyenouragegreater
lobbyingeortbythestakeholders,thentheyarewasteful. Ifitisknownwhih
aetedpartyisdisproportionatelypowerless,thentheeienyoftheoutome
anbeimprovedbyequalizingthepowerbalane. However,insodoing,thereis
theriskthatawastefullobbyingwarandevelop. Further,suhshiftingofthe
powerbalanepresupposesthattheregulatorknowsinwhihdiretionthee-
ientamountofdevelopmentis,apossibleproblem. Theresultsofthisanalysis
alsosuggestthat trendstowardslargerdevelopmentprojetsandonentrated
developmentrmsmayin partbearesponseto thethreatoflobbying.
Referenes
[1℄ Adams, G., Rausser, G., Simon, L., 1996. Modelling multilateral nego-
tiations: An appliation to Californiawater poliy. Journal of Eonomi
BehaviorandOrganization30,97111.
[2℄ Amman,H.M.,Kendrik,D.A., Rust,J.(Eds.),1996.Handbookofom-
putational eonomis.Elsevier.
[3℄ Beierle,T.C.,Konisky,D.M.,2000.Values,onit,andtrustinpartiipa-
tory environmentalplanning.JournalofPoliyAnalysisandManagement
19(4), 587602.
[4℄ Bellman, R., 1957. Dynami Programming. Prineton University Press,
Prineton,NewJersey.
[5℄ Blume,L.,Rubinfeld,D.L.,Shapiro,P.,February1984.Thetakingofland:
Whenshouldompensationbepaid? TheQuarterlyJournalofEonomis
99(1), 7192.
[6℄ Brooker, J. F., Young, R. A., 1994. Modelling intrastate and interstate
marketsforColoradoRiverwaterresoures.JournalofEnvironmentalEo-
nomisandManagement26,6687.
[7℄ Carraro, C., Marhiori, C., Sgobbi, A., June 2005. Appliations of nego-
tiation theory to water issues. Poliy ResearhWorking Paper 3641,The
WorldBank,NewYork,NewYork.
[8℄ Cummings, R. G., Holt, C. A., Laury, S. K., January2002. TheGeorgia
irrigation redution aution: Experiments and implementation. Georgia
stateuniversityworkingpaper,GeorgiaStateUniversity.
[9℄ Daubert, J.T., Young, R. A., November1981. Rereationaldemands for
maintaining instream ows: A ontingent valuation approah. Amerian
Journalof AgriulturalEonomis63(4),666676.
and revenue division, applied to environmental poliy. European Journal
ofPolitialEonomy14,281301.
[11℄ Dueld, J. W., Brown, T. C., Allen, S. D., September 1994. Eonomi
value of instreamowin MontanaBig-Hole and Bitterrootrivers.Roky
MountainForestandRangeExperimentStationResearhPaperRM-317,
United StatesDepartmentofAgriultureForestServie.
[12℄ Dueld, J.W.,Neher,C.J.,Brown,T.,September1992.Rereationben-
ets of instream ow - appliation to Montana Big Hole and Bitterroot
rivers.WaterResouresResearh28(9), 21692181.
[13℄ Fishel,W. A.,Marh2001.Homevoters,muniipalorporategovernane,
and the benet view of the property tax. National Tax Journal 54 (1),
157173.
[14℄ Fishel, W. A., February2001. Why are there nimbys? Land Eonomis
77(1), 144152.
[15℄ Graihen, P. R., Requate, T., Dijkstra, B.R., August 2001. How to win
the politial ontest: A monopolist vs. environmentalists. Publi Choie
108(3-4),273293.
[16℄ Green, G.P.,O'Connor, J.P., April2001.Waterbankingandrestoration
ofendangeredspeieshabitat: AnappliationtotheSnakeRiver.Contem-
poraryEonomiPoliy19(2),225237.
[17℄ Grin, R.C., Hsu, S.-H.,May1993. Thepotentialforwatermarket e-
ieny when instreamowshavevalue. AmerianJournalof Agriultural
Eonomis75(2), 292303.
[18℄ Hartwik,J. M.,1977. Intergenerational equityand theinvesting ofrents
from exhaustibleresoures.AmerianEonomi Review67,972974.
[19℄ Henderson,V.,Beker,R.,2000.Politialeonomyofitysizesandforma-
tion.Journalof UrbanEonomis48,453484.
[20℄ Hirshleifer,J.,1989.Conitandrent-seekingsuessfuntions: Ratiovs.
dierenemodelsof relativesuess.PubliChoie63,101112.
[21℄ Innes, R., Otober1997. Takings,ompensation, andequaltreatmentfor
ownersof developed and undevelopedproperty. Journalof Lawand Eo-
nomis40(2),403432.
[22℄ Innes, R., Polasky, S., Tshirhart, J., Summer 1998. Takings, ompensa-
tion and endangered speies protetion on private lands. The Journalof
Eonomi Perspetives12(3),3552.
[23℄ Irvin,R.A.,Stansbury,J.,Jan/Feb2004.Citizenpartiipationindeision
making: Isitworththeeort? PubliAdministrationReview61(1),5565.
funtions intullok'srent-seekingontest.EonomisLetters56,5961.
[25℄ Landry, C., Otober 1999. Market transfers of water for environmental
protetionintheWesternUnited States.WaterPoliy 1(5),457469.
[26℄ Leah, W. D.,Pelkey, N. W., Sabatier, P. A., 2002. Stakeholderpartner-
shipsas ollaborativepoliymaking: Evaluation riteriaappliedto water-
shedmanagementinCaliforniaandWashington.JournalofPoliyAnalysis
andManagement21(4),645670.
[27℄ Linster, B.G.,1994.Cooperativerent-seeking.PubliChoie81,2334.
[28℄ Loomis, J. B., Otober1998. Estimating the publi'svalues for instream
ow:eonomitehniquesanddollarvalues.JournaloftheAmerianWater
ResouresAssoiation34(5),10071014.
[29℄ Loomis,J.B.,November2000.Environmentalvaluationtehniquesinwa-
ter resoure deision making. Journal of Water Resoures Planning and
Management126(6),339344.
[30℄ Lubell,M.,2004.Collaborativeenvironmentalinstitutions: Alltalkandno
ation? JournalofPoliyAnalysisandManagement23(3),549573.
[31℄ Lubell,M.,Feiok,R.C.,Ramirez,E.,July2005.Politialinstitutionsand
onservationbyloal governments.UrbanAairsReview 40(6), 706729.
[32℄ Murphy, J. J., Howitt, R. E., June 1998. The role of instream ows in
a water market: Using experimental eonomis to address environmental
issues. Presented at the World Congress of Resoure and Environmental
Eonomis.
[33℄ Pye-Smith, C., Feyerabend, G. B., Sandbrook, R., 1994. The Wealth of
Communities: Stories of Suess in Loal Environmental ManagementC
Pye-Smith, GB Feyerabend, R Sandbrook - 1994- Earthsan℄. Kumarian
Press.
[34℄ R Development Core Team, 2006. R: A Language and Environment for
Statistial Computing. R Foundation for Statistial Computing, Vienna,
Austria,ISBN 3-900051-07-0.
URLhttp://www.R-projet.org
[35℄ Riddiough, T. J., 1997. The eonomi onsequenes of regulatory taking
riskon landvalueand developmentativity.JournalofUrban Eonomis
41,5677.
[36℄ Shumaher,E.F.,1999.SmallIsBeautiful: EonomisAsIfPeopleMat-
tered: 25YearsLater...WithCommentaries! Hartley&MarksPublishers.
[37℄ Skaperdas,S., 1996. Contest suess funtions.Eonomi Theory 7,283
290.
approahto deiding whowillsupply instreamowwater.Journalof the
AmerianWaterResouresAssoiation38(4),959966.
[39℄ Supalla, R. J., June 2000. A game theoreti analysis of institutional ar-
rangements for platte river management. International Journal of Water
ResoureDevelopment16(2), 253264.
[40℄ Tullok,G.,1967.Thewelfareostsoftaris,monopolies,andtheft.West-
ernEonomiJournal5,224232.
[41℄ Tullok, G., 1980. Eient rent-seeking. In: Buhanan, J., Tollison, R.,
Tullok,G.(Eds.),Towardatheoryoftherent-seekingsoiety.TexasA&M
UniversityPress,CollegeStation, pp.97112.
[42℄ Woolley, J.T.,MGinnis,M.V., Kellner,J.,Winter 2002.TheCalifornia
watershedmovement: Sieneandthepolitisofplae. NaturalResoures
Journal42(1), 133183.