The effects of the nonmarket urban land use system in Eastern Europe: the case of Belgrade

9 Pages·0.416 MB·English

by  Begoviс B., Mijatovic B.

Treballs de la SocietatCatalana de Geografía - N° 35 - vol. VIII
The effects of the nonmarket urban land use
system in Eastern Europe: the case of Belgrade
Boris BEGOVIé
Centrefor Economic Studies MECON, Belgrade
Bosko MUATOVIé
The Economic Institute, Belgrade
1. Introduction
The politicalchanges which took place inthe late 1980sin EasternEuropecleared the
wayforthe radicalreforms oftheeconomies ofthese countries, takingthem from acom
mand (planned) to the market system. Although the general strategy of such a transition
is clear, there are many distinctive waysto apply the general strategy. The choice varies
from countrytocountryandfrom onesegmentoftheeconomy toanother. Acrucial pre
requisite formaking thecorrectchoice ofaparticularstrategy isknowledgeofthecurrent
(previous) nonmarket system, its features and how it worked. An appropriate diagnosis
of the current (previous) system and its effects can be solid ground for the formulation
ofasound transitionstrategy.Therefore, considerableeffort has been put intoananalysis
of the nonmarket economic system and its various segments.
Nonetheless, one of the segments of the economic system which has not been widely
studied until nowisurbanland use systeminEastern Europe. Hence, the intention ofthis
paper.isto examine the crucial features of the nonmarket urban land use systemand to
analyze its effects on urban development. This will be done using the case of Belgrade
as a representative Eastern European city. Accordingly, the institutional framework and
the crucial features of the current (nonmarket) urban land use system will be analyzed.
Itseconomicand financial effectsaswellasitsimpact onurbandevelopment willbe sur
veyedand evaluated. Such a procedure will hopefully provide a solid basis from which
toconsiderthereform ofthenonmarketurban landuse systeminEasternEuropeancities.
7
2. Theinstitutionalframeworkofthe nonmarketurbanlandusesystem
¡
The economic systemwhichexistedinYugoslaviabeforethe 1991civilwarwassimilar
to the Soviet economic system. The crucial features of such a socialist system were:
a) the monopoly of public ownership of virtually all assets and wealth;
b) a nonexisting market; Le. no market allocation ofeither production factors or goods;
e) administrative (command) allocation of both production factors andgoods.
At the core of Yugoslaveconomic systemwaspublic ownership, called «socialowner
ship»according to Yugoslavlegal terminology. Social (public) ownership dominated the
major industries (the manufacturing industry, transportation, commerce, banking, insu
rance, etc.) aswellasnationalwealth(urbanland, etc.). There werejustafewcaseswhe
rethismonopolydidnotapply,soprivateownershipwasallowedonIyinagriculture (agri
culturallandup to 10ha) andinthe smalltrade and technical services sector (firms with
up to five employees).
Nonetheless, the Yugoslaveconomic system(also known as socialist selfmanagement)
had several distinctive features setting it apart from the Sovietsystem.' The first feature
wasthesocialownershipsystemwhich, contrary totheSovietstateownershipsystem,was
characterized by the absence of a recognized property owner and well defined property
rights. Hence, socially owned assets or wealth did not belong to anyone specific (either
tothe Stateor tothe employees), butbelong to society.This rather nebulous concept has
eliminated arecognized owner(withrelatedpropertyrights) ofthevastmajority ofassets
and wealth. The concept ofsocial ownership wasdeveloped in compliance with the idea
thatpropertyrightsarenotimportant foreconomic performance(efficiency),butonIyfor
the distributionofretums (earnings). The basic idea ofthe architects ofthe system(Kar
delj (1971)) was to abolish every type of proper ownership, thereby abolishing
all revenues based on ownership (non-labor revenues).2
The seconddistinction wasthatthemarket wasnotreplaced byacentral planbodybut,
inthelast twentyyearsatleast, bytheconcept of«socialbargaining», wherebyallocation
resultedfromasystem,inwhichthenumerous parties concernedwouldnegotiateandset
tleall important economic issues (supply,prices, wages, etc.) through relevantinstitutio
nal mechanisms.
The architects ofthe systemconsidered that social bargaining wouldlead to the settle
ment of all conflicts based on partial interests.
Since there was no production factor market allocation, there was no labor or capital
market (stockexchange), no securities (capitalwasimmobile), employedlabor could not
be fired (lose their jobs), wageswere determined exogenously,and interest rates (which
werefixedbythe state)werethe sameatallbanks and alwayslowerthan inflation (hence
there were no earnings from capital).
Contrary to the Soviet system, there wasa goods market in ex-Yugoslavia, although it
wasnot quite similar tothe goodsmarkets inWesterncountries. It wascarteled andpro
tected from foreigncompetition andmarketinformation wasnot crucial fordecision ma
king in firms.
8
Although there is controversy as to whether such a system provided dynamic growth
in its early phase,? there is consensus that the system as such did not provide economic
efficiency: there was no way for the efficientallocation of resources. Maximizing wages
wasthe goalofthe self-managementfirms, sosuch firms characterizedadistortionalreac
tion to the changing conditions on the market. Furthermore, the benevolent socialist go
vernmentmadethe budgetconstraintofself-managedfirms very softbymeansofa gene
rouscreditandsubsidypolicy.Theresultwasapathy,mismanagement,inefficiency,inflation,
unemployment, zero growth and big structural debalances.
Thefeatures ofthegeneraleconomicsystemdeterminedtheinstitutionsand mechanisms
ofthe urbanland use system: the monopolyofpublic ownershipand the lackofa market
allocation of resources.
3. Tbefeatures andmechanismsofthe nonmarketurbanlanduse system
Thereisamonopolyofpublicurbanlandownership, known in Belgradeas «social lan
downership», The actuallandowneris the municipal (city) authority. This monopoly was
establishedbythe nationalization of urbanland which was completed in 1958.4 As there
is a monopoly of public urban landownership there can be no trading and other related
transactions conceming urban land.
Since the urban land trade is not feasible, the necessary condition for the existence of
anurbanlandmarketismissing. Hence, thereare no markettransactions, no marketland
price, or marketrentoThe nonexistenceofan urbanlandmarketimpliesthereis no mar
ket allocation of urban land, but only administrative allocation.
Sincethere isa monopoly ofpublicurbanlandownership, privatelyownedagricultural
land on the urban fringe is transformed into publicly owned urban land by compulsory
purchase. This compulsorypurchaseisappliedinall cases, regardlessoffuture (planned)
landuse. Acompensationispaidtothepreviousownerand the levelofthis compensation
isarbitrarily(unilaterally)determinedbythe municipalauthority. Ithas nothingtodo with
the marketvalue (price)ofagiven plotofland(the compensationforasq.met. ofpurcha
sed land equals 0,5% of the average price of a sq.met. social housing floor space). As
the city grows, allplots oflandontheurbanfringe are transferredtopublic landownership.
Theessenceofthe administrativeallocationistodonateurbanlandtodevelopersand/or
users. Theplot ofurbanland isstillpubliclyowned,butthedeveloperand/oruserisgiven
the rightto develop this plot and to retain the building in his ownership. This regulation
is somewhatsimilartothe institutionoflandleaseholdin marketeconomies, with the ex
ception of two points: a) unlike a leasehold arrangement the urban land donation is not
restricted in time, Le. no expiration date is specified up to which time the plot ofland
can be used; b) no market rent is paid by the developer/user. Furthermore, the right to
developaplot ofdonatedurbanlandis nontransferable. Hencedevelopmenthas tobe ca
rried out by the selected developer/user.
Urbanlanddonation isbasedon the discriminatory right ofthe local authorities to se
lectpotentialusersaccordingtoarbitrarilyestablishedcriteria. AlIpotentialusers(private
9
housing as well as socially owned companies) apply to the municipal authorities for the
urbanlanddonation. Theprocedureto selectauserisnot strictlydefined, soapplications
are evaluated according to arbitrarily established (even ad hoc) criteria..The municipal
authorities can refuse any application without giving reasons for doing so.
Three fees are paidby the developers/users ofthe urban land. An urbanland donation
fee is paidas apremiumto the municipalauthoritybythe developer/userbeforedevelop
ment begins. This fee is calculated according to the blueprint floor space. The amount
ofthe feeisdeterminedbydecisionofthe municipalauthorityand this amountisdifferen
tiatedaccordingtothe locationandplannedlanduse. The level ofthe urbanlanddonation
feecan bechangedunilaterallybythe municipalauthorityat anytimeand byany amount,
butnowadays it is usuallychangedonce ayear. The amountofthis fee usually comprises
8-10% of the total property investment (development costs).
An urban land servicing fee is also paid as a premium to the municipal authority by
the developer/userbeforedevelopmentbegins. Urbanland servicingincludessite clearing
(only in the caseofredevelopment) and providingthe on-siteinfrastructure, whichinclu
des connectionstothe watersupply, sewage, powersupply and gas supplynetworks, hea
ting system and telecornmunications. This feeequals the actual land servicing costs of a
givenlocation, sothe average fee (amountoffeepergroundsq. met.) variesconsiderably
from location to location. The amount of this fee usually comprises 35-45%ofthe total
investment.
An urban land use fee is paid regularly (monthly) by the user of a givenplot of urban
land (location). Thisfee is calculatedaccordingto the floor space. Theamountof the fee
is determined by the municipal authority and is differentiated according to the location
and the actualor plannedlanduse. Theamountofthe urbanlanduse feecan also be chan
ged unilaterallybythe municipal authority's decisionat any timeand by any amount, but
nowadays it is usually changed quarterly. For many years the levelofthe urban land use
fee has been virtually negligible."
AlIthe financial resources obtainedthroughthe paymentofsuchfees are usedto finan
ce infrastructureinvestmentsrelatedtourbandevelopment, Le. urbanutilitiessupply(wa
ter supply and sewage, power and gas supply, the heating system, public transportation
and telecommunications)and tofinancethe supplyofurbanpublicutilities(roads, streets,
pavements, pathways, parks, etc.). Theinvestments relatedtourbanutilities cannotbe fi
nancedfrom marketrevenues sincethe pricesoftheseutilities havebeendistorteddown
wards as the result of a policy prescribing that prices should cover operating costs only,
Thereforetheseinvestmentsmustbe financedexternally. On the otherhand, the character
of public urban utilities is such that their supply and development must be financed ex
temally.
4. The economic and financial effects of the nonmarket urban land
use system
Themost significanteffectofthe nonmarketurbanlanduse system is thatthere are no
market land prices, or market rents. As the existence of competitive market prices is a
10
necessary conditionfor the efficient allocation ofurban land as a resource (changing the
parterooflandusefrom lowertohigher valueuse), theglobaluseofurbanland inBelgra
deisinefficient. Therearetwobasic inefficiencies: thephysicalinefficiency ofurban land
use and its economic inefficiency. There are also two cases where both types of ineffi
ciency occur: inefficiency in new development and/or redevelopment and inefficiency in
theuseofurbanlandbytheactualoccupants (users). Anexamination ofallcases ofineffi
ciency follows.
Physical inefficiency in the case of new development results from the urban fact that
theland donationfeeisanextremely downwarddistorted surrogateofthe urbanlandpri
ce. It is reasonable to estimate that the market price would be much higher, particularly
inthecitycenter.Arelativelylowurbanlanddonationfeeisnotanincentiveforthephysical
efficiency ofland use, so a relatively low floor-to-ground index and low densities result
from this distortion. Furthermore, the urban land donation feeis calculatedaccording to
the blueprint floor space, not he ground (land) space. Such a method of fee calculation
means that there is no incentive whatsoever for the efficient physical use of urban land
(ground). Hence, the averagecost (amount ofthe feeper floor sq. met.) is constant and
financially itisallthe sametothedeveloper/userwhetherasingle flor or ahundredfloor
building is constructed on a given location.
Economic inefficiency in the case of new development is a direct consequence of the
absence of a urban land market. If there were a market there would be competition for
a plot of land at a given location, so that the location would be used by the prospective
user who offered the highest bid, Le. the user to whom a given plot of land is the most
valuable.Sinceefficientallocationrequiresthatresources (urbanlandinthiscase)bechan
neled to their highest valueuse, agiven plot ofland would be used efficiently. However,
the actual nonmarket, noncompetitive manner ofselecting the users bythe local authori
ties offersno guaranteeofeconomicefficiencyinthe case ofdevelopment and or/redeve
lopment.
Physical inefficiency by the actual occupants is encouraged by the negligible level of
theurban land use feewhich theyhavetopayo The actuallocationusers donotpayamar
ket rent which, it is reasonable to expect, would be far higher than the urban land use
fee. Hence this fee is a negligible expenditure for the occupants..That is why the ineffi
cientusers ofspacecharacterizedbylowdensities, suchasmanufacturingindustryplants,
etc., are not encouragedto leavethe best and otherwisethe most expensive locations, so
theyhaveretained them. Furthermore, theurban land use feeisalso calculatedaccording
tofloor space, notground (land) space.Suchamethodoffeecalculationimplies thatthere
isnoincentive for the physically efficientuse ofurban land (ground) becausethe average
cost (the amountofthe feeper sq. met.) isconstant regardless ofthe floor to ground in
dex. This arrangementis the reason whythere are still worthless ground floor buildings
in the very center of Belgrade.
Economicinefficiencybytheactualoccupantsistheconsequenceofphysicalinefficiency.
Thereason Hesinthefactthatphysicallyinefficientusers ofspace, suchasthemanufactu
ring industries, the building industry, etc., thathaveretained theirlocations are economi
cally inefficienturbanland users. Ithasbeen demonstratedthat these industrieswhen10
catedinbigcitiesarerelativelyinefficientcomparedtothesameindustries locatedinsmall
11
cities. On the otherhand, serviceindustrieslocatedinbig citiesaremoreefficientcompa
red to thoselocated in small cities." Hence, the restructuring process that leads to increa
sing the economicefficiencyofacity isimpededbythe nonmarketurbanlanduse system.
Since the regulations governing the allocation of a site do not include any time limit
for land use, there is no solid basis for the land user's financial calculation, especially
the calculationofnetpresentvalues. Furthermore, sinceurbanlanduse fee can be chan
ged unilaterally at any time by the municipal authority's decision the developer/investor
cannotcontrolthe costswhichcomprisethe amountofthe urbanland servicingfee. This
implies that an investment project which is profitable according to the current fee level,
can become unprofitable according to sorne future fee level, EvidentIy such an arrange
ment discourages property investment.?
Moreover, such arrangementsforurbanlandallocationimplythatmunicipalauthorities
are empoweredtowithdrawthe righttouse alocationat anytime, meaningthatthe actual
usercan be evictedfromthe locationatanytime. Thisincreasesthe uncertaintyforpoten
tial users/investors which discourages property investment. On the other hand, a lower
levelofinvestmentdiminishesthe potentialfor economicrestructuringwhichleads toper
petuating the inefficient economic structure of cities.
As there is a monopoly of public urban landowership, it is not possible for potential
urban land investors to invest vast financial resources. This is why significant potential
investorssuchaspensionfunds, insurancecompaniesand otherfinancial institutionscan
not invest in land and real property, thus inducing a decline in property investment. Fur
thermore, the nonexistenceof an urban landmarketaccounts for the nonexistence of nu
merous financial mechanisms that are cornmon in free market systems (mortgage loans,
property bonds, property companies' shares, etc). Accordingly, the concentration of fi
nancial capital for funding real property investments has been slow and inefficient, Le.
capitalisnot mobile. This has resultedin aconstantshortageoffinancial resourceswhich
in turn reduces property investment activities.
The currenturban land use system (as well as the legal system) is rather vague and is
inconsistentwiththeurbanlanduse system inmarketeconomies. Hence, potentialforeign
property investors are amply discouragedto invest. This accounts for a further reduction
in property investment and a further delay in the internationalization of the economy.
Thus, itisevidentthatthe currenturbanlanduse system in Belgradeshowsavery poor
performance in the domainof economic efficiency.
Furthermore, the nonmarketurbanlanduse system has resultedinthe poorperforman
ce ofthe municipaland localauthorities' financing. As the system is nonmarketthereare •
no urbanlandmarketprices or marketrents. It is reasonableto expectthatmarketprices
and rents would be muchhigherthan the currenturbanlanddonationand urbanlanduse
fees. Hence, if there were an urban land market prices, municipal and local au
thority revenues from urban land sales and rents would bemuch higher.
As there is monopoly of public urban landownership there is no solidbasis for urban
land taxation. One could reasonably assume that municipal and local authority revenues
12
from thetaxationofprivately ownedurbanland wouldbeconsiderable. Hence, themono
polyofpublicurbanlandownershipandthenonexistenceofanurbanlandmarketandmarket
prices leads to a poor performance in the realm of the municipal and local authorities'
financing, as revenues are considerably smaller than what they could be potentially.
In addition, the municipal authorities' financing policyhas induced afurther reduction
in local revenues. FirstIy,feeshavenot been properly indexedto the high inflation rates.
For example, during 1987the averagemonthIy feeindexation was3%,while the average
monthly inflation rate was 10%, in real terms, revenues were significantIy reduced. Mo
reover, asthere werenoconstrains, urbanland users/occupants were encouragedtopost
ponepaymenttheurban landusefee.This wasacomponentofthepolicyaimed atdecrea
singthetaxburdenoftheenterprisesandthohelptheminthetroublesometimesofeconomic
crisis.
An «informal»practice existed of writing off the debts of socially owned firms where
the urban land feeswere concerned. This actually meant that these firms evadedfeepay
ments.SuchapolicywasappliedveryfrequentIyinthecaseofbiglossmakerswhichusua11y
employ ahuge labor force."All these elements haveled to afurther reduction inmunici
pal and local revenues. VeryfrequentIy this reduction is so vastthat the revenues are not
sufficient to finance land servicing and provide public urban utilities. These deficits are
dealt with by borrowing money from clearing (commercial) banks.
5. The effects of the nonmarket urban land use system on urban
development
Urban development in Belgrade is affected in sorne cases directIy, but in many more,
indirectIy astheresultofthenonmarketurbanlanduse system.Oneofthemostimportant
effects is that there is no conventional central business district in Belgrade. This is due
to the negligible urban land use fee. It has already been pointed out that such afee level
haspersistentIy encouraged thephysicallyandeconomicallyinefficient useofurban land.
Thus lowdensity housingstillexistintheverycenterofBelgrade. Asthere isnoincentive
fortenants torelocate (duetonegligiblehousing costs)itisnotpossible for.business(offi
ce space, retail, etc.)tobehighIyconcentratedinthe very center ofBelgrade. Hence, the
process of urban redevelopment is strongly discouraged.
As it has been pointed out, the urban land donation fee and the urban land use fee are
virtuallyidentical foralllocations. However,theurban land servicing feeequalstheactual
land servicing costs at a givenlocation. Hence in the case ofredevelopment all the costs
ofclearingthelocation shouldbepaid bythe investor/developer. These costs are highbe
cause appropriate accornmodation (new houses or flats) must be provided for all the te
nants located inthebuilding due fordemolition. Furthermore, the whole operationofde
molition, clearingthesiteandfindingappropriateaccornmodationfortenantsisexclusively
carried out by the municipal authority, so there can be no cost control by the investor.
Therefore, potential investors are discouraged to invest in redevelopment.
These obstacles toredevelopment haveresulted inthe concentrationofdevelopment on
the urban fringes and so Belgrade's urban development tends towards an urban spraw1.
13
Such apatternofurbandevelopmenthasbroughtonareductioninglobal urbandensities,
whichintumrequireinvestmentinlong urban infrastructurelines (roads, pipelines, etc).
Furthermore, the reductioninglobal densitiesgivesrise tothe need forvariousurbanuti
lities (public transportation, etc). Such apatternofinvestment has reducedthe economic
efficiency in providing all urban utilities.
An additional factor in such a tendency towardsurban sprawl is apersistentexcess de
mand for housing and office space. This excess demand is also one of the results of the
actual nonmarket urbanland use system. Namely, one may sayit is reasonable to expect
that market land rent would be much higher than the urba~ land use fee.
This means that the cost ofliving has been significantly reduced. According to the re
sults ofa theoretical considerationofinternal migrations,?these costs are a significant ge
neratorofmigrations. The lowerthecost ofliving inagivencity,thelargerthepopulation
migration will be towards that city. Hence the low cost ofliving in Belgrade.índucedby
the nonmarket system have significantly influenced a large migrational flow toward Bel
grade. Such a flow has created a high growth rate ofBelgrade's urban population, hence
the excess demand for housing and office space.
6. Conclusion
The nonmarketurbanland use systeminBelgradederives from the socialistnonmarket
economicsystem. The essenceofsuch asystemisthe monopolyofpublic ownershipover
assets and resources, and the administrative (nonmarket) allocation of these resources.
The institutions and mechanisms ofthe nonmarketurbanland use system are very dis
tinctive from the institutions ofamarketsystem. They correspondto administrative allo
cation and are based on the discriminatory rights ofthe municipal and local authorities.
There are no market prices or rents, rather unilaterally determined urban land fees.
The economiceffectsofsuch asystemarevery poor. The crucialeffectisthephysically
and economicallyinefficientuse ofurbanland. Suchinefficiency has asignificantimpact
on the general economic inefficiency ofa socialisteconomy. Additionally, the nonmarket
systemhas causeda significantreduction.inreal property investmentas well as in global
investment activity. This leads to adeteriorationofthe prospects for economic restructu
ring, particularly with respect to industries located in cities. The financial performance
of this system is also poor, as there is a perpetual deficit in the municipal budget.
The nonmarketurbanlanduse systemhashad adverse consequencesonurbandevelop
ment inBelgrade: urban developmenthas astrong tendency towardsurbansprawl andur
ban redevelopmentis neglected, which implies a further reduction in the urban land use
eficiency.Apropercentralbusinessdistricthas notbeen established. Due tothe relatively
lowcost ofliving, ahugeimmigrationhas taken place, which has inturnactivated ahigh
population growth rate, which led to the deterioration of the quality of life.
.The overallperformanceofthe nonmarketurbanlanduse systeminBelgradehasthere
fore been evaluated as very poor. This is the result of the essence of the current urban
14
landusesystem:themonopolyofthepublicurbanlandownershipandadministrative(non
market)allocationoftheland. Thatiswhysuchasystemcannotbesuccessfullymodified.
It should be replaced by a market urban land use system.
Notes
1 Afruitful analysis ofthe Yugoslaveconomic system,itsorigin and its distinctions compared totheSoviet
systemhas been given byProut (1985).
2 Thisisconsistentwiththepremisesofcommunistideology(basedonMarx'svaluetheory)thatonlyrevenues
based on the employment oflabor are sociallyjustified.
3 See Madzar (1990)and Milanovié (1991) forthe discussion.
4 AccordingtotheUrban Land Nationalization Act,althoughnationalizationofthevastmajority ofproperty
rights wasdone bya Federal decree in 1945.
5 ItcanbearguedthatfeeisessentiallyrentoHowever,itshouldbetakenintoaccountthatnofreewillofusers
hasbeenrevealed,nocontracthasbeenestablished,norentrevisionprocedurehasbeenagreedupon.Accor
dingly,this feecannot be labeled rento
6 For the empiricalevidence which support this thesis inthe case ofYugoslaviasee Begovié(1991a).
7 Itcanbearguedthatcyclicaltendenciesontherealproperty marketsinWestemcountries havethesameef
fects.However,itshouldbepointedoutthatmarketpricefluctuations canbepredicted, atleastinprincipal,
while the behavior ofthe stateis rather unpredictable.
8 This isconsistent with thejob preservation policy,very popular in all communistcountries.
9 For a good reviewofthese considerations see Greenwood(1985).
References
BALCHIN, P.N., KIEVE, 1.L. andBULL, G.H. 1988: UrbanLandEconomicsand Pu
blic Policy. London: Macmillan
BEGOVlé, B.1991a: TheEconomicApproach tothe OptimalCitySize, Progress inPlan
ning. Oxford: Pergamon Press
BEGOVlé, B. 1991b: The Reformofthe Urban Land Use Systemin Belgrade, Lisbon:
Paper presented at the 31st. RSA European Congress
BEGOVlé, B. and MIJATOVlé, B. 1992:The Introduction ofthe Market in the Urban
Land Use System. Disp, Vol.28: 3-17
Departament ofEnvironment. 1988: UrbanLandMarkets in the United Kingdom. Lon
don: Her Majesty's Stationary Office
GREENWOOD.M. 1.1985:HumanMigration: Theory, Models, and Empirical Studies,
Journal ofRegional Science, Vol.25: 521-544
KARDELJ,E. 1971: Slobodan udruzeni rad(FreeAssociatedLabor). Beograd:Radnicka
stampa
MADZAR, Lj. 1990: Suton socijalistickihprivreda(The Dusk ofSocialistEconomies).
Beograd: Ekonomika
MILANOVlé, B. 1991:Prikaz knjige Suton socijalistickih privreda (The Book Review
of The Dusk of Socialist Economies), Ekonomska analiza, Vol. 25: 431-439
OECD, Project Group on Urban Markets. 1990:FinalReporton Urban Markets. Paris:
OECD
PROUT, C. 1985: Market Socialism in Yugoslavia, New York:Oxford University Press
15