one lump sum (a âspecialâ dividend equal to the present value of all future ... for rapid housing privatization concerns the distribution of the dividend from the ...
TWURD WP #12
URBAN DEVELOPMENT DIVISION TRANSPORTATION, WATER AND URBAN DEVELOPMENT DEPARTMENT ENVIRONMENTALLY SUSTAINABLE DEVELOPMENT THE WORLD BANK
RAPID HOUSING PRIVATIZATION IN REFORMING ECONOMIES: PAY THE SPECIAL DIVIDEND NOW
Robert M. Buckley Patric H. Hendershott Kevin Villani
February 1994
WORKING PAPER
The TWURD Working Papers present preliminary research findings and are intended for internal review and discussion. The views and interpretations herein expressed are those of the authors and should not be attributed to the World Bank, to its affiliated organizations or to any individual acting on their behalf.
Copyright 1993 The World Bank Washington, D.C.
AJl Rights Reserved October 1993
This is a document published informally by the World Bank. In order that the information contained in it can be presented with the least possible delay, the typescript has not been prepared in accordance with the procedures appropriate to formahy printed texts, and the World Bank acceptsno responsibility for errors. The World Bank does not accept responsibility for the views herein, which are those of the author and should not be attributed to the World Bank or its affiliated organizations.The fmdiigs, interpretations, and conclusions are results of research supported by the Bank; they do not necessarilyrepresentofficial policy of the Bank. The designation employed, the presentation of material, and any maps used in the document are purely for the convenience of the reader and do not imply the expression of any opinion whatsoever on the part of the World Bank or its affiliate concerning the legal status of any country, territory, city, area, or of its authorities, or concerning the delimitations of its boundaries or national affiliation. This paper was prepared by Robert M. Buckley, a Principal Economist at the World Bank, Patric H. Hendershott, a Professor at the Ohio State University and Kevin Viiani, a Professor at University of Southern California. This paper was prepared as a joint USAID/World Bank project to develop an Urban Action Plan for Budapest and was presented at the First AREUEA International Conference in Los Angeles, the BudapestMeeting of the European Network for Housing Research, and at a seminar at the Royal Institute of Technology in Stockholm. Discussionswith Hans Wijkander and others at RIT were especially useful. Comments by Esra Bennathan, Alain Bertaud, Peter Feider, Donald Ham-in, Michael Helter, James Follain, David Oliiger, Bertrand Renaud, Jeff Telgarsky, Margret ‘lhalwitz, and Barry Voret were also helpful.
THE WORLD BANK
Rapid Housing Privatization In Reforming Economies: Pay The Special Dividend Now
Robert M. Buckley Patric H. Hendershott Kevin Villani
WORKING PAPER
TABLE OF CONTENTS I. INTRODUCTION AND OVERVIEW
1
II. TEl.EREAL ESTATE SECTOR IN MARKET ECONOMIES A. Long-Run Equilibrium B. Adjustment to Equilibrium
5 5 6
III.
THE REAL ESTATE SECTOR IN CENTRALLY PLANNED ECONOMIES A. The Housing Sectorof a PCPE B. Over or Under Investmentin Housing?
IV. TIUNSITION TO A MARKET ECONOMY A. A Glacially GradualTransition B. A Rapid Transition C. Equity Concernsand Political Objections
9 9 12 15 15 17 19
v. SUMMARY
21
REFERENCES
22
Table 1: The Ratio of Real EstateValue to ReplacementCost Figure 1: The Rental Market in Market Economies
25 8
Figure 2: The Rental Market in Controlled Economies
11
Figure 3: The Rental Market After 45 Yearsof Control
14
I. INTRODUCTION
AND OVERVIEW
The benefitsof marketeconomiesover centrally plannedeconomiesstemfrom economic agentsfreely respondingto pricesthat better reflect real socialcosts. The liberalizationof prices to reflect true economicscarcityshouldthus be a top priority for policymakersin the previously centrally plannedeconomies(PCPEs).’ Liberalizingrents andprivatizing housingought be the top priority in the PCPEsfor two reasons. First, rents were the most distortedof all prices, owing to the statusof housingas a social (as opposedto economic) good.* Moving this price to market will thus generate substantialefficiency gains in reallocatingthe existing stock [Daniel and Semjen(1987) and Tolley (1991)]. Second,well-functioninghousingmarketsare an essentialpreconditionfor the restructuring of the businesssectorto a market economy. Labor market efficiency is enhanced through vastly improved householdmobility (Blanchard etal., 1991), and the efficiency of capital accumulationandallocationis improvedby funding entrepreneurialeffort via borrowing againsthomeownerequity (Newberry, 1992). Conversion of the housing sector is said to be high on the list of reform minded governments,as well as internationalagencies. 3 Nonetheless,there has beenlittle progressto date, andprivatization initiatives appearto be losing steam(seeKalinina, 1992, on Russia,and Kingsley and Struyk, 1992, on EasternEurope). Indeed, some countries, suchas China, are planningon phasingin housingreforms over severaldecades.4 Blanchardet& (1991, pp 36-39) make the casefor rapid privatization of state assets
l This priority of price reform is emphasized in Calvo and Frenkel (1991), Blanchard et.al. (1991) and McKinnon (1991). Many of the conclusions in this paper follow from the transition models presented in these works. 2 Renaud (1991) shows that in the late 1980s the simple average gross rent-to-income ratio for Bulgaria, China, Hungary, Poland, Romania, Yugoslavia, and the USSR was less than 3 percent. Similar figures hold for Czechoslovakia (Kingsley and Struyk, 1992). Administered rent levels are even lower when measured as a fraction of economic rent because income in the PCPEs was also severely repressed. In market economies, in contrast, the average rent to income ratio generally exceeds 20 percent (Mayo a, 1992). 3 The Russian Republic has identified housing as one of its most foremost reform areas, and in 1991 adopted a law “On Privatization of the REFSR Housing Stock” (Kalinina, 1992). One of Poland’s first loans from the World Bank was a $200 million housing loan; in both 1988 and 1991 the Chinese embarked on a major housing reform effort (World Bank, 1991); and a number of World Bank and USAID studies of housing policy in Eastern European countries have pointed out the broader problems posed by socialist housing delivery systems and the need to reform these systems. 4 China’s 1991 plans are based on a premise that it will take until the year 2015 to develop a housing market and that privatization will not occur on a significant scale for at least 10 years.
2 throughgiveaways,whereasMcKinnon (1991)presentsthe casefor “going s~ow..“~The stakes regardingthis policy choice are high. Housing markets will not function properly until the transition to a free market is essentiallycomplete; the go-slow approachperpetuateshousing market distortions with few offsetting benefits and underminesthe entire reform agendaof convertingeconomies.Moreover, the distortionsintroducedby the phasingin of partial reforms can be destabilizing.6 The argumentsagainst rapid privatization generally are analyzedin Blanchard,&al, (1991). We addressthose argumentsin the context of housing reforms. Our conclusionis identical:the arguments“are ill-thought out. ” The first argumentis that existing housingstock is unaffordablein private ownership. The secondis that governmentscannotafford to forego the revenuefrom the housingstock. Both are basedon a partial analysisof how centralplanning distortedrents, ignoring its simultaneousdistortionsto householdincome and wealth. In marketeconomies,housingaffordability argumentsalwaysreduceto somehouseholds havingrelatively lessincomethan others; hence,housingsubsidiesare introducedas an in-kind transfer. This household-specificargument cannot be extendedto the total populationof an economyvis-a-visits existing housingstock. While enormoussubsidiesare providedto renters in the PCPEs,the full cost of thesesubsidiesis currently being paid by the existingpopulation. Conceptually,freeingrents while simultaneouslyreturning the incomefunding current subsidies would make private market rents as affordable in the aggregateas the pre-reform subsidized rents. Moreover, the overwhelmingefficiency gains make it relatively easy to developgiveaway schemesin which everybodyis unambiguouslybetter off, mitigating equity considerations relative to the statusquo. The secondargumentis that governmentsneedto retain the housingstockbecauseit will generateeither an amplestreamof revenuesor a large lump sum when sold at “market price~.“~ While we sympathizewith the need for governmentsto reduce fiscal deficits, retaining the housingstock will not do so; as shown by Renaud (1991), the current operation of public housinglosesmoneyin all the PCPEs. Giving such “assets”away would increasegovernment revenuesevenif no taxeson private housingreturns or capital were collected. This conclusion
’ Actually, Blanchard et. al. (1991) argue for rapid housing distribution on page 29, but for administered increases in rents and thus housing prices prior to distributing the housing stock on page 50. They are unambiguously in favor of rapid distribution of industrial assets. 6 Calvo and Frenkel(l991) conclude that “the benefits from administratively spreading the price rise over time may be illusory” (p 21), and that “the more effective is the domestic safety net designed to protect wage-earners, the more likely it is that more economic problems (inflation) ensue” (p 19). ’ This is a general theme of McKinnon (1991): “Indeed, the tendency of the fiscal position of the government to deteriorate because of the liberalization itself militates against any ‘giveaways’ of industrial assets or the housing w (p 148, italics added). Katsura and Struyk (1991) specifically state that “these @lousing] assets are so valuable that if sold at or near full value, they could provide the state with substantial funds to help during the economic transition.” @ 1)
3 is only altered slightly by proposedrent increases,which in most caseswould increasethe present value of net rents to less than twenty percent of reproduction cost, or less than the present value of the tax revenuethe governmentwould be expectedto collect from private owners. Neither is the conclusionaltered fundamentallyby apparentsalesof somepropertiesat significantlyabovefundamentalvalue. Suchsaleswould either shrink the domesticallyoccupied housing stock or fuel inflation. Moreover, the resultant perceptionof unaffordablerents is largely responsiblefor the heightenedconcernof the populationregardingeviction, foreclosure and tenancyrights that is stifling seriousreform. Preventingsuchsalesis thus anotherargument for, not against,housinggiveaways. The persistentmisconceptionthat the housingstock has large fundamentalvalue relates to the gap betweenthe level of subsidizedadministeredrents and of long-run equilibrium free market rents. Of course, if rents were at unsubsidizedfree market levels, fundamentalvalue would be substantial. But rents cannot get to such levels unlessthe income funding current subsidiesis returned to the population. As a result, all analysesof reform haverecommended rent and wage increasesas a necessaryfirst step. At this point we needto emphasisthe m sourcesof subsidiesin existingrents and the reforms necessaryto restore demandto free market levels. As is well known (Komai, 1990, andBlanchardet.al., 1991), socialisteconomiesgenerallyrepressedwagesandimplicitly “taxed” businessincome at rates of 100 percent or more. This amountsto an incometax on wages.8 The current implicit wage income tax in the PCPEsfunds numeroussubsidies,including the ongoingoperatingsubsidiesfor housing.g While rents have risen sufficiently to about cover operating costs in a number of countries(Kingsleyand Struyk, 1992), the gap betweenlong-run equilibrium and administered rents is still huge. This gap representsthe secondsubsidy, the return on capital from past housinginvestments. Householdsfinancedcapital subsidiesfor past investmentswith “forced savings” through the wage repressionmechanismjust described. Rents can be raised to free market levelsonly if the “dividends” from investmentin the existinghousingstockare restored to householdincome. The restoration shouldnot be provided through wage reform (be tied to hours worked) becausedoing so would severelydistort labor marketsandreduceinternationalcompetitiveness. Rather, the dividend income shouldbe distributedeither as an ongoingrental allowanceor in one lump sum (a “special” dividend equalto the presentvalue of all future dividends), which is fiscally equivalentto giving away the existing housing stock. Identification of the state’s
* See Pogodzinski (1991) for a discussion of labor supply effects of this wage repression. ’ The implicit tax on wages to finance the rent subsidies and the need for wage reform are discussed in Tolley (1991) and Tami (1991).
4 housingdividend as the major source of current subsidizedrent levels explainsthe seeming dichotomybetweenthe low fundamental(rental) valueof public housingandthe perceivedhigh market value (which includesthe presentvalue of future dividends). While the dividendcouldbe paid either way, we strongly advocatethe one-time special dividend. As Olsen (1983) argues, the fall of communismoffers policymakersa unique but fleeting opportunity to break the existinginstitutionalpolitical obstaclesby irretrievably giving away the housingstock. A continuingdividendboth keepsthe housingstockin the handsof the governmentand is all too easilycancelableat somefuture date, when rent controls could all too easilybe reintroduced. Restoring effective demandis not the only conditionnecessaryfor robustprivate housing markets. As Russianpolicymakerslearned,liberalizingpriceswhile productionis still controlled by statemonopoliesdoesnot necessarilyproducea supplyresponse. The primary focus of this paper, however, is the needto restore housingmarket demand,assumingthat the supply side reforms describedin Renaud(1991) are implemented. More specifically,the central question for rapid housing privatization concernsthe distribution of the dividend from the existing housingstock, rather than the much thornier wage and rent level questionsthat have beenthe emphasisof all other studies. From this it follows that policy shouldfocus on developingan equitableformula (e.g., homeownershipvouchers)for giving away the vast majority of the housingstock, while implementingall supplyreforms (breakingup inefficient statemonopolies, developingproperty rights, etc.). The remainderof this paperis divided into four parts. Part II describesthe working of real estatemarketsin market economies.Part III identifiesthe distortionsintroducedby central planning,and Part IV discussesalternativetransitionsto a free real estatemarket. The final part summarizesthe paper.
Il. REAL ESTATE SECTOR IN MARKET ECONOMIES A discussionof real estatemarkets in market economiesis best presentedin two parts: the long-run equilibrium and the adjustmentfrom a disequilibriumto the equilibrium. A. Long-Run Equilibrium
In full equilibrium, net (of operatingexpenses)rent equalsthe rental or user cost (UC), the quantity of spaceequalsthe demandfor it, and the value of the spaceequalsits replacement cost. Measuringthe user cost as the sum of the fmancingcost (real after-tax financingrate) and depreciation,grossedup by unity less the tax rate, after-tax net (of operatingexpenses)rents must cover the financing cost and depreciation. Solvingthis relation for the pretax grossrental rate: Pr/RC = (fin+depr)/(l-7) + oper/RC,
(1)
where Pr is the price of (grossrent on) a unit of space,RC is the replacementcost of a unit of space,tin is the real after-tax financing rate, depr is the depreciationrate, oper is operating expenses(utilities and maintenance)per unit space,and7 is the rate at which rents lessoperating expensesare taxed. In general,value equalsthe presentvalue of expectedfuture after-tax net rental income. But in equilibrium, where thesefuture rents are expectedto equalexpectedfuture user costs, new constructionexactly earnsnormal profits and the real assetvalue of a unit of spaceequals its replacementcost: V = RC.
(2)
In a well-functioningmarket economy,demandand supplyareequalbecauseprice adjusts to bring demand into line with the existing stock. We expressthe demandfor spaceas a positive function of income (Y) and a negativefunction of the price (rent) of spacerelative to the price of other goods (Pr/Po) and equatedemandto the existing supply or Q: D = D(Y, Pr/RCPo) = Q.
(3)
We abstracthere from vacancyrates. With vacancies,we would convert Pr to an effective rent and Q to an effective quantity (would multiply both by one less the vacancyrate), and, in disequilibrium, we would allow for differences between the actual and “natural” vacancy ElkS.‘”
lo For an analysis incorporating vacancy rates, see Hendershott and Kane (1992b). For a discussion of the “natural” vacancy rate (the rate that maximizes profits for owners of space), see Hendershott and Haurin (1988).
6 Figure 1 describesthe rental market for space. The solid schedulesdescribe full equilibrium; the quantity of spaceis suchthat demandand supply are equatedat (Pr-oper)/RC = UC(in which caseV =RC). Excessdemandis illustrated by the dasheddemandschedule, where Y’ > Y and thus Pr’ >Pr. In this situation, value exceedsreplacementcost and new constructionwill be induced until Q rises to Q’ and full equilibrium again exists, as we shall now explain. B. Adjustment to Equilibrium Three more equationsare neededto capturethe essenceof the adjustmentprocess. First, we expressnet construction(C) as a positive function for the value/replacement-cost ratio: C = C(V/RC).
(4)
With V/RC exceedingunity, new spacecanbe soldfor more than the costof producingit. Thus constructionin excessof depreciationwill occur. With V/RC lessthan unity, constructionwill be less than depreciation. In fact, if the value of spacein this use is below the value in an alternativeuseby more than the cost of conversion,conversionto the alternativeuse will occur. To capturethe impacts of net constructionand net conversions(CONV) out of the stock, we expressthe stock in period t in terms of the previous period stock and net construction and conversions(both betweent-l and t): Q, = Qtwl+ C - CON-V. The last, andprobably mostimportant relationshipfor what follows, is the determination of fundamentalvalue, the presentvalueof expectedfuture net real rentalincome. Assumingthat the operatingexpenseratio, tax rate, depreciationrate and financing rate are constant, v = $
(lwT)
(prt
- oper) (l-&p)
(l+fin)t
t-1
. (6)
Equation (6) illustrates three fundamentalpoints. First, and most obvious, valuation requires market participants to make long term forecastsof real rents.” Second,if rents only cover operatingcosts (and are expectedto do so “forever”), then value is zero. Third, if the market is in full equilibrium, market rents equal their long-run equilibrium value [(Pr-oper)/RC = (fin+depr)l(l-T)] and V = RC. Equation (6) is usefully rewritten as
I’ Generally, real rents are specified as tijusting to discrepancies between natural and observed vacancy rates. For estimates in which real office market rents adjust to discrepancies between both natural and observed vacancy rates and equilibrium and observed real levels, see Hendershott (1993).
7
v = (1 - PVBERI)RC,
(W
where PVBERI is the presentvalue of expectedbelow equilibrium rental income mr53BRX = c t-1
(1-t)
[UCRC - (Prt-oper) (l+fin)t
I (l-depr)
t-1
.
(1
Numerous “disturbances” can changefundamentalvalue: changesin UC(m fin, T or oper), in RC (real labor costs, production technologies)or in the expectedpath of the Pr’s (owing, possibly, to changesin governmentpolicies). As a general rule, we would expect changesin thesevariables over any short period to be small (costs and technologiesgenerally changeslowly) or to be temporary in nature(real after-tax interestrates rising and then cycling back down). Combinedwith a fairly rapid productionresponse,reaI value would not fluctuate sharply. The 1980sprovided an exceptionto this rule in the United States. Owing to a breakdown in prudentlendingpractices, an unprecedented commercialbuilding boom occurred. Real office market rents were halved, andreal fundamentalvaluefell by over a third relative to replacement cost (Hendershott and Kane, 1992b). This “lending-frenzy” is estimated to have reduced commercialreal estatevaluesby $250 to $300 billion dollars and to have wastedabout $125 billion of economicresources(HendershottandKane, 1992a). The frenzy is largely attributable to mistakengovernmentregulatory and private sector lendingpractices. The recentU.S. experienceshouldbe a warning to countriesreforming their real estate sectors. Evenin a purportedly stablemarket economy,lendingand building errors of enormous magnitudecan occur. The potential for such errors in economieswith highly distorted real estatemarketsmustbe far greater. And with tenuousmacroeconomicstability, the socialas well as economiccosts of sucherrors would be multiplied many times over.
co
PI’ Pi
l*
0
‘* 0. 0. l* ‘. ‘* l b l. . . . . -. . . . . . . . . . . .c
Q
Q’
Figure 1 Rental Market in a Market Economy Real Rent
b.....................,.....................
‘1
Qumbtityof Space
IIL
THE REAL ESTATE SECTOR IN CENTRALLY PLANNED ECONOMIES
Real estatesectorsin centrally-plannedeconomiesfunctionedfar differently than those in market economies. Constructionwas subsidized,interest rate and rent subsidieswere given for owners andrenters, respectively,and wageswere repressed(“taxed”) to fund the subsidies. We beginby describinghow thesepoliciesdistort marketsand then discusswhether the PCPEs have over or under investedin housingduring the last 45 years. A. The Controlled Housing Sector of a PCPE Construction costs in the PCPEs were previously directly lowered by four “capital” subsidies: land was provided for residentialuse by the state at little or no cost; the price of labor and materialscontainedsubstantialsubsidies;direct grants from the state budget were provided to finance construction of both owner-occupied housing and state cooperative developments;and long-term mortgage financingfor cooperativeand owner-occupiedhousing was provided at below market rates. There were also ongoing “operating” subsidies: utility rates (central heating, electricity, water and sewer, gas, and trash collection) were set below economic scarcity value, and state-ownedcompaniestypically provided upkeep and maintenanceof the housingstock at administeredprices.‘* The PCPEsraisedrevenuesby suppressingwageswell below productivity and “taxing” (confiscating)enterpriseprofits. Thetax revenuesfinancingoperatingsubsidiescanbe described as “transfers,” whereasthe tax revenuessubsidizingcapital represent“investments”in the stateowned housing stock. The tax was broadly based, but housing investmentswere made in discreteintervals. At any point in time, the tax on householdincome to fund capital subsidies was the implicit failure to pay “market” dividends on the existing housing stock (i.e., the implicit tax rate on housingdividendswas 100%). With 01andB, respectively,of replacement-costcapitaland current operatingcostsbeing subsidized,the total income suppression(SUP) necessaryto fund all housingsubsidiestoday is: SUP = (Pr - pr’)Q = [crfinRCl(l-r) + /3oper]Q,
(7)
where Pr-m, the differencebetweenmarket and controlledrents, is the rental subsidyper unit housingandQ is the total quantity of (stateowned)housingspace. The current dividendincome missing from householdincome is cr[finRCl(l-r)]Q, and the wage tax is BoperQ. Current householdincomeis reducedby both the lack of dividendsand the wage suppression. The present value of the expected missing after-tax dividend income is the wealth
I2 See Renaud (1991) for a more complete description of these subsidies and their relative use, e.g., among public housing, cooperatives and private dwellings.
10 (homeowner equity) missing from PCPE householdbalance sheets.13 To understandhow crucial homeowner equity generally is to the “affordability” of market rents, consider the following. During the periodof high interestratesin the U.S. in the late 197Os,housingmarket analystsconcludedthat, based on current prices and interest rates and ignoring substantial homeownerequity, only aboutten percentof homeownerscould afford the housesthey lived in. This percentagewould likely be evenlower at the higher real interestratesandproductioncosts characterizingthe PCPEs. That is, without homeowner equity, the vast majority of PCPE householdswill not be able to afford their current housing. Figure 2 illustrates how a real estatemarket so controlled would deviatefrom a free market. With low rents, excessdemandand a queueresult: the income suppressionreduces the demandfor all goods and services,but all of the income is usedto fund rent reductions. The queuesare perhapsthe most damagingconsequenceof the controlled market. Households are assigneda particular unit and are effectively prohibited from moving. That is, the existing housing stock is inefficiently allocatedamong households,and householdsare suboptimally distributed geographicallyrelative to their job locations. Moreover, effect zero household mobility greatly underminesattemptsat industrial restructuring.14 Thesesubsidiesresult in hugedifferencesbetweenrents for controlled (residential)and uncontrolled (commercial) real estate. The result is the developmentof “gray markets” for illegal conversionsto alternativeuses. Thus the distortionsin residentialmarketsspill over into other real estatemarkets. Housing market distortions also result in hoarding and an inefficient allocation of housing. Roth Daniel and Semijen(1987), for Hungary, and Tolley (1991), for China, show that the gains from permitting a market to reshuffle the entitlementsto the existing stock of housingare equivalentto an increasein housing servicesgeneratedby a number of years of production. Thesepotentiallylarge efficiency gains,however, require marketpricesto facilitate reallocationof the stock among households,to encouragerenovation activity, and to provide proper signalsfor new construction. The involuntary taxes to finance housing demandplacedthe statein the position of a price-insensitivemonopoly developer. Kingsley and Maxiam (1992) estimatethat real supply costsin Czechoslovakiaexceedcompetitivesupplycostsby 40 percent. We suspectevengreater resourcemisusein the former Soviet Republic, where price distortionswere larger.
I3 If expected future values of 01 and fiu equal current values, the missing wealth is arRC*Q. I4 Renaud (1991) fmds household mobility in reforming economies to be less than 10% of market economies. Again, see Blanchard et.al. (1991) for a discussion of the potential costs of this impediment to restructuring.
.
Awouoq paljoJluo3 e UI.
(0d’A)Cl
.
12 B. Over or Under Investment in Housing? The conventionalwisdom is that the cumulative effect of a half century or more of planningand administrativeallocationof housinghas resultedin substantialunder-investment in housing(Renaud, 1991). If this were the case,then demand,after the return of the housing stock to households,would exceedthe existing stock at Pr = UC, and market rents would rise aboveUC. Value would exceedreplacementcost, and new constructionwould be forthcoming, assuminga liberalized housing supply sector. Becausethis responsewould take time, the economywould be at a point suchas A in Figure 3. On the other hand, while shortagesare expectedat the low administeredrent, increases in rent to the long-run equilibrium level could easily createan excesssupply. Recall that the shareof investmentdemandallocatedto &l consumergoodswas low in most PCPEsrelative to market economies. That is, after removal of rent subsidiestheseeconomiescould be at point B in Figure 3, and little constructionwould be expectedfor many years. We would anticipatethat somelocal marketswould be representedby point A and others by point B. However, several reasonslead one to doubt that the pre-reform state of disequilibriumis generally one of shortage. First, data on the PCPE sharesof investmentin housingare difficult to sort out and comparewith thosefrom market economies. Prell(1989), for example,arguesthat the post WWII growth of housingproduction in the SovietUnion was “considerable”and almost twice as high as estimatesmadein the West. Similarly, Alexeev et-al. (1991) show that during the 1957-64period Soviethousingproductionper capitawas the highestin the world, and the World Bank study of housingpolicy in Hungary indicatesthat during the 1970’sHungary was probablythe world’s leadingproducerof housingon a per capita basis. Finally, Goldsmith’s (1985) dataindicatesthat Sovietand Hungarian levels of housing assets-- the only two PCPEsfor whom there are suchdata -- as a shareof wealth are higher than many market economies. Second,while the shareof savingsdirectedto nonresidentialinvestmentsmay havebeen higher than for market economies(seeKomai, 1986), the value of theseinvestmentsprobably was not (seeTanzi, 1991). World Bank estimatesindicate that averagereturns on industrial investmentfor PCPEsover the last severaldecadeswere negative(Bleaney,1988). Discounting the existingcapital stock at a rate reflectingthe real marginalproductivity of new capital would likely suggestthat housingrepresentsa proportionatelygreatershareof wealth for PCPEsthan for most market economies. Third, even if housingis in short supply, there may well be a surplusof the generally dislikedpanelconstructionflats. The removalof all subsidiesand a simultaneousboostin aftertax incomes(and housingwealth) could leadto an increasein the demandfor, and thus supply of, “luxury” housingand a permanentsurplusof panelflats. We suspectthat a substantialamountof renovation,repair and conversionwork will meet
13 the hurdle rate for new capital. But the shareof GDP devoted to new housing construction projects that are viable at a 15% real cost of capitalwill probably be quite small, and the share devotedto panel constructionproducedby statemonopolieswill be negligible.”
..’
Is McKinnon (199 1, pp37 and 52) suggests that the hurdle rate for new investments in the converting economies should be near this value.
08 ‘.
-9
l.
‘.
‘*
‘*
Figure 3
-0
‘.
‘0
*a
: : : : : : : :
‘a l *.: i”=.., *a . ‘. ‘. ‘0 i l. i
8 \
I* ..*.
l ‘a
UC WY,Po)
D(Y-SUP,Po)
6
Rental Market After 45 Years of Control Real Rent
*.
I-*.~***~~~...*.....,.,.,.,,.......................,......*.....,........~..............,~,
t
i
IV. TRANSITION TO A MARKET ECONOMY The needto convert the housingsectoris universallyrecognized. Disagreementsarise regarding the optimal pattern and speedof adjustmentand the priority of housingover other reforms. While someadvocategoing slow on housingreform, we, like Tolley (1991), contend that speedis of the essence.16Our reasonsare three. First, many of the gains from reform accrueonly whenreform is essentiallycomplete. Second,gradualreform is likely not politicaIly feasible. To illustrate, during an electioncampaignin the summerof 1993,Hungary, which has beentrying to raiserents to their full-equilibriumfor decades,froze nominalrents, meaningthat real rents would decline. Third, the major argumentsagainstrapid reform are false: privatizing the housingstock will improve, not worsen,the fiscal positionof governments,and the housing stock can be privatized in an affordable and equitablemanner. This sectionaddressestheseargumentsdirectly. We begin with an analysisof a very gradualtransitionto a free housingmarket, includingthe implicationsfor real estatevalues,and concludethat deeplysubsidizedhousinghasvirtually zero value in the government’shands. We then describethe meansto a rapid transition, and last we discussequity concernsandpolitical objections. A. A Glacially Gradual Transition Supposethat rents are raised, but not even sufficiently to clear the market. Then virtu&y noneof the benefitsof housingreform are gained. Hoardingof large flats will continue (no efficiency gainsfrom reallocationof the existingstock will occur), and labor force mobility wiII not increase. Achieving market clearingrents for the vast majority of marketsis a minimal initial step. Referring back to Figure 2, the housing queuewould be eliminated, and some governmentsavingswould be achievedas rents rose from Pf to Pf’. Generally, though, the rent increaseand budget savingswill be small becausein most reforming economiesfew householdscan afford much higher rents. The fundamentalproblem is, of course, the income (and wealth) suppression. Until after-tax incomesor wealth are raised, few can afford to pay full-equilibrium rents. How long will it take to get to full-equilibrium rents? Considertwo cases. In the first, which approximatesChina, real rents are only one-tenthof full-equilibrium, and they will be raised by 10 percent a year. The adjustmenttakes25 to 27 years, dependingon the level of financingrates. In the second,which appearsto approximateRomania,real rents are 25 percent of full-equilibrium (operating costs are covered), but real rents will be increasedat only 5 percent a year. The adjustmentis still 25 to 29 years. And just what is the housing stock currently worth under these scenarios? We can
I6 Stiglitz (1991, p 15) argues that state “commitments” to competition and against subsidization are the primary economic advantage of privatization. See also Sappington and Stiglitz (1987).
16 answerthis questionby usingthe valuationequations(6a) and (6b), in which the ratio of value to replacementcost is relatedto the presentvalue of below equilibrium rents. Table 1 reports this ratio for the two slow transitionsdescribedabove(real rents starting at 10 percentof market or coveringoperatingexpenses,respectively). The calculationsare performed for two financing costs(high and low). In the first, fin = 0.12, 7 = 0.2, oper = 0.06 (0.03 eachfor maintenance and utilities), and depr = 0.024. Thus, from equation(l), UC = 0.24 and rentsjust covering operatingcosts initially equal a quarter of equilibrium rents (.06/.24). In the second,fm = 0.08, and thus UC = 0.19 and initial rents are about a third of equilibrium (0.06/0.19). The valuesin the table are ratios to replacementcost, and thosein the first row (RC = 1) are basedon the further assumptionthat real constructioncosts have not changedsince the housingwas built. As canbe seen,valuesare negativeif real gross rents start out at only onetenth of equilibrium and rise at 10percentper year. If rents initially cover operatingcostsand are expectedto rise at 5 percentper annumin real terms, then valueis positive, but equalsonly 6 percent of (depreciated)replacementcost in what we considerto be the more realistic high interest rate/user cost case. In the low real interest rate scenario,value is almosta quarter of replacementcost. The assumptionof unchangedreplacementcost seemsinappropriate, Estimates by Kingsley and Maxiam (1991) suggestthat the presubsidyproduction costsof stateconstruction companiesmay be 40% greater than thoseof their private market counterparts. If this is true, the developmentof private constructioncompanieswill reducereplacementcost by 30 percent (0.4/l .4). In this case (row 2)) value would be only 4 to 16 percent of original (depreciated) replacementcost, even when real rents initially cover operatingcosts and are expectedto rise at 5 percentper year. The values in Table 1 may seemlow to someand certainly appearat odds with casual observationof isolatedhigh salesprices. The questionis whether suchobservationssupport a go-slow privatization approach. We think not, as we explain by consideringspecific types of and motivationsfor sales. One reasonfor high prices in somedesirablemarketsis purchasesby foreigners. They are willing to pay close to reproduction-costprices becausethey (1) have income and wealth (their income has not been suppressed)and (2) are not eligible for domestic rent subsidies. While such sales raise revenue, they also reduce the housing stock available for citizens. Whether the net deficit reduction -- the declining availablestock will increasepressure for subsidizedproduction -- is worth the net reduction in availablehousingis unclear. Another reasonfor high pricesis the conversionto other uses. Commercialrents in most PCPEsare unregulatedwhile supply has typically beenrestricted. Theserents are thus often above,not below, equilibrium levels. With suchrents, valuesabovereplacementcost are to be expected,and conversionsfrom residential,thoughtypically illegal, will follow. Here again, suchconversionsreduce the supplyof housingfor citizens.
17 Considerablecashbalances(“liquidity overhang”)havebeenaccumulatedby someowing to the lack of availableconsumergoods, and thosewith this cash might well be willing to pay abovefundamentalvalue for units continuingin housinguse. Macroeconomicpolicy calls for a 100 percent tax on such balancesto reduce inflationary pressures, which is one reason householdswith balancesare willing to overpayfor housingunits. Salesby local governments at above-fundamental-value prices are one way to imposethis tax, but the proceedsmust then be taxed at 100% by the central bank if inflation is to be controlled.17If this secondtaxation is not possible,housinggiveawaysare necessaryto prevent inflation. Moreover, distributing the housingstock to holdersof liquidity overhangis generallyconsideredinequitable,due to the prior inequitabledistribution of the overha.ng.18 Finally, governmentsmight be ableto generatesalesprices abovefundamentalvalue by providing excesscredit. Not only is this inflationary, but the immediaterevenuegain is likely to be offset by default lossesin the future. lg Housing giveawaysalso avoid this problem. B. A Rapid Transition Two mechanismsfor a rapid transition to a free-market housing systemare an income supplement(rental housing allowances, such as those being employed in the former East Germany) and a wealth transfer (home ownership vouchers, such as those contemplatedin Russia). Annual housingallowanceswould provide housingsubsidiesto the populationthat, in the aggregate,equal the foregone annual regular dividend on the rental stock. Ownership voucherswould provide a one-timespecialdividendto householdsequalto the existingphysical stock of governmentresidentialcapital.*O Roth mechanismscanbe structuredto reflect desired equity considerations. While both mechanismswould entail substantialadministrative, managerial,and other difficulties, we believe the problems of the regular dividend mechanismto be more severe.21 l’ Calvo and Frenkel(l99 1) note that ” . . .sales by state enterprises generate (non recurring) revenue (however such a revenue cannot be spent for, otherwise, a new liquidity overhang would be generated).” (p. 34, emphasis in original). Salimano (1990) also concludes that “sterilization of the proceeds of the auctions of state-owned assets is required” (Pp. 14-15). ‘* Blanchard et.al. (1991, p. 34) note “the two domestic groups that are in a position to acquire a disproportionate fraction of the capital are the ‘nomenklatura’ and those who have become rich from black market activities. ” I9 See Buckley, Hendershott, and Villani (1993) for a discussion of the likely moral hazard and adverse selection problems associated with lending in the PCPEs. p Rental housing serv es an important role in market economies. Most important, renters are more mobile than owners. Moreover, rental housing management is more effective for dense multi-family housing units, such as those characterizmg urban PCPE housing. Privatization schemes should thus encourage active rental markets. 2* Kornia (1990) discusses problems involved with transferring the ownership of a large part of the capital stock.
18 First, becausethere are currently no explicit housing dividends, the desired/neededlevel of housingallowancesis unknown.** Second,the governmentas the singlelandlord would have the added practical difficulty of searchingfor the equilibrium rent, as well as the political problem of collecting the rents. Third, the statewould continueas the (inefficient) managerof the housingstock. The surestand most efficient way to restore housingmarket equilibriumwithout creating additionaldistortions during the transition is to unwind totally the socialisthousinglegacy by eliminating all housingsubsidiesand the taxes that finance them and giving away the existing stock.” With homeownerequity, market rents would then be affordable to most households. The transitionwould, of course,needto focuson creatingan environmentin which the giveaway would be acceptedby tenantswho in somecasesare now paying very low rents. This issueis probably highly idiosyncraticacrosscities and countriesand is beyondthe scopeof this paper. A common objection to this proposalis that a valuableassetis being given up by the impoverishedgovernment. We have alreadyaddressedthis objection, at leastin part: if rents are below operatingcostsor expectedto rise by lessthan 5 percentper year in real terms in the absenceof the giveaway, then the housing“asset” is really a liability. But what if rents equal operatingexpensesand are expectedto rise by 5 percent per year in real terms? And further that the financing rate is only 8 percent? According to the calculationin Table 1, the stock is worth 0.23 of replacementcost. However, the presentvalue of tax revenue,usingthe same20 percent tax rate, expectedto be collected on earnings from a replacement-costunit of the deregulatedhousingstockis evengreater than0.23. Calculatingthe presentvalueof taxesfrom $1 of stock as T(uc-oper)/(fin+depr), we obtain 0.25 for any assumedlevel of the fmancing rate.% The deregulatedhousingstock is worth more to the governmentin private handsthan the regulatedhousingstock is in the government’sown hands!
22 As a fmt step, dividends equal to fid(l7) on a stock valued at reproduction cost could be allocated to subsidies. If households used part of their subsidies to consume nonhousing goods or if PCPEs have overinvested in housing, market-clearing rent levels will be below long-run equilibrium levels. Subsidies would then have to be cut temporarily to reflect the below-market return on existing housing capital. Over time, an excess of absorption over net new construction would increase rents toward the long-run equilibrium level, and subsidies would increase.
z3The state need not gi ve away all publicly owned units. The share of public housing in the total housing stock in market economies varies from 2% in the U.S. to 30% in Britain (Renaud, 1991, p 38, note 15). The criteria for robust private markets is that public housing not be an alternative at a subsidized rate. Thus, the percentage of housing that remains public needs to be kept low, and the supply strictly rationed. 24 The present value of taxes per unit of stock is calculated as
TA;y=~ * t(ucRCt-1
oper) ( 1 -depr) t-1 .
(l+finJt
19 C. Equity Concerns and Political Objections A major concernwith housinggiveawaysis the priority often extendedto the existing tenant and thus the perpetuationof past inequities. This concern could, as mentionedearlier, be addressedthrough careful designof an ownershipvoucher plan. The voucherswould be distributed basedon a politically determinedformula. The state (or local government)could then set an initial price, in voucherunits, for the housingstock. Voucherswould be issued.for the entire stock. Then, an activesecondarymarketcould be createdto facilitate housingmarket adjustmentsz The voucher schemecould, for example, allow those in the queue who have not yet receivedunits to be compensated by thosewho have. In practice, an explicit “vesting” scheme would be created“ex post”. Age (pensionerswill likely be protected; they have, after all, had their wagessuppressedlonger than the young have), time in unit, etc. are all issuespoliticians might address. Becausethe current housingsystemis inequitableto almost all and the voucher distribution schemeis infinitely malleable,equity concernsare not a legitimate obstacleto rapid market reforms. We believe the political problems associatedwith this reform have been greatly aggravatedby the previousfocus on settingsalesprices high and failing to considera dividend distribution. The likelihood of a shrinking housing stock, owing to sales to foreigners and conversionsto other uses,and the obviousadvantageof those with accumulatedliquid wealth raiselegitimateconcernsregardingeviction, foreclosureandtenancyrights. Housinggiveaways, on the other hand, would reducetheseconcernsby greatly increasingthe affordability of most current tenantsandthusreducingthe political issueto the more manageableestablishmentof the appropriate socialsafetynet. Another reasonfor the delayin housingprivatizationis the perversepolitical incentives createdby the transfer of the statehousingstock, along with the liability for housingservices, to local governments. Local governmentshavenot yet developedalternative revenuesources, have no influenceover housingwage reform, and are not taxed on liquidity overhangcaptured by assetsales. Consequently,they havean enormousincentiveto garner high salesprices and little incentiveto speedreform (Alm andBuckley, 1992). The incentivemechanismfor housing reform can be improved by the central governmentrecoupingproceedsfrom housingsales, adjusting revenuesharingand providing the legal infrastructure that supportshousingmarkets (Renaud, 1991). Encouragingthese actions should be a top priority for the international agencies.
25 While formal mortgage fmance is not necessary prior to the privatization of the existing stock, some form of finance is necessary to facilitate efficiencyenhancing trades and is desirable to allow individual households to move up or cash-out and entrepreneurs to purchase and develop rental housing. However, the system must be limited because excess credit in transitory housing markets could destabilize housing markets and undercut macroeconomic stabilization policies, just as excess real estate credit in the 1980s is undermining the U.S. economy today (Hendershott and Kane, 199&Q.
20 Finally, we note that simply giving housingto existingtenantsis no more inequitablethan a glaciallygradualincreasein rents where the presentvalue of the future below full-equilibrium market rents equalsthe fundamentalvalue of the housing. In both casesthose with better housing benefit equally relative to those with worse housing. Given the above described efficiency gainsfrom the giveaway, it is preferred to the glacially gradualrent increase.
V. SUMMARY
Transformingthe housingsectorto a market orientationought to be a top economicand political priority of formerly socialist economies. This prioritization is appropriatebecause effectively-functioning housing markets would improve the utilization of the existing stock, facilitate labor mobility and thus productivity gains in the industrial sector, and provide strong collateral to tap existing wealth for venture capital, as well as lower the cost of new housing construction. Despitetheseputativebenefits,tangibleprogresstoward housingsectorreform has been very limited in most of thesecountries,using as a benchmarkthe rent levels chargedin public housingor the shareof public housingstock that has beenprivatized. * The failure to establisha well-functioning housing market stems primarily from two fundamentalmisconceptions.The first is that giving the housingstock away will erodethe fiscal position of governments.The secondis that the existinghousingstockis unaffordablein private ownership. The first misconceptionrelates to valuing real estate. In many reforming economies, housingprivatizationdecisionsare being madeat the local governmentlevel, and thesecashconstrainedgovernmentsoften expect assetsalesto finance their operations. But real estate valuesdependon expectedfuture cash flows. Without bonafide rental reform, which cannot occur without distribution of the housingdividend,the housingstock in the PCPEshas trivial fundamentalvalue(currentcashflows arenegative). While somesalescan be achievedat prices abovefundamentalvalue, such saleseither reducethe stock of housingavailableto citizensor fuel inflation. Giving away the housing stock is an excellent meansto prevent such sales. Taxing the increasedproductivity at rates significantly below the 100 percent marginal rate implicit under state ownership improves the fiscal positions of both the government and households. Rent to incomeratios in the PCPEsare on the order of one-eighthof those in market economies. As is well recognized,theselow ratios follow (in part) from a tax being placedon current wagesto financepart of current rental operatingcosts. Not previouslyrecognizedis that the low ratios are evenmore the result of a 100percentimplicit tax on the returns (“dividends”) from pasthousinginvestments.Recognizingthis tax is important for two reasons.First, without removal of the housingdividend tax, along with the wage tax, householdswill not be able to afford long-run equilibrium rents for at least a decade,and until they can, a well-functioning housing market will not exist. Second,understandingthis tax helps clarify the very serious equity concernsthat characterizeand often obscurethe privatizationdiscussion. While the return on the governmentowned housing can be distributed as either an ongoing regular dividend or a one-time specialdividend, the one-time giveaway is preferred. This method both assuresthe transition to a system in which housing servicesare allocated accordingto market signalsand gets the governmentout of the housingmanagementbusiness. To not pay the specialdividendcould be a policy error of enormousproportions.
REFERENCES Alexeev, Michael, etal., Overview of the Soviet Housing Sector, Washington,D.C: USAID, December1991. -
“Market v. Rationing: The Case of Soviet Housing,” Review of Economics and ‘Statistics,70, 1988.
Alm, James and Robert M. Buckley, “Privatization by Local Government: A Net Worth Perspective,”World Bank, 1992. Blanchard,Olivier, RudegerDombusch,PaulKrugman,RichardLayard, andLawrenceSummers, Reform in EasternEurope, Cambridge,Mass.: MassachusettsInstitute of Technology Press,1991. Bleaney, M. F. Do SocialistEconomiesWork? The Soviet and East EuroneanExnerience, Basil Blackwell, 1988. Buckley, Robert F., Patric H. Hendershottand Kevin E. Villani, “The Viability of Thrifts in a PrematurelyDeregulatedEnvironment:Are They All ZombiePonzies?”presentedat the 2nd Annual AREUEA InternationalMeeting, Mystic CN, October 1993. Calvo, Guillermo and Jacob Frenkel, “From Centrally-PlannedTo Market Economies:The Road from CPE To PCPE,” NBER #3698, May 1991. Daniel, Zsuzsa,“The Effect of HousingAllocation on SocialInequality in Hungary,” Journalof ComnarativeEconomics,December1985. Daniel, Zsuzsaand AndresSemjen,“Housing Shortageand Rents: The HungarianExperience,” Economicsof Planning, 1987. Fischer,Stanleyand Alan Gelb, “Issuesin SocialistEconomyReform,” World Bank Working Paper565, Washington,D.C.: World Bank, December1990. Goldsmith, Raymond, Comparative National Balance Sheets,Chicago, Ill.: ChicagoPress,1985.
University of
Hendershott, Patric H., “Valuing Real Estate in Overbuilt Markets: Application to Office Buildings in Sydneyand Melbourne,” presentedat 2nd Annual AREUEA International Meetingsin Mystic CN, October 1993. Hendershott,Patric H. andDonald R. Haurin, “Adjustmentin the Real EstateMarket, AREUEA Journal,Fall 1988.
23 Hendershott,Patric H. andEdwardJ. Kane, “CausesandConsequences of the 1980sCommercial ConstructionBoom,” Journalof Applied CornorateFinance,Spring 1992. -
, “Office Market Values during the Past Decade: How Distorted Have Appraisals Been?,”presentedat theFifth AnnualAustralasianBankingandFinanceMeeting,Sydney, November 1992.
Kalinina, Natalia, “An Account of Housing Privatization in Russia,” Russian Academy of SciencesReport of the World Bank TechnicalCooperationProject, World Bank, Washington,D.C., June 1992. Katsura, Harold and Raymond Struyk, “Selling Eastern Europe’s Social Housing Stock: Proceedwith Caution,” Urban Institute, January1991. Kingsley, Tom and Marian Maxiam, “Housing Costs and Affordability in Czechoslovakia: The Opportunity for PrivateHome Building,” Urban Institute, April 1992. Kingsley, Tom and Raymond Struyk, “Progress in Privatization: Transforming Eastern Europe’sSocialHousing,” Urban Institute, April 1992. Komia, Janos, “The Hungarian Reform Process,Visions, Hopes, and Reality”, Journal of EconomicLiterature, December1986. -,
The Road to FreedomShifting from a Social Svstemthe ExamDleof Hungarv, New York: Norton, 1990.
Mayo, Stephenu.,
Housing:IndicatorsProiect, World Bank, 1992.
McKinnon, Ronald’ The Order of Economic Liberalization Financial Control in the Transition to a Market Economv, JohnsHopkins Univ. Press, 1991. Newberry, David, “Reform in Hungary: Sequencingand Privatization,” in Graham Bird, ed., EconomicReform in EasternEurope,London: Elgar, 1992. Olsen, Marcus, The Rise and Decline of Nations, 1983. Pogodzinski, J.M., “Reform of the Polish Housing Market: Manuscript, May 1991.
Issues and Methods,”
Prell, Michael, “The Role of Housing in Soviet GNP Estimates,” Review of Income and Wealth, 1989. Renaud,Bertrand, “HousingReform in SocialistEconomies,”The World Bank, April 1991.
24 Sappington,David and JosephE. Stiglitz, “Privatization,Information andIncentives,”Journalof Policv Analvsisand Management,6, 1987,567-582. Solimano, Andres, “Macroeconomic Adjustment, Stabilization and Growth in Reforming SocialistEconomies,”The World Bank, 1990. Stiglitz, JosephE. “TheoreticalAspectsof the Privatization: Applicationsto EasternEurope,” Institute for Policy Reform Working Paper,September1991. Szelenyi, Ivan, Urban Ineaualities Under State Socialism, New York: Oxford University Press, 1983. Tanzi, Vita, “Mobilization of Savings in Eastern European Countries: The Role of the State,” in Anthony Atkinson and RenatoBrunetta,eds.,Economicsfor the New Europe, New York: New York University Press,1991. Tolley, George S., “Urban Housing Reform in China: An Economic Analysis,” World Bank Working Paper 123, Washington,D.C.: World Bank, April 1991.
25 Table 1. The Value of Real Estate Relativeto ReplacementCost Rents = 10% of Market uc=.19 uc=.24
Rents = Operatine Costs uc=.19 uc=.24
RC = 1.0
-.09
-.14
.23
.06
RC = 0.7
-.06
-.lO
.16
-04
.
