Loading...
HomeMy WebLinkAboutAgenda Report - June 20, 1991 (68)�CW CITY OF LODI COUNCIL COMMUNICATION 4Fpw+. AGFIM TITIE: 1991-92 Appropriations spending Limit MEETING DATE: June 20, 1991 PREPAM BY: Finance Director RF)OOMWM ACTION: That the City Council provisionally set the 1991-92 Appropriations Spending Limit at $33,441,797 as calculated using: 1. The percentage increase in the California Per Capita Inccane; and 2. The greater of the percentage .increase in the City's own population growth or the population growth of the entire San Joaquin County. These figures were suppliedby the California Department of Finance. The 1991-92 Appropriations Spending Limit YnF require adjusturent if the County As se. sor ` s Office provides data showing that percentage change in the local asseskrent roll from the preceding year due to the addition of local non-residential construction is greater than the percentage increase in the California Per Capita Income. The County does not have the ability to provide this information at this titre . BACKGROUND IWORMATION: Article XIIIB of the California State Constitution specifies that an annual Appropriations Spending Limit shall be established to place limits on the amount of revenue which can be spent by the city. In J1une, 1990, proposition Ill was passed which modified the earlier proposition 4 and the corresponding legislation regarding calculation. The current legislation has dwwgud the annual growth adjustrrent factors used in the calculation of the Appropriations Spending Limit- The City Council must choose between: 1. The population growth of the city; OR 2. The population gz Lvth within the County. The recomrendation of the Finance Dunt is to use the percentage which Will result in the highest Appropriations Spending Limit. The population growth of the County is higher in fiscal year 1991-92. The population growth of the City was higher during fiscal years 1987-88 through 1990-91. APPROVED: t .liter THOMAS A. PETERSON City Manager L] CC..APR0/CO.00M CC -1 1991-92 Appropriations Spending Li dt June 20, 1991 Page Two Currently, the county does not have the ability to provide the data for the change in non-residential assessment. Therefore, vie have no choice but to use the California Per Capita Insane as part of the calculation. In future years, the Council w i 11 also have to maks a selection between these two items. In calculating the 1991-92 Appropriations Spending Limit, the new growth factors were applied to the 1986-87 Appropriations Limit and each year thereafter. This did not change the limits €or these years but allowed the accumulated growth in these years to be applied to the current year. See attached Exhibit A for the appropriate growth rate factors and Appropriations Spending Limit calculations. F'JNDING: Does not apply, H. Holln Finance Director RHH:DW: ss Attachment (Exhibit A) Prepared by Diana White, Assistant Finance Director OC APRO/CO.CCM City of Lodi Appropriations Spending LL.Li.t Growth Factors/Calculations GROWr ! FACTORS: FY 87-88: (1.0347)x(1.0572)=1.0939 (1.0939)x$22,654,787=$24,782,072 ISE: $22,654,787 was 86-87 Appropri at3ons Spending Limit. EY 88-89: (1.0466)x(1.0496)=1.0985 (1.0985)x$24,782,072-$27,223,106 FY 89-90: (1.0519)x (1.0252) =1.0784 (1.0784)x$27,223,106=$29,357,398 FY 90-91: (1.0421)x(1.0226)=1.0657 (1.0657)x$29,357,398=$31,286,179 FY 91-92: (1.0414)x (1.0264)=1.0689 (1.0689)x$31,286,179=$33,441,797 FOIA ASI,/00.OD14 % Increase Fiscal Per Capita city county Year Incxme . Population Population 87-88 3.47 5.72 3.33 88-89 4.66 4.96 3.32 89-90 5.19 2.52 2.20 90-91 4.21 2.26 2.23 91-92 4.14 1.19 2.64 FY 87-88: (1.0347)x(1.0572)=1.0939 (1.0939)x$22,654,787=$24,782,072 ISE: $22,654,787 was 86-87 Appropri at3ons Spending Limit. EY 88-89: (1.0466)x(1.0496)=1.0985 (1.0985)x$24,782,072-$27,223,106 FY 89-90: (1.0519)x (1.0252) =1.0784 (1.0784)x$27,223,106=$29,357,398 FY 90-91: (1.0421)x(1.0226)=1.0657 (1.0657)x$29,357,398=$31,286,179 FY 91-92: (1.0414)x (1.0264)=1.0689 (1.0689)x$31,286,179=$33,441,797 FOIA ASI,/00.OD14 RESOLUTION NO. 91-119 RESOLUTION SELECTING N44 -PL GROWTH ADJUSTMENT FACTORS AND ADOPTING A PROVISIONAL APPROPRIATIONS SPENDING LIMIT FOR 1991-92 1 N COMPLIANCE WITH PROPOSITION 111, ARTICLE X I I I B OF THE CALIFORNIA STATE CONSTITUTION RESOLVED, that the City Council of the City of Lodi does hereby select the annual growth adjustment factors and adopt a provisional Appropriations Spending Limit for 1991-92 in compliance with Proposition 111, Article X I I I B of the California State as follows: I. 1991-92 Growth Factors Used: A. Population growth within San Joaquin County of 2.64% B. Growth in California Per Capita Income of 4.14% 11. Appropriations Spending Limi t: Total Appropriations Spending Limit for 1990-91 (allowing for accumulated growth due to new growth factors being applied and to FY 1987-88 through 1989-90) $31,286,179 Increased by allowable San Joaquin County population growth of 2.64% multiplied by California Per Capita Income of 4.14% Therefore, the Appropriations Spending Limit for 1991-92 $33,441,797 The 1991-92 Appropriations Spending Limit may require adjustment if the County Assessor's Office provides data showing that percentage change i n the local assessment roll from the preceding year due to the addition of local non-residential construction is greater than the percentage increase in the California Per Capita Income. The County does not have the ability to provide this information at this time. Dated: June 20, 1991 1 hereby certify that Resolution No. 91-119 was passed and adopted by the City Council of the City of Lodi in an adjourned regular meeting held June 20, 1991 by the following vote: Ayes: Council Members - Pennino, Pinkerton, Sieglock, Snider and Hinchman (Mayor) Noes: Council Members - None Absent: Council Members - None Alice M. Reimche City Clerk RES91119/TXTA.02J �. Approved by ty Council 4/12/89 Amended 4/11. 25/90 and 12/19/90 FIVE-YEAR PHASING SCHEDULE CITY OF DAVIS PHASED ALLOCATION PLAN The following schedule represents the initial schedule for the eligibility of approved projects to apply for residential building permits. The figures represent the maximum number of residential building permits which may be processed for each project with each fiscal year for single-family units. The multi -family unit allocations may be processed at any time during the five-year period. - - - - - - - - - - HOUSING UNITS - - - - - - - - - - - - -- N. Davis Farms 20 4 24 Northstar 107 133 80 20 340 Wildhorse 75 85 104 264 Crossroads 85 135 85 70 100 475 Stonegate East 90 59 83 284 516 Mace Ranch 105 105 105 105 1053 210 735 Oakshade 30 124 72 1 45 17 1 203 1 491 Willowcreek 50 53 103 Waggener 130 130 MacDonald 1 45 1 4 FDC/Southfield 45 40 85 Sunnyside 30 38 68 Evergreen 1 25 25 119 169 willowbank 13 13 I In 1989 only, building permits may be issued one month prior to the beginning of the 1989-90 fiscal year. 2 Single-family units to be sold for less than $200,000: Crossroads, 50 units; Oakshade, 49 units; Sunnyside, 30 units. 3 Living groups, co -housing, and housing developed by non- profits for permanent affordability, with no more than 150 such units annually. The City Council approved 105 units to either Single- family for 1995/96 or Multi -family for 1990/96. NUMM aAL rrarAr u as "39s RM ar+q r r'A►uLr eMr•I�t� snt TABLI&A im rrrw Arwy urrY . . . nr f�ti • nYYr*u�•w • . titi •NYf•tY.A.�YMrINr IWVYTIYM• N r nwTK4IRm i s Asia Y+r uw fi WiYM � � • w r�� MYt YY►rr.r. .rYs V V rN. C.tbtMVATIri N'M4AMU a Ams wr naI .uw.•. MwW ..M.ibY{Ytir PYMYf tYNNr. prWMN{ V QU4" ruin w a A"es r M MIa1r1AL rkwAr111 Is Aralm �ot.uanal Muu• :.. • A t Y f f{ t f ��.ssals�l i, i CHILES j. P A R K i aAw _ CALVO{M4 1"Snr RAM --==M- Neumiller & Beardslee A PROFESSIONAL CORPORATION FOUNDED AS ASHLEY s NEUMILLER ATTORNEYS AND COUNSELORS JANUARY f9O3 FIFTH FLOOR WATERFRONT OFFICE TOWER II 509 WEST WEBER AVENUE STOCKTON. CALIFORNIA 95203 GROWTH CONTROLS AND HOUSING PRICES: SELECTED NAILING ADDRESS: PO BOX 20 STOCKTON. CALIFORNIA 3 82 01-3 02 0 TELEPHONE 12091 946-5200 FAX (209) 9a8-4910 -NOVEMBER 1989 0 LINES LINCOLN ,,INSTITUTE '-OF LAND POLICY Inside this Issue Case in Point 2 Lincoln Calendar 3 b a d Titles 4 Deeds 5 Siting$ 6 THE LINCOLN INS1irt1TE OF LANDPOLICY is a non -pmt educational institution that enables policymakem administrators, andother stu• dents toexplorethe complex linkages betweet. publiepolicies, including taxation. and land policy. and the impact of these policies cn major issues of our society. The major goal of the Institute is tointegrate t! ! . practice end understanding of land policy and those forces influencing that policy, especially taxa- tion, which hawsignificant impact uponthe lives and livelihood ofall people. The Institute is a tax exempt school providing advanced education in land amoomics, including property taxation, and offering challenging opportunitiesfor learning, research and publi- cation. LAND rMMSis published six times each year by the Lincoln Institute of Land Policy ® Copyright 1989 Linco!n Institute of Land Policy 26 Trowbridge Street Cambridge, NIA 02138 617/6613016 Do Growth Controls Really Matter? William A. Fischel Professor ofEconornics Dartmouth College This article is abstracted from a comprehensive review ofgrowth controls that Professor Fischel is now completing for the Lincoln Institute ofl and Policy, with funding from the Urban Land Institute Thestudy reviews the empirical evidence and findings of&w, IZO published studies. Here Professor Fischel provides the conclusions that he has reached from the evidence. The full study, available from the Lin- colnlnstitute later thisyear, criti- cally reviews empiriealstudies on zoning andgrowth controls. that growth controls are efficient or in- efficient. While many studies show that growth controls do have effects, few at- tempt tomeasure both benefits and costs of land use regul ation. The few cosUbenefi t analyses that exist indicate that growth controls are likely to be inefficient. The major costs seem to be wasteful decen- tralization of employment locations and too much commuting. The focus of the literature surveyed for this study was local government control of development, not national environmen- tal policy. Theselocat controlis iinidude tightening Of traditional zoning laws as well as mom Itoria'on the extension of water andsewer tines andnonprice ra- tioning on building permits. Causes and Consequences of Growth Controls Recent growth management programs most frequently occur in two types of com- munities. Small, relatively affluent cities and suburbs are the typical locus of exclu- sionary zoning policies.' I have argued elsewhere' that the growth control move- ment was in part caused by judicial and state legislative attempts to limit exclu• sionary zoning. These limitations may have led the affluent communities that did not want to accept large amounts of low-income housing to adopt a fail -back policy of excluding all new housing. Growth controls are seemingly beyond ju- dicial reproach on exclusionary grounds because they democratically exclude everyone. Indeed. many growth manage- ment programs go out of their way to mention that what little growth does oc- cur should contain a low and moderate income housing component. Such benevo. lence may not offset the overall effects of reotrirtion on the housing market. Growth controls also seem to arise often in states in which citizen ballot in- itiatives are common. Direct democracy allows for little of thecompro—i— and bargaining that goes on in representative government. Measures that provide a small benefit for a iarge number of voters and impose a iarge cost on an isolated group of citizens are more likely topass in a plebiscite than in a legislature. Growth controls adversely affect a rela- tively small number of voters in the jurisdiction—landowners and business interests—while providing financial gains or community amenities to a large num- ber of existing residents. Courts of law might offset this political imbalance if they were to respond to de- veloper complaints about such practices by requiring that the commucity pay just compensation for the devalued land. No state court, however, has intervened sole- ly on the basis of landowner devaluation unless the errant regulation is so extreme that it ieaves the landowner with almost no use for his or her land. The consensus of legal observers is that the California courts have been the most accommodat- ing to community regulation and the least sympathetic to landowner co=lt plaints.' This. combined with the wide- spread use of voter initiatives. has made (Continued on pg 2) WILLIAM A. FISCHEL is a Professor of Eco- nomics at Dartmouth College. where he has been a faculty member since 1973. Author of The Economics of2oning Laws: A Property Rights Approach toAmerican Land Use Con- trols (Johns Hopkins University Press, 1985). Recently, ha has organized a conference of legal scholars on the takings issue (papers pub- lished in the Columbia Law Review) end a con- ference of economists on land use controls (papers to be published in Land Economics) -2 i LINCOLN INSTITUTE (IF LAND POLICY LAND LINES Do Growth Controls Really Work? (Continued from pg. 1) California the undisputed leader in growth controls both in the 1970s and the 1980s. The result of these two settings— small affluent communities or communi- ties that adopt controls by referendum— is that growth controls are apt to go too far. In situations where some type of controls may be efficient in facilitating reasonable development, communities will tend to adopt controls that are too extreme. The cost of voting for extreme controls is not brought home to the voters or suburban councils because those adversely affected are either a small fraction of the electorate or not resident in the community at all. Aside from their adverse effects on the cost of housing, inefficiently restric- tive grnwth controls probably cause metropolitan areas to be too spread out. This is not to deny that growth controls may make development in individual municipalities more compact. Such local ordinances cause developers to go to .other communities. The most likely al- ternative sites are in exurban and rural communities, where the political climate, at least initially, is more favorable to -� development. As these communities in turn become partly developed, the new. comers wrest the political machinery from the progrowth farmers and busi- ness interests. Then these communities, too, adopt growth controls, sending de- velopment still farther from employment and commercial centers, Eventually, errF ployment and cannercial activities also disperse from traditional population centers as they find that employees and c sb=ew ate harder to find. The long -run effect of this is a lower standard of living. People will commute more than they otherwise would, which reduces their real inwmes; Dispersion of residences and Jobe promotes more au• tomobile travel and longer trips, mat- ing more congeation and pollution and eventually requiring more highway con- struction. A tic>ne subtle consequence of ineffi- ciently dispersed homes and businesses is the Ions of agglomeration economies for firms, 7be advantages of operating a business in the proximity of many other businesses is one basis for urban econo- mies. Location in a city allows firms to have access to a more skilled and flexi- ble labor force. It also permits the face- to-face exchange of ideas, which pro- motee innovation. Ebroas that tend to disperse firms erode euch advantages and reduce potential output from the in- dustry. Though advances in telecominu- nications and electronic media have induced at least some businesses to leave urban areas without any loss of ef- ficiency, such firms are still the excep- tion. Face-to-face contact is an essential ingredient of most growing businesses. Conclusions Is Land use controls do provide some benefits that would be difficult to ob- tain under less coercive conditions. Aboli:,ion of zoning and related con- tmals would create a demand for alter- native controls It is not clear that these alternatives would be less costly to administer or mare efficient in their effects than zoning. ■ Gbmkh controls and other aggressive extensions of land use regulations probably impose costs on society that are larger than the benefits they pro- vide. The higher housing prices as- sociated with communities that impose growth controls are more like- ly the result of wasteful supply con- straints than benign amenity production The last conclusion is more tentative because only a few studios have addressed it in a persua- sive framework. NOTES 'Rolleston, Barbara Sherman, "Determinants of Restrictive Suburban Zoning: An Empiri- cal Analysis." Journal of Urban Economics 21 (January 1987Y 1.21. 71whel, William A. The Economics of Zon- ing L(uvs: A Property Rights Approach to American Land Use Controls. Baltimore: Johns Hopkins University Press, 1985. chap- ter 15. 'Ellickson, Robert C.. and A. Dan Tarlock. Land Use Controls: Cases and Materials. Boston: cattle, Brown, 1981, page 75. Nu%t?t esr.tt lwn Capitalization of Regulations in Land Values study of growth controls in Fair- fax County, Virginia, provides good evidence of the effects of controls on land values. For a 1974 un- published study on "Land Prices and Factor Substitution in the Metropolitan Housing Market" (Urban Institute work- ing paper 1207-24), George E. Peterson obtained a sample of almost all of the vacant, residentially zoned parcels that sold in Fairfax County during the period 196373. He estimated price per acre of each parcel with and without the zoning constraints included. The addition of zoning constraints showed that zoning influences land values. Peterson calculated price per acre ef- fects by distance. He found that on land next to the central business district (CBD), there was a seven -fold price difference between land zoned for 20 units per acre and land zoned for one unit per acre. On land fifteen miles from the CBD, this differential shrunk to "only" a three -fold difference between 20 units per acre and one unit per acre. This answers one objection to zoning studies. which is that even in the ab- sence of zoning. lots would be larger in the suburbs. This is true, but in Peter- son's sample, the large minimum lot nzeswere still abinding constraint in the farther suburban areas of the county. Moreover, this finding shows that restrictive controls applied to a large fraction of suburban land can have significant effects on urban structure, pushing development to remote locations as close -in development is precluded. Peterson's observation period over- lapped the beginning of Fairfax County's sewer moratorium, which be- gan in late 1972. Peterson found that by 1973, the sewer moratorium's effects radically changed his model's estimated effects. Having a grandfathered and thus permissible sewer connection pushed the value cF a lot way up, while the implicit value of other characteris- tics. such as proximity to the CBD , actu- ally fell. Even his measure of permitted land use intensity, the zoning variable, became much less significant. This sug- gests an important override effect of growth controls. The existence of new controls map reduce the apparent impor- tance cf old controls. such as traditional zoning. FROM: "DO GROWTH CONTROLS REALLY MATTERT' By William A. Fischel v !i. . v PRINCIPLE HOUSING FINDINGS CT7\rrIVT.TT TW o An average of 315,000 housing units need to be built annually through 1985. o Approximately 4 percent (365,000) of existing housing units need to be replaced. o Nine percent (860,000) of existing housing units need to be rehabi 1 i to ted. 0 23% of all low income households pay more than 250 of their income for housing. o The median price of a home in California in 1980 was $97,961 while nationwide the median price was $62,060. o 430,000 households are overcrowded. The following conditions contribute to California housing problem: o The post-war baby boom generation is moving into the household formation period. o Net immigration into California is on the rise. o The number of households has increased due to high divorce rates and professional men and women marrying later. o Housing lots have become increasingly scare in California's metropolitan areas. o High inflation has caused savers to turn to other investments, thus making mortgage funds from banks and savings and loans scarce and available only at high interest rates. o.;,, use -.regulation's` pernttsI . and=everchanging building 'standards are increasing housing Costs. o Californians now pay an average of 37% of their income towards house payments. (Nationally, house payments average 24% of inane). l� SOURCE: Ca Tax Research Bulletin, October 198i pg. 3 UT i 14EASURE A THE GREENBELT INITIATIVE On August 25, 1981, the voters of the City of todi approved an initiative ordinance which eliminated the City's Planned Urban Growth Area from the Land Lke Element of the General Plan. The effect of this Ordinance was to establish the new urban growth boundary at the city limits as shown cn Exhibit A. At the present time, annexation of County property to the City for urban development purposes is not possible without an amendment to the Land Use Element of the General Plan. The effects of this initiative upon housing cannot be determined at this time. It has affected the assumptions concerning housing since properties once considered potentially buildable are now excluded. If construction is to occur, it will be limited t c those areas already within the City limits. Any direct relationship betvieen the Greenbelt Initiative and fluctuations in construction activity cannot be proved due to innumerable other variables which include interest rates availability and development of land in nearby areas, and weather conditions. It#<f iaied ghat as theamoc�n of -=vacant Cy`tand r a l.'iite prte o�emm#nfng` va�ani: "tend wit ! increase, -and 2. t'he devetopm� t= tfi8' with occur, . in att probability, be at `higher density �untts:per acre) dee to increas`d grid casts. Ix REAL ESTATE ?:.'^�r r,:' f :wr-:� F:': .-�:.;Ti•. ,,y.. K - F_qy ,:i;��;•` .:.`• r..,K v -`::. ' HE LONG-RUNNING ROOM in Southern California is the stuff most chambers of commerce can only dream of a soaring popula- tion, office buildings sprouting everywhere. and prices on some single-family homes climbing in value by more than 52,000 a week. But lately something has changed. Growth isn't such a kick anymore. Increas- ingly it's looked upon as a threat to a pleas- ant and prosperous way of life, and as something to be resisted. The growth -con- trol movement that has resulted jus: may be the harbinger of a national trend. For much of the population of Califor- nia. and particularly that of the vast moun- tain-ringed Los Angeles basin. economic growth now conjures up visions of stupen- dous traffic jams, overflowing sewer sYs- terns. and pollution -filled air. The natives are growing rebeltious. In 1986.69% of the voters in Los Angeles approved a plan to slash by half the allowable density of future commercial and industrial buildings in most of the city. Last year a slow -growth candidate won a seat on the Los Angeles city council, defeating the council presi- dent. Across California 14 of 20 growth - control initiatives carried the rote in 1987. Eight of ten cities in Ventura County. just west of Los Angeles, have passed slow - growth measures in recent years. In all. 57 cities and eight counties in California have voted to limit growth. according to Made- lyn Glickfeld, an urban planning consultant in Malibu. "There was a time when Los Angeles was going forward and people wanted growth," says Sandy Brown. a physician's wife in Westwood, an upscale neighbor- hood that includes the campus of UCLA. But then a few years ago. Brown stared out her kitchen «indow and saw a bulldozer pulling down yet another home nearby to make room for apartments. "All of a sud- den 1 wid to myself, 'What is going on here?"' She proceeded to help create Friends of Westwood, a neighborhood as - Massive housing projects forced into areas far fromwhere residents work have led to jammed Riverside and Orange Countyfreeways. R � ,,. ■ t A �--, '�. .;let � Sharon Browningtri es to keep the peace. sociation that has become the scourge of developers 'For years growth didn't in- trude on my comfort mn� says Brown. "But now it's become an infringement on my way of life." UCH SENTIMENT in favor of stowing or halting growth could prove a more virulent national movement than the tax reform measures that began in California with Proposition 13 in 1978 and spread to many other states. "No -growth is more fundamentally grass roots than Proposition 13," says Dwight Worden. a Solana Beach, California. attorney who has written nearly a dozen growth -control measures for different ballots. "Proposi- tion 13 had a charismatic lead- cr in Howard Jarvis," observes Worden. "This movement has no single leader. It is spontane- ous in city after city." Less vigorous strains of the antigrowth virus are already flourishing around the country in parts of New York. Virginia. and North Carolina. Former U.S. Senator Paul Tsongas of Massachusetts is publicly op- posing uncontrolled growth on Cape Cod, and a senior aide to Governor Thomas Kean in New Jersey declares growth management the "biggest looming public policy issue in the state." But (ew regions have ever succeeded in using this concern to generate support for useful. comprehensive p hn- ,. ning. J. Ronald Tem illiger, head ofTram- mell Crow Co.'s residential division, the biggest builder of apartments in the coun- try, notes that antigrowth feeling was rare ten years ago. "``ow." he saes. "of the 60 cities where we operate, wc� see step it in about half" If you go back to (he 1970s, "It's 11 you can find some precursors of the present drive for slower peoplf growth. The movements then ostensibly were prompted by bfebo concern for the environment. . but sometimes were just dis- a ee guised opposition to construe- Save f tion of low-income and multiracial housing. which op- Selye! ponents thought were a threat not le, to properry values. The currenr sentiment, on the other hand. more often stems from immedi- anyotl arc and infuriating problems on bol created by rapid commercial or residential development, says Harvard economist Joseph Kalt: "Thistime it is the reality of inadequate roads. sewers. water systems. and other infrastructure," Across the country citizens have re- sponded in ways cften heavy-handed. City councils and planning authorities arc being REAL ESTATE pressured into forcing developers t o pay ev- ery nickel of the cost of the added public services that their projects make necessary. This usually means that new businesses and new -home buyers wind up paying just that much more. Ballot measums have also imposed severe and inflexible limitations on new construction. Such extreme in a steps reflect 3 deeply felt re- t sentment—often justifiable— that citizens have lost control Ig to of their local governments to developers. em- For all the sedottsrtess of the conditions that spark it, the by push to slow growth can end up ;fng causing severe problems, even for the people behind the movement. The short-term re - else sult of limiting growth in a A," community iS often to create a housing shortage, sending home prices showdrtg skyward and forcing many of those vdthjobs in the am to live far away. This only dogs high- ways all the more. Typicallysach initiatives curtail the growth of housing more than they do any increase in the number oflobs, compounding the problem. Ultimately, a community can price itself out of business expansion, and in fact drive some businesses out. leading to a loss of tax revenues and deterioration of services and Property vAues. As young workers are forced to live else- whem employers halve to pur- sue them. The movement also sets up a wrenching conflict between the haves and the have-nots, -It's like people in a lifeboat agree- ing to save themselves by not letting anyone else on board." says Frank Mittelbach, a pro- fessor of urban economics at UCLA. "The people in the wa- ter should have a vote." In Los Angeles consider- ations as diverse as geography and taxes have come together to push slow -growth sentiment onto fact -forward. The five - county Loc Angeles area, with its beaches and good weather, I,;,% attracted many new resi- dcntc anxious for a more pleas- anI life. Between 1975 and 1987 the reputation of the met - Developer Ray Watsonsays his company's c o sts h ave soared. 41A k: il 9 w REAL ESTATE ropolitan area grew rapidly, adding almost three million people, 22 times as many as the metropolitan New York area gained. Los Angeles also has become a prime ex- ample of urban sprawl -95% of the re- gion'sjobs are outside the downtown area. Even as the region has boomed, its abili- ty to handle growth has diminished. Since the mid -19705, California has fallen to49th among the 50 states in per capita spending on roads. Adjusted for inflation. nation- wide federal spending for many infrastruc- ture programs has decreased since 1980. At the Sametime. the California tax initiative, Proposition 13, has limited property taxes on homes that have not been sold since 1978. Still other measures severely limit what a municipality can spend even if it is growing rapidly. Nowhere is the pace of development more visible than in southern Orange County. down the coast about 50 miles south of downtown LA What not too tong ago were vast ranches still largely in- tact from the rime they were granted to set - tiers by the King cf Spain, now crawl with bulldozers and earthmovers. Even in the face of antigrowth campaigns, vast stretches of hillside are being stripped bare for construction. A double whammy is at work here: Those Proposition 13 - style tax -cutting measures have wound up tempting many fi- nancially desperate municipal- ities to ignore good planning in pursuit of another source of tax revenues -commercial growth. "Proposition 13 is driving changes in land use," says C. Bradley Olson, an executive at the Irvine Co.,which owns and is developing a 100 -square - mile tract of land in Orange County. "The city of Costa Mesa near here gets huge sales Wes from a big shopping cen- ter it allowed there. But people from all over drive there. so Costa Mesa has huge traffic problems." Sometimes the hunger for tax revenues is almost comical. To win city approval for a big project on its land in Tustin. the Irvine Co. had to agree to include a dozen auto- mobile dealerships in its plans so Tustin could colkct sales taxes on all the new cars sold within its borders. More often towns pursue developers of high-rise office buildings while vigorously REPORTER ASSOCIATE Kate Gallen Tom Rogers battlesagainst developers. opposing new housing, which would re- quire adding expensive new municipal ser- vices such as schools. "There is an absolute imbalance of jobs and houses," says John Martin. the Jrvine Co.'s head of residential marketing. "There has been an average of 53,000 new jobs a year created in Orange County for the last five years. and for that we should have built 42,000 housing units. But at tops, Orange is adding only 20,000 units a year." As 3 re- sult, workers clog the freeways, driving long distances to get to work. "The commute to River- side used to rake 40 minutes during rush hours." says Mar- tin. "Now it takes 2% hours." Talk about lifestyle. Infrastructure problems dog outlying areas too. With plan- ning commissions there less sensitive to the growth issue, builders are able to put up mas- sive housing developments in what was previously untouched terrain. packing the units together as tightly as if they were in the middle of a city. In some big developments south of the Irvine properties. one noteworthy amenity is in short supply: main roads. Commuters froni thousands of homes are funneled into a handful of strecis before reaching the near- est freeway. With frustration over growth mounting, developers usually reel the heat. and no or.c is turning up the lemper,irure more vigorously in Orange County than Tum "it was so- cialism for the wealthy, an attempt by develop- ers to put the cost of their sins on the public." 126 FORTUNE DECEMtlLR 5. 19XX Rogers, who has emerged as an unlit leader of the slow -growth movem there. Rogers, 63, a small-scale shopp center developer himself and a fort chairman of the Orange County Repu can Party ("I'm about as right wing you can get"). proudly displays at, graphed pictures of Ronald Reagan on wall. To drive with him through bulido: and newly developed areas of Orat County as he points out monotone rows of houses and clogged intersectio is to see the full fury aimed at develops "They're greedy bastards." he fum "Thugs in three-piece suits." Rogers's transformation from grov booster to basher follows a fairly comm pattern. "I used to be part cf that grow is -progress crowd," he explains. "But changed when the government showed couldn't plan for growth." Orange Cour traffic already was dreadful in 1984 wh county supervisors proposed a S5 billi transportation improvement program be financed by a sales tan increase t1 required voter approval. Rogers argu that the major beneficiaries would be s velopers, who would use the new roads justify new development. "It was soci. ism for the wealthy, an attempt by dev opers to put the cost of their sins on t public." says Rogers, who pulled togeth environmentalists. open -space enthusias antitax groups, and others to oppose t. measure. HEN THE SALES TAX Ic by more than 2 to 1, Roge figured that was a message government officials to sic the pace of development. "Instead, th. kept right on just like they always did." 1 says. In 1987 he began circulating pet i t i o for an elaborate Citizens' Sensible Grow and Traffic Control Initiative, which else. Bally required developers in Orange Cou, ty's vast unincorporated areas to provit roads and public safety facilities for the projects. His measure was vigorously of posed by builders and lost last June. B; polls in Orange County show that two thirds of the voters continue to be in favi of slowing growth. The combination of overburdened it frastructure and increasingly sophisticate antigrowth groups is forcing big chang� in how developers do thing,,,, In the S,, Fernando Valley to the north of dowt town Lo; Angeles, Jack SNund faces bizarre chore: if he is to obtain an exemt Hotel BeijingToronto is situated in the heart clfBeijing's business and diplomatic area, on Jianguo-Menwai Street; 20 minutes from Tfan'anmen Square. andwithin walking distancefrom the Fhmdship Store We offer superb business and pleasure amenities, that include o business center, conference rooms, and the finest Cantonese and Continental cuisine. Along with the unique and hospitable personal service that has come to char acterfie Nikko Hotels International. Where the heart Is always content hotel beij ng-toronto Aangtk Menwal street, Beijing Tet: 5002266 Telex: 210011/210012ILHCN Cable: 5650 MING Man Wd by T nikko hotels internatonal Foi reserval*ns. Calc yow travel agent. Japan Air Unes or Mkko Hotels intetnaional. Toll tree to US. and Canada t-soli-NIKKO-us t64S5"71. 128 FORTUNE 1)ECEMIJER 5. 19$R REAL ESTATE tion from the city's limit on new Sewer connections. So that his proposed office project won't add to the strain on the sys- tem. he will refit 8,000 toi!ets anywhere in Los Angeles with plumbing that uses three gallons per flush instead of the usual five. To overcome the opposi- tion cf homeowners near his proposed buildings. he hired Sharon Browning. a former social worker with a blossom- ing career in helping develop- ers work with slow -growth groups. Browning persuaded Spound to go door to door through the neighborhood asking how he could change his design to make it accept- able. He's gained support but still awaits approval by the planning commission. In Los Angeles, Friends of Westwood brought a Iat,d- mark suit that focused on a proposed 26 -story office tow- er, winning a ruling from the state court of appeals that the city has to assess the environmental impact of major projects before issuing building permits. Though the development company. Center Wast, had won zoning approvals for the ofCoe, its president. Kambiz Hekmat, negotiated an agreement with Friends of Westwood to cut the building's size by 18.5%. In ad- dition, the city is requiring that he "plant, water, and prune the trees I put on the sidewalk," says Hekmat. That's nothing: As part of settling a dispute over parking at a new hotel near Beverly Hills, the owners agreed to pay $250,000 to Friends of Westwood and other slow -growth groups, money that presumably can be used to finance growth -control efforts'-- against ffortsagainst other new projects. Developers around Los Angeles rou- tinely find themselves picking up the tab for much of what municipalities once pro- vided. Some have been forced to donate land and build new ,- firehouses, police stations, even city halls, in order to win municipal approval for their plans. In parts of orange County there is talk that de- velopers may soon have to pay the Salaries of policemen' and firemen trade necessary by their projects. Orange County forced the Irvine Co. to build a S45 million, six -lane road to accommodate a Planned 2,600 -tank communi- ty, Normal planning would call for two lanes for that size "in the long term these antigrowth areas be- come elite communities without economic dynamism." community. 'ihe cost of meeting such re- quirements _ rs staggerili& For a. neiv busi- ness center, the' Irvine_ Co.' is payirtg.225 million for road construction and improve- ments.' Says once ''chairaian. ;Watson: "In 1963 `imptovemen s on tand`.cost about 515,000 a acre Now rt `costs $250,000 an NTiGROM f movements may. bei' emotiouslly.'.satisfy'* to the participants, but. the ;eventual outcome inn be a disaster, "No - growth; -movements hurv. ft people who wanted them,--.;argaes.prviessoc Kenneth Rosen, ldmbmin . of the Center : for Real IEstafe'and i3risari Ecatcuairs.at the Uni- Earthmovers strip the hills bare to prepare for more building In southern Orange County, '7'Sa" -r... f, f Hotel BeijingToronto is situated in the heart clfBeijing's business and diplomatic area, on Jianguo-Menwai Street; 20 minutes from Tfan'anmen Square. andwithin walking distancefrom the Fhmdship Store We offer superb business and pleasure amenities, that include o business center, conference rooms, and the finest Cantonese and Continental cuisine. Along with the unique and hospitable personal service that has come to char acterfie Nikko Hotels International. Where the heart Is always content hotel beij ng-toronto Aangtk Menwal street, Beijing Tet: 5002266 Telex: 210011/210012ILHCN Cable: 5650 MING Man Wd by T nikko hotels internatonal Foi reserval*ns. Calc yow travel agent. Japan Air Unes or Mkko Hotels intetnaional. Toll tree to US. and Canada t-soli-NIKKO-us t64S5"71. 128 FORTUNE 1)ECEMIJER 5. 19$R REAL ESTATE tion from the city's limit on new Sewer connections. So that his proposed office project won't add to the strain on the sys- tem. he will refit 8,000 toi!ets anywhere in Los Angeles with plumbing that uses three gallons per flush instead of the usual five. To overcome the opposi- tion cf homeowners near his proposed buildings. he hired Sharon Browning. a former social worker with a blossom- ing career in helping develop- ers work with slow -growth groups. Browning persuaded Spound to go door to door through the neighborhood asking how he could change his design to make it accept- able. He's gained support but still awaits approval by the planning commission. In Los Angeles, Friends of Westwood brought a Iat,d- mark suit that focused on a proposed 26 -story office tow- er, winning a ruling from the state court of appeals that the city has to assess the environmental impact of major projects before issuing building permits. Though the development company. Center Wast, had won zoning approvals for the ofCoe, its president. Kambiz Hekmat, negotiated an agreement with Friends of Westwood to cut the building's size by 18.5%. In ad- dition, the city is requiring that he "plant, water, and prune the trees I put on the sidewalk," says Hekmat. That's nothing: As part of settling a dispute over parking at a new hotel near Beverly Hills, the owners agreed to pay $250,000 to Friends of Westwood and other slow -growth groups, money that presumably can be used to finance growth -control efforts'-- against ffortsagainst other new projects. Developers around Los Angeles rou- tinely find themselves picking up the tab for much of what municipalities once pro- vided. Some have been forced to donate land and build new ,- firehouses, police stations, even city halls, in order to win municipal approval for their plans. In parts of orange County there is talk that de- velopers may soon have to pay the Salaries of policemen' and firemen trade necessary by their projects. Orange County forced the Irvine Co. to build a S45 million, six -lane road to accommodate a Planned 2,600 -tank communi- ty, Normal planning would call for two lanes for that size "in the long term these antigrowth areas be- come elite communities without economic dynamism." community. 'ihe cost of meeting such re- quirements _ rs staggerili& For a. neiv busi- ness center, the' Irvine_ Co.' is payirtg.225 million for road construction and improve- ments.' Says once ''chairaian. ;Watson: "In 1963 `imptovemen s on tand`.cost about 515,000 a acre Now rt `costs $250,000 an NTiGROM f movements may. bei' emotiouslly.'.satisfy'* to the participants, but. the ;eventual outcome inn be a disaster, "No - growth; -movements hurv. ft people who wanted them,--.;argaes.prviessoc Kenneth Rosen, ldmbmin . of the Center : for Real IEstafe'and i3risari Ecatcuairs.at the Uni- Earthmovers strip the hills bare to prepare for more building In southern Orange County, '7'Sa" -r... I don't know doctor, ever since we've made this argrrisition, I've been seeing doublelf ci3`! '.AWN. of •VICE ►�t?r'Sto;.. FN7 �,�C� 1'�fStrx-til= ` � � y . 6 •t d t •' R z If you find yourself with two phones, t m baskets, too much inventoryry call EAL. We If turn your doubfie vision Into tax deductible donations that generate scholarships for needystudents, and supplies for worthy cofieges. FAlucational Assistance Ltd Phone (312) 004)(HO. or write P0. Box 3tt21.Glen Hlyn.It. 0(1133. t3h FORTUNE DECEMBER S. {9M REAL ESTATE versity of- California .at; Berkeley._ ",Th.have a negative;eit`e .. '66"'tite'localrecon, MY as firms relocate: and* real income go down In -the long.: term these antigrow areas becoiae elite communities wither economic dy ai m`isM. Trying to send newcomers elsewhe probably sounds just fine to slov growthers, but it won't solve Southern Co ifornia's problems. Only one-third of tl region's projected growth will come frog migration. The rest will derive from a hig birthrate. Even if Southern Californiat put a wall around the place, the populatie would grow by three million—equal to ar other Orange County—in 22 years_.- Faili�y to provide housing:;toacoommodate sun N THEORY expensive or inftastruc tune -short surroundings should driv unwanted residents away. Troubk is it doesn't happen that way, observe Robert Paulson,_ managing director. 0 McKttiseyr 8t Co's:1.os Angeles cilce an, „ head of a chambe. or cxir►iaterce ,study rn �:: Brants, the people `who . the''tmos services, tare the ones who remaut.'', Com par its evep!,"lly poll their operatiorig ou . IDU, slow-growth regeott:.io trod .surtablr `. romia s 1. tagy or:Browtrt- trot a '4 i tlic LA L�-Anaetes rcaion reduced hotisint growth by 15% and the rate of new com mercW space by 25%, unemploymen would double in the 1990s. Unlike the environmental movement which was arguably more altruistic, slo% growth has a distinctly selfish. "sock it to the newcomers" side to it. That means that in- convenient changes in lifestyle that might help accommodate even carefully planned growth—car pooling, say. or recycling waste—will be slow in corning. The same spirit, but on a grander scale, will likely get in the way ofcoordinated regional solutions to planning prohlem. "Everyone wants is see it happen." .,%- Ruth Galanter, the slow -growth cano ,late elected to the Lo: Angeles city crnri,:il last yrar, "but without giving up their prrri,gatives." a up' Public Report Bulletin of the INSTITUTE OF GOVERNMENTAL STUDIES r EUGENE C. LEE. Director HARRIET NATHAN. STANLEY SCOTT. Editors UNIVERSITY OF CALIFORNIA, BERKELEY Vol. 21 October 1980 No. 5 GROWTH IN CALIFORNIA: PROSPECTS AND CONSEQUENCES l..arry J. Kimbell Graduate Schooi of Management University of California, Los Angeles David Shulman Graduate School of Administration University of California. Riverside Introduction Several factors influence California's growth in employ- ment, population 2nd housing supply. These three elements are interrelated so that they act upon each other, and are acted upon in turn. To learn more about these interactions, we propose models and projections highlighting changes already taking place, and discuss policy options that can help encourage future change in desirable directions. Between 1972 and 1979 the California economy ex- perienced extremely rapid employment growth (3.7 percent annually), while population growth was substantially slower (15 percent).' Consequently, during that period the gross, labor force participation rate (total population divided by the cMlian labor force), went up from 42 to 48 percent. We assume that the participation rate will continue to in- crease in the next two decades, although more slowly. But these increases cannot continue indefinitely, because obviously the civilian labor force cannot exceed total population. Further, basic demographic factors will prcb- ably keep the gross labor force participation rate fr= risirg much beyond 55 percent, since about 35 percent of the population is either over 64 years or younger than 17. When the gross participation rate is constant the Iabor force and population must necessarily grow at identical rates. Therefore, if the high employment growth trend of 1972-1979 were to continue into the future for perhaps 20 years, population growth would have to accelerate rapidly. (One implication of this alternative is staggering: a continuation of the 1972-1979employment growth rate would yield a population of 52 million people in California by the year 2000, contrasted with current estimates pro- jecting about 30 million by the turn of the century.) On the other hand, if the comparatively low population growth of the 1970s were to continue for the next 20 years, employment growth would have to slow down dramati- cally. In short, something has to give. Current policies are often at emss purposes; some try to promote job growth, while others try to restrict popu• lation growth. As long as employment goes up much more rapidly than population, however, these policies do not outright conflict. But when it becomes more difficult to increase the number of workers in a given population, the dilemmas wall become acute. Thus strong employment growth under those conditions will lead to strong demo- graphically based demand for more housing space.: But, quality,' and it6.......1n thousing sup; y1. in combination, these and :other factoxs (tg ; inflation) stimulate powedaal forces raising,Mime prices, producing eztraordi:;ardy high pricas in most n a2or urian areas afCaiifoinla:` fn such cir- cumstances. can we develop policies that will provide for adequate and affordable housing for many more Califor- nia workers and residents, while still preserving environ- mental quality? Our study examines the magnitudes of pressures that seem likely to become much m>Q:Ie intense between now and the turn of the century. What follows is divided into several sections. First canes a discussion of the recent history of employment and population for the state, and all 17 of its standard metro- politan statistical areas (SMSA's). Next is a brief discussion of state and local legislation for environmental protection that has worked to restrict the supply of housing. This is followed by a section discussing the implications of high employment growth and land -use restrictions with respect The California Rilicy Seninar was established in 1977 as an experimental program in University -state government cooperation on the study of longer-term policy problems. Chaired by UCPresident David Saxon. the Seminar membership includes the Governor. President Pro Tempore of the Senate. and Speaker of the FIAT LUX Assembly, plus other designated governmental and University appointees. Each year the Ser irwoommissions a number of reswch projects. selected from among research proposals suggested by University faculty mem- bers. Chosen projects are fended at 150,000each. for work extending overs two-year period. This issue of the PubficAffairs Report, the first to feature research sponsored by the California Policy Seminar, summanz�s CALIFORNIAPOLICY SEMINAR a longer technical report for a general sudicnce, Copies of the technical report are available at cost from the Institute of Governmental Studies while supplies last. Copyrigh) 1980 ID by the Rcgcnts of the University of California ISSN 00333417 I 20 15 Figure I Job Growth in California, 1940-1979 (Annual Percent Change in Non -Agricultural Employment) 40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 70 72 74 16 78 Source: Sec end -note I. Dashed Kne denotes 6 peroent growth to show high growth years. to housing prices, using a model that links house prices and employment growth. Next are 20 -year population and employment projections for California and its 17 SMSA's, followed by policy implications of the analysis, and a few concluding comments. Employment Gmwth and Population Typically California has experienced its most rapid em- ployment increases during wartime. The rapid increase in the late 1970s, however, was not associated with a sharp escalation in military expenditures. (Figure 1 charts changes in nonagricultural employment for 1940-1979.) The fifth fastest growth in employment occurred in 1978, a striking occurrence because it came before the increase in military spending associated with events in Iran and Afghanistan. Job opportunities in California affect in -migration to the state, and net in -migration is linked with the difference between the California unemployment rate .and the US. unemployment rate (see Figure 2). Thus in the early 1960s, when there was no unemployment rate differential, ap- proximately 300,000 people a year came to California, and in the early 1970s when the California unemployment rate exceeded the U.S. unemployment rate by 3 percent, net in -migration approximated zero. This empirical relationship supports a prediction that, when California and national unemployment rates are the same, approximately 300,000 people will again move to California annually. We can also predict that for every percentage point the California unemployment rate exceeds the U.S. rate, 100,000 fewer people tend to come to Cali• fomia each year. Thus it would take a Californiaunemploy- ment rate three percentage points higher than the national figure to eliminate inmigration. But as Figure 2 shows, the relationship broke down in 1979. Instead of the predicted increase in inmigration, there was an actual decline. Job availability may have beom-e less of a factor in determining population growth, being replaced by the rising prices of houses in California's major urban areas relative to house prices in the nation, as a key factor that caused the reduction. If employment growth rates of the 1972-1979 period persisted, and the dynamic described above did not break down, California's population would reach 52 million by the year 2000. 2 -is suggests over 12 million people in Las Angeles County, and 7 million people in Orange County—giving Orange County the same population as Los Angeles County today—and Napa Valley's vineyards would be shared vith nearly one million people (see Table 1). Obviously, however, California population as large as this estimate would exert tremendous pressure to increase housing prices, and a rise would in turn slow down em• ployment growth, in turn slowing population growth. In A - 1-, population projections cannot be bawd on a continuation of the rapid employment growth observed in the 19705. The demand pressures that force up housing prices are likely to reduce California's future employment growth rates. 350,000 300,000 250,000 200,000 150,0150 100,000 ,a 50,00C 0 Predicted vs. Actual Net in -Migration as a Function of the California Unemployment Eate Minus the U.S. Unemployment Rate 1961-1979* 60 62 64 66 69 70 72 74 76 78 79 'The regression equation is: Net 1n -Migration - 313,982" 103.026 (California Unemployment Rate -US. Unemployment Rare). Standard Errors in parentheses: (15.790) (10.140) TM "pply ;.eshidioas Commission. In addition, in t970 the California Environ- mental Quality Act mandated the preparation cf environ - While employment was increasing rapidly in the 1970s, mental impact reports for substantial building projects. a host of physical and legal restrictions on new housing Other more direct: attempts to recttkt:growth'uy._to supply were developing. As the most readily available maintain {local; amruity valets, and ;lake sikk fortes -as building sites are used up, the supply of new urban build- buV4ot zoning at lim tationson permits issued for housing ing sites requires more extensive capital investments to units. The ,pixysilina traAd in`=zoning'actlons has hues' to make them buildable, i.e., comprises land that is not flat. WsiM One has only to compare the San Fernando and Santa Clara is lanc4 has, beefs 46i1itssffied `or "dowri , zoned. m many valleys cf the 1940s and today to dramatize the consump- areas. More'.explidt slo!w•gioroi+th potIcteshave:ifeea "adopted tion cf flnt land, much less of which is now available for in. such cities; as;.Petaittma;-:Davis, Thotuas d .Oaks and construction sites. Riverside.. Thesut. iesttktions stave beedespecially harsh on: In addition, two maaor.categorfesof lust restriciioi's oil multiple-unit,con development,_have 'Siad siVdffcant--irripactsi environments Perhaps unlntentioiliHk, Proposition 13. appears to have constraints, and=tsf 4' " ta: lirttit growtlt? Environmental accelerated thc,trend fOW "final zonings vrhich occurs: constraints associated t�ffi identifiable costs include the when s-eommt+alty-_allows only_:those developmersts that creation or strengthening of the Air Resources Board, pay their. own .way, a ,'when projected iddltioiiall tax Water Resources Control Beard, Energy Resources Conser revenues produced iqual or.exeeed estirrtated public spend- vation and Development Commission, and the Coastal ing necessary to service a development 3 Table 1 Implied Year 2000 Population if Employment Growth Persists at the 1972-1979 Rate 1979 Population 2000 Projected population SMSA (in thousands) (in thousands) Percent increase California 22,694 52,122 130 Anaheim -Santa An a -C a rd e n Grove 1,874 -7,494 300 Bakersfield 375 956 155 Fresno 485 1,180 143 Los Angeles -Long Beach 7,128 12,295 72 Modesto 251 651 159 Oxnard -Simi -Ventura 500 1,415 183 Riverside -San Bernardino -Ontario 1,445 3,435 138 Sacramento 980 2,447 150 Salinas -Seaside -Monterey 281 537 91 San Diego 1,800 5,562 209 San Francisco -Oakland 3,194 5559 74 San Jose 1,253 3,953 215 Santa Barbara -Santa Maria -Lompoc 295 695 136 Santa Cruz 176 507 188 Santa Rosa 280 1.053 276 Stockton 320 613 92 Vallejo-Fairfield•Napa 312 901 189 Source: See end -note 1. Who Benefits? Who Loses? Who has benefited and who has lost as a result of such land -use controls? If participation in heated public debate were a valid guide to who wins and loses, then surely the local California land developers would appear to have lost. This may not be the case, however, although the building industry has publicly denounced both statewide and local land -use controls. While builders clearly do not wish their current Iand holdings to be restricted in usage, this does not mean that they will not, given time and new opportunities, benefit on balance f= land -use restrictions. Thus, for example, the aggregate value of land could increase as a result of supply restrictions. California home builders experienced record profits after many of the land -use restrictions were imposed: (In 1980 the industry may have been depressed by severe financial strains stemming from US, inflation, but this situation is clearly not caused by local California land -use restrictions.) Restrictions on land supply do not stop the demand for housing, but instead increase the return to entrepreneurs who can maneuver building permits through the processes and bring new houses onto the market. A very general, basic economic principle is at work. A mobile factor of production cannot readily be deprived of competitive returns. if restrictions on development in California should take away the profits of builders, they would tend to leave California. Indeed, by making building permits available only to the most adroit. persistent, and 4 knowledgeable builders, governmental regulations may have enhanced the profits of local entrepreneurs, at the expense of the nationwide construction firms that can no longer compete as readily as they did before more extensive land - use restrictions were imposed. Tighter Iand-use restrictions mean that nationwide firms have to invest more time and money to learn how to process permits successfully, If thU is true, negative consequences of land -use restrictions probably do not fall on local California builders mnuch as on others. A similar argument holds for landowners. In any given case of a proposed local restriction, the owners of the land will almost certainly scream loudly in protest—for good reason. Actions limiting the use of their individual parcels of land will not alter California's general land housing prices, while the individual landowner stands to lose in the decision at hand. But no such presumption holds when all landowners in California are "threatened" with more extensive land -use restrictions. In fact they may collectively benefit from land -use restrictions in general, in the form of higher prevailing land prices. As will be discussed below, since the early 1970s home ownership in California has shown an extraordinary rate of return, especially for those who enjoyed small down payments and higher mortgages. Home prices increased rapidly (see Table 7), whereas the mortgage liability for a given homeowner remained fixed, Thus dwe who now own homes do not appear to have suffered from these restrictions. Table 2 California and U.S. Home Price Increases, 1971-1979 (Percent change) Year Southern California San Francisez) Bay Ares Sacramento US. 1971 3.8 5.2 3.3 7.8 1972 4.6 4.7 6.7 7.6 1973 - 6.3 6.8 7.1 85 1974 9-5 11.4 9.2 10.7 1975 15.4 13.5 11.2 10.1 1976 17.8 14.8 14.3 8.0 1977 26.7 23.7 15.4 12.3 1978 27.9 21.9 21.8 13.9 1979 20.8 15.9 21.6 14.1 Sources: Red Estate Research Councils of Southern and Northern California and National Association of Realtors. Others who may have benefited include those who place a relatively high value on preserving the environment, improving air and water quality, or reducing crowding and congestion. Who 1csW. While no defmitive answer can be given, it seems plausible that nonlandowners and nonhomeowners have suffered losses, unless they place a higher value on the environmental benefits they enjoy, than on the increased housing costs they confront. Residents of other states who would prefer to work and live in California, but who cannot afford the high housing costs, may fall into this class. Obviously, we cannot prove with certainty that land- owners and builders have benefited from the land -use restrictions they have opposed so vigorously. But it would be a serious mistake to ignore this possibility. HousingPrices and Economic Growth Califomia's rapid employment growth in the 1970s, and the emergence of supply restrictions on buildable sites, combined to intensify the home price explosion of the late 1970s. California's price increases clearly ran ahead of the rest of the United States. Table 2 represents home price data for Southern California, the San Francisco Bay Area, Sacramento and the US. (The prices reported for Cali- fornia are estimates for an unchanging sample of houses. This eliminates problems caused by changes in the mix of houses sold. The US. prices are based on actual trans- actions and are subject to changes in the types of houses sold.) As noted eadier, relative increases in home prices can slow employment growth. Anecdotal and empirical evi- dence-i.e„ the decline in in -migration in 1979 -supports this hypothesis$- Corporate ";aad`governmental ''recruiters tellus that.-the,prici. of,housirrg is the ngle.gniatest ob stacic. to' fectuiting out of -state personnel `for California 6 High home prices :end to keep new people from moving to California, and they tend to motivate current residents to leave, since equity values in modest California houses will buy much larger houses in other states. A More Realistic Set of Projections The hypothesis -that California's comparatively high home prices act as a constraint on growth -leads to our second set of projections. The model employed recognizes the influence of population growth on home prices, as well as that of home prices on employment growth. Thuswhen employment growth tends to increase in -migration, higher population in turn puts pressure on home prices, while limit population growth. (IlDtal employment can still grow if labor force participation rates increase.) In the model, t -L relationship is calibrated so that population growth stops completely when the home price/ income ratio for a given area reaches 12, i.e., the home price is 12 times the per capita income. A home price/ income .ratio of 12 means that housing costs comprise 45 percent of average household income. (The specific variables translating home prices and per capita income to household budget share are shown in Table 3,) Several different home price/income ratios were ana- lyzed before 12 was chosen as the most plausible value, bearing in mind related population projections. A value as high z$ 20 would lead to a population projection for Table 3 Salient Points in Considering Impact of a 1112 x Per Capita Income" Constraint for Housing Costs in California • Per capita annual income (1979) $9,992 • Average number of persons in household 2.64 • Average annual household income (59,992 x 2.64) $26.378 • House price at 12 times per capita income (59,992 x 12) S 119,900 • Annual payments on such a house equivalent to S 11:598 Payments as percent of average annual household income ($11,898 _ $26,378) 45% • Payment level as acceptable percent of average annual household income 30% Average household annual income required for the annual payments (S 11,898) to amount to 30 percent of income 539,660 • Based on: 30 -year, 10 percent mortgage of $95,000; property tax payments of 12 percent of market value; and insurance of .3 percent of market value. Note that the 10 percent mortgage rate reflects a long-nmestimate of mortgage interest rates rather than current rates. California of 35.7 million persons in the year 2000, and the distribution of that population across SWs seemed implausibly high in some areas. The chosen value of 12 leads to a population projection of 303 million, which is more consistent with other current population projections. Moreover home prices as high as 20 times average house- hold annual income would mean that far more than 45 percent of average household annual income would be spent on housing, and such a high budget share also sears; implausible. Further, employment growth in the major coastal urban areas already appears to be somewhat con- strained by home prices, even though the home price/ income ratios are now closer to 10. (In a few years we presumably will have more experience with extraordinarily high home price/income ratios, and this empirical back- ground may permit more rigorous econometric estimates of the appropriate value.) In areas where the ratio of home prices to per capita income is as low as 5 or 7, and where building site avali- ability peanits substantial population increases without inordinate pressures on home prices, employment can continue to grow at roughly the same rates as in recent years. The Riverside -San Bernardino SMSA is an example. In 1972 house prices were 5.7 times per capita income, and by 1979 the home price/income ratio was still only 7.4. Not surprisingly, from 1972 to 1979 population increased by 22.6 percent and payroll employment in- creased by 36.4 percent. On the other hand, in areas where home prices are as high as 10 times per capita anaual income, and where home prices are also quite sensitive to slight increases i -r population (e.g., Loo Angeles), employment growth is likely to slow down considerably as it becomes more difficult to achieve higher labor force participation rates. In Los Angeles the home price/income ratio was 6.8 in 1972, and had reached 10.7 by 1979. During that time, population increased by only 2.1 percent, whereas payroll employment, including commuters, grew by 23.8 percent. In our model, each SMSA is treated independently of the other urban areas in California. Furthermore, as noted, home prices in some areas appear to be much more sen- sitive to population pressures than others. Thus home prices in Los Angeles appear significantly more sensitive to demographic pressures than those in Ventura County or the Riverside -San Bernardino SMSA, in part, no doubt, because Los Angeles has much less buildable open space near its urban core than the other two St's. Because of differences in sensitivity, using the same home price/ income ratio does r&- mean that all SMSA's will grow at the same rates between now and the year 2000. Table 4 shows the projections using the home price/income ratio of 12. (It is important to note that relative home prices are the only constraint on population and employment explicitly provided in the modal. Other factors—e.g., the availability of water and energy, or an increase in air pollution— presumably could act as ouxtraints long befione the house price constraint took effect. But the possible impacts of those other constraints is the subject of another research effort, and is not examined here.) Projections Under Home Rice Restraints We:project: the totalpopulation of ..California to be 30.3. million ; persons - by the'"year, 2000. using the home price constraint. (Table -4 shows the population estimates using that constraint; contrasted with Table I, which shows Population; projections without any tioosirtg cost con- straint) if employment is'constrairted ;by, housing uosts, there is likely to be a slowdown in employment growth in the urbanized ;coastal'areas of California. California's new growth centers are -therefore likely to , be inland (e g., Fresno, Bakersfield) rather titan on the coast. Table S presents the recent rates of employment growth for all 17 of California's SMSA's, as well as estimates of employment A Table 4 Estimated Year 2000 Population if Employment Growth is Constrained by Housing Costs Sources: See end -note 1. I 1979 Population 2000 Estimated population SMSA (in thousands) (in thousands) Percent change California 22,694 30,292 33 Anaheim -Santa Ana -Garden Grove 1,874 2,986 59 Bakersfietd 375 733 95 Fresno 485 1,061 119 Los Angeles -Long Beach 7,12b 7,266 2 Modesto 251 538 114 Oxnard -Simi Valley -Ventura 500 679 36 Riverside -San Bernardino -Ontario 1,445 2,875 99 Sacramento 980 2,047 109 Salinas -Seaside -Monterey 281 354 26 San Diego 1,800 2,093 16 San Francisco -Oakland 3,194 3,325 4 San Jose 1,253 2,180 74 Santa Barbara -Santa Maria -Lompoc 295 396 34 Santa Cruz 176 219 24 Santa Rosa 280 427 53 Stockton 320 568 78 Vallejo -Fairfield -Napa 3I2 460 47 Sources: See end -note 1. Table 5 Percentages of Employment Growth 1972-1979 and 1995-2000 Estimated 1972-1979 1995-2000Estimated SMSA Growth rate growth rate California 3.7 L6 Anaheim-SantaAna-Cyarden Grove 7.8 3.2 Bakersfield 41 4.6 Fresno 4.5 4.9 Los Angeles -Long Beach 22 OS Modesto 4.6 4.7 Oxnard -Simi Valley -Ventura 4.6 31 Riverside -San Bernardino -Ontario 39 41 Sacramento 43 4.7 Salinas-Seaside-Montercy 33 2.5 San Diego 4.7 L7 San Francisco -Oakland 3.0 0.4 San Jose 5.1 3,2 Santa Barbara -Santa Maria -Lompoc 4.1 2.6 Santa Cruz 5.3 1.6 Santa Rosa 5.6 19 Stockton 3.0 3.7 Vallejo -Fairfield -Napa 4.2 4.5 Sources: See end -note 1. I growth for the period 1995-2000. The inland areas showing the greatest employment growth in the 1995.2000 period are also those where easily developed land, i.e., relatively flat land. is available and housing prices are lower. The areas of high housing cost are, of course, the slowest growing. i,lajor implications Major implications can be drawn from this analysis of the prospects and ccnsequences of future growth in Cali- fornia. First, almost inevitably Califomia's recent rapid employment growth will slow significantly. also slowing economic growth. This development is of considerable significance to public and private planners alike. Second. if the state attempts to achieve employment growth that is greater than our constrained projections, state and local governments will have to encourage the production of housing units by relaxing restrictions. This would mean un- doing many of the zoning restrictions enacted in the 1970s and increasing the zoning capacity of neighborhoods already built up. Such a change would represent a major policy reversal, and go to the heart of the issue of who has the right to say "no" to new housing development. It would in all likelihood mean higher density housing—e.g., townhouses, condominiums, apartments, and zero -lot line houses—closer to employment centers. It would also mean a reduction in local autonomy over Iand-use decisions. Although the model suggests that California's inland areas wM grow faster than other areas during the next two decades, it is of course uncertain whether the inland areas will in fact accommodate growth by supplying and financ- ing the needed infrastructure of facilities and services necessary to urbanization. In addition to imposing infra- structure costs, such growth would also conflict Po�th the goal of preserving agricultural land. Failure to supply the jnfrastntsture, or to achieve some reconciliation between growth and agricultural land protection policies, will tend to slow the growth of inland areas. The model also raises a host of other issues, including the linkage between rent control and growth controls. The emergence of supply restrictions helped to increase home prices. That condition plus the in -migration generated by a booming economy led to higher rents, and ui.imately to political pressure for rent control. To increase employment further without expanding the housing supply would exac- erbate the problem. In -migration has already helped price existing residents out of their apartments. Rent control can In fact be viewed as a form of growth control, because it prevents new residents from bidding housing away from existing residents. ALS» related to rent control is the overall impact of high relative house prices on the nonhomeowning population. The model implies that population increases would push relative home prices higher, thus noldrrg it even more difficult for nonhomeowners to enter the housing market. Consequently unless there is either a significant slowdown in employment gxowffi, or a substantial increase in housing supply, the hone ownership opportunities are likely to be foreclosed to a large part of the population for the foresee- able future. This could exacerbate the existutg tension be- tween renters and real estate owners. Another development involves government finance Onc of the conditions favoring the creation of our huge state surplus was California's ability to increase employ ment faster than population. While this continued. revenues could increase faster than costs, revenues being more a function of employment, and costs more a function of population. But if population growth catches up to employ- ment growth, state and local government budgets could be severely affected by costs that rise faster than revenues. Concluding Comment California is undergoing major changes. Familiar past trends are unlikely to continue, and a simple extrapolation of recent employment and population trends does not provide a reasonable or plausible guide to the future. Although land -use restrictions have probably limited the amount of congestion and crowding in California, other considerations are now looming in importance. Thus the trade-off between faster employment growth and a better environment will intensify once it appears unWkely that labor force participation rates can grow substantially. The fault lines of the conflict can be identified, even though it is impossible to predict where and when the political quakes will occur. We expect political tensions to grow over the issues of rent control, restrictive zoning, agricultural land protection, the funding of public infra- structure and services, higher densities in existing urban neighborhoods, the overburdening of transportation facili- ties, and the allocation of urban space among different socioeconomic groups. The conclusions presented aie intended to assist in the policy evaluation of these issues. We do not here attempt value judgments on how these issues ought to be resolved. Voters and their elected repre- sentatives will be deciding these matters. In the process, coalitions that have supported increasingly restrictive land• use policies are likely to fragment, and conflict among voting groups is likely to intensify in the coming decades. In essence, until now California has been able to have its cake and eat it, too. We have restricted land use signifi- cantly, while also enjoying very rapid employmrstt growth. This combination is almost certain to find. Accordingly in the near future we must either actively fadlitate rapid expansion of the housing supply, cr accept a dramatically slowed rate of employment growth. NOTES Larry J. Kimbell is Director of Economic Models, UCLA Busi. ness Forecasting Project; and Associate Professor of Business Eco. nomics, Graduate School of Management, University of California, Lor A ngelcs, David Shulman is Assistant Professor of Administration and Economics. Graduate School of Administration, Univcrssty of California, Riverside. For the find report on which thin paper is based, the writers have provided the following acknowledgment: We arc grateful for the assistance of Peter Jaquette, Research Associate at the UCLA Business Forecasting Project. for his assistance in developing the economic models discussed herein and for his important insights in the economic issues involved. We are grateful to Martin ticlmke, Art Packenham, lames Patterson, Isabel Hambright and John Cummins for their helpful comments on an cailiet draft.'rhe study benefited from the substantive and editorial suggestions of Ms. Ruth Galanter. I. All employment and population data were retri-ed (non thu Security Pacific National Bank computerized data bank. The origi- nal source documents are from the Population Research Unit ofthe Califcrnia Department of Finance and the California Employment Development Department. 2. See, for example. Fred E. Case, "Housing Prices and Environ- mental Impact Reports (EIR)," California Heal F.stare !n+dicators, Graduate School of Management, UCLA (Spring 1980). pp. 2 and 4; and Robert Kneisel, "The Impact of the California Coastal Zone Conservation Commission on the Local liousing Market: A Study of the South Coaw Regional Commission." unpublished Ph,D, dis- sertation. University of California, Riverside (December 1979). Both studies reported higher house prices as a result of regulation. 3. See. for example, Jeffrey 1. Chapman and John J. Kiran, "Land U9eConsequcnas of Proposition 13," pp. 95-I24, and David 1980 Shulman, "Proposition 13 and the Spatial Allocation of Economic Activities." pp. 125-137, in Southern California lay.• Review 53: i (November 1979). 4. See. for example. Shap -ell Industries. Annual Report. 1978. Shapeil is California's largest homebuilder with 3 sales volume of 5243 million in 1978 when it delivered 2.026 houses. Its operating profit increased from S21.7 million in 1976 to 562.2 million in 1978. During thai same period operating income as a percentage of saics increased from 16.5 percent to 25.6 percent. 5. See, for example, Scott LcFaver, "µ`ill Success Spoil Silicon Valley?" inPlannrng 46 (4). 22-25 (April 1980). 6. See, for example, John dlenvin. "When Cities Cct Too Pop- ular," iibz*w 126 (6): 72-76 (September 15, 1980); this article discusses the adverse Impact of high housing costs on the Cali- fornia economy. RECENT INSTITUTE PUBLICATIONS Butler. Lewis H., Harold S. Luft, Hekne L. Lipton and Joan B. Trauner Medical Life on the Western Frontier: The Competitive Impact of Prepaid Medical Care Plans in California (California Policy Seminar Monograph Number 6). 19pp 53.50 Leary, Paul M. The Northern Marianas Covenant and American Territorial Relations. Research Report 80-1 55pp 53.75 Mitchell. Douglas E and Laurence Iannacoone The impact of California's Legislative Policy on Public School Performance (California Policy Seminar Monograph Number 5). 24pp 53.75 Roenier. Ruth and William Shonick Private Management of California County Hospitals (California Policy Seminar Monograph Number 4). 27pp f3.75 Schutt:, Howard G., Lloyd D. Musoif and Lawrence Shepard Regulating Occupations in California: The Role of Public Mem- bers on State Boards (California Policy Seminar Monograph Number?). 22pp $3.50 1979 Balderston, Frederick E, I, Michael Heyman and Wallace F. Smith Proposition 13, Property Transfers. and the Real Estate Markets (prepared for the Commission on Government Reform). Re- searchReport 79.1 5 6 p p + Appendices $3.00 Bowen, Frank M. and Eugene C. Lee Limiting State Spending: The Legislature or the Electorate (prepared for the Commission on Government Reform). Rc- snarch Report 79-4 100pp+Appendices f4.50 Dealt, Terry J. and Ronald J. Heckart, eompilert Proposition 13 In the 1978 California Primary; A Pre -Election Bibliography. Omslonal Bibliographies Number 1. 88pp 56.00 Fletcher,Thomas, Dennis Hermanson, JohnTayior, Shirley lienueil and Dean LineWrger Alioealing the Cne Percent Local Property Tax In California: An Analytis (prepartd for the Commission on Government Reform). Research Report 79.7 35pp + Appendices 33.25 hfcWatters, Ann Robertson Financing Capital Formation for Local Governments (prepared for tha Commission on Govemment Reform). Research Report 79.3 51pp $3.00 Scott, Stanley Policks for Seismic Safety: Elements of a State Governmental Program. 94pp 175.75 1978 Ames, Bruce N. Environmental Chemicals Causing Cancer and Genetic Birth Defects: Developing a Strategy to Minimize Human Exposure (California Policy Seminar Monograph Number 2). 28pp 53.50 Bradshaw, Ted K. and Edward J. filakely Policy implications of California's Changing Life Styles (Cali fornia Policy Seminar Monograph Number 3). 30pp 53.50 Cooper. Michael D. California's Demand for Librarians: Projecting Future Require- ments. 128pn 56.50 Eckbo, Garrett Public Landscape: Six Essays on Government and Environmental Design in the San Francisco Bay Ares. I35pp 57.75 Cater. Paul W., Ronald B. Roble, John T. Knox and Norman Y. Mineta Four Persistent Issues: Essays on California's Land Ownership Concentration. 1Vater Deficit;, Sub -State Regionalism. and Congressional Leadership. 79pp 55,75 Mullins, Phil, Thomas O. Leatherwood and Arthur Lipow, eds. Political Reform in California: Evaluation and Perspective, Research Report 78.3 160pp 56.50 Nathan, Harrict and Stanley Scott, eds., with I 3local authors ]Experiment and Change in Berkeley: Essays on City Politics, 1950-1975. 501 pp $10.75 Phillips. Kenneth F. and Michael B. Teitz liuusing Conservation in Older Urban Areas: A Mortgage In- suranm Approach. Research Report 78-2 39pp 53.50 Pyk, David H. Changes in the Financial Scrvices Industry in California (California Policy Seminar Monograph Number I). 17pp 53.00 Stebbins, Kobcrt C.. Theodore J. Psp,enfuss and Florence D. Amamoto Teaching and Research in the California Desert. Res arch Report 78-1 26pp f3.00 Weeks, Kent M. Ombudsmen Around the World: A Comparative Chart. 2d. ad. 163pp 57.50 Monographs, Bibliographies, Research Reports and a full list of Institute publications are available from the Institute of Govern- mental Studies. 109 Moses liall, University of California, Berkeley, California 94720. Checks should be made payab;e to the Regents of the University of California. Prepay all orders under f30.California residents add 6% sales tax; residents cf Alameda. Contra Costa and San Francisco counties add 6'h% sales tax. Prices subject to change. Note to Readers While reading recent issues of the Public Affairs Report, have you thought of questions, suggestions, criticisms or comments that you might like to consider sending in to the editors? Do you have any observations about the choice of topics. or their treatment in the pages of the Report' 1}o you have any advice or suggestions regarding future issues? Letters from readers and requests for copier arrive regularly, but by including this note, we try to keep open a "reader' window" that will give us a better sample of what subscribers think about the Public Affairs Report, and what directions it might take. Your comments may also provide substantive observations that cxauki be acknowledged or excr:rpied in future issues. -The Editort c ® Institute of University of bLurl Business and California, Economic Research Berkeley CENTER FOR REAL ESTATE AND URBAN ECONOMICS WORKING PAPER SERIES WORKING PAPER No. 94-173 HODS ING ALLOCAT ION AND METROPOL I TAN CEVELWMENT BY JOHN M. QU I GLEY T1ao� pepen� are preNm&rary .. *k1 ; tlNa pacpos0 iS { , . `to: sprmilats discussion and camnerC Therefore. they -we not io be rxed or quoted in any pubkotion without the express permission of the author. GRADUATE SCHOOL OF BUSINESS ADMINISTRATION Reduced local government support in the wake of Proposition 13 (limiting property tax rates in California) and the federal cutbacks in grants for local services have made it difficult to finance expansion of the infrastructure and public services necessary for housing development. Because communities can not increase property taxes enough to pay for local services needed by new residents (schools, sewers, etc.), fees are imposed on new housing development, increasing v housing prices. Because }locate; revenues are•~ roughly _proportional .to housing values and''the 'dezriands`"for' services - are .. >roughly-proportiorta' 'to the number . - of -l ausaii ii local governments have strong incentives `to-adopt zoning regulations Th requiring ..targe amounts °"of hoersing cans2mption. ese =regi Iatfons, = sa=called° scal � - zoning; n are intended ` to re. ;new`residents 'to'-=consume (and''to =pay"p _j taxes yan).":mare reap estate than-they 3d`otiierwise choose. As a result, much of the new development excludes housing that could serve lower income persons. New rental construction approved at the local level tends to be small units that cater to older couples (who use relatively few local services); new housing for young families (who use many public services, particularly schools) is less widely available. The problems created by Proposition 13 are unresolved; in consequence, local governments continue to resist rapid development of housing. 20 Institute of University of Business and California, Economic Research Berkeley CENTER FOR REAL ESTATE AND URBAN ECONOMICS WORKING PAPER SERIES WORKING PAPER NO. 90--173 HOUSING ALLOCATION AND METROPOLITAN DEVELOPMENT BY JOHN M. til l GLEY Ttiesa ps�trs aetp prsuminary : ;„. 4t±rtaetwe a+u purpose is Wstittaipl�s diudisabn auxt '' .. -COmme t,1herefore. they '-are-reed be cited or quoted in any Publication without the express permission of the Sutter. GRADUATE SCHOOL OF BUSINESS ADMINISTRATION locations. The most frequently mentioned reason for site selection was the proximity to th he residences of key employees and managers. Two fifths of firms mentioned this factor, and another 15 percent were concerned with proximity to the workforce. About 20 percent of larger firms (with more than fifty employees) were concerned with proximity to the homes of key employees, while 40 percent were concerned with the location of the workforce, Transportation access and the cost of space concerned one-fourth of all firms, and were of somewhat greater importance to larger fi3=. Further, interviews with tenants in newly constructed space indicate that firms seek greater labor force availability, either through reduced commutes or by capturing secondary earners, and they seek a more highly educated local work force. Significantly, interviews with many of the largest developers and builders of the facilities suggest that they had chosen these building sites for similar reasons in anticipation of the demand. b. Competition Between the Bay Area and Other Regions: Housing Affordability The San Francisco Bay Area is well Down for its natural resources and mild weather, its relatively low levels of pollution and congestion, and its striking architecture, making it one of the most desirable locations in the U.S. in which to work and live. Per capita income in the region is 18 sixty percent above the national average, and it has attracted the highest average educational level of any US local labor market. Not surprisingly, housing prices are high. By August 1988, the median price of owner occupied housing in the region was $216,000, and prices had increased by 71 percent in six years. 13 In six years median housing prices increased from 1.8 times the national median to almost 2.5 times the median for the US as a whole. At current prices, only about one household in eight already living in the area could afford to purchase the median priced house, given widely accepted rules of thumb. In large part, of course, high housing prices are to be expected, given the desirability of the region, There are, however, strong indications that the regional price level. for shelter has begun to act as a deterrent in the competition for new business activity. 14 to`increase°housing grf", above competitive levels: here the answer'is almost-certainly.yes: 13 These and other background statistics are discussed in Hird, et al, op cit, and Kenneth T. Rosen and Susan Jordan, "The Ban Francisco Real Estate Market," Berkeley, CA: Center for Real Estate and Urban Economics, 1988. 14 For a recent example from the popular press, see Fortune Magazine, Oct 2, 1989. is Reduced local government support in the wake of Proposition 13 (limiting property tax rates in California) and the federal cutbacks in grants for local services have made it difficult to finance expansion of the infrastructure and public services necessary for housing development. Because communities can not increase property taxes enough to pay for local services needed by new residents (schools, sewers, etc.),_ fees are imposed on new housing development, increasing housing prices. Because local revenues are'=rot�ghly s -r-. �--• .� ra. -. a:_2,Y•; a --.a-"' __--mss. ;_ , proportional to hot7slnq values and the-- -demands - for services - are. -_roughly propsrtioial to the number of households, local = --governn►ents have strong`incentives to'adopt zoning regulations requi=uiq large wamoun'ts 'of 'Housing uconsuznption. These aegulations, so-called _ f isca3` zoning, " are inten ` a� requte.enew `residents=_to consume {and tol- --payproperty taxes estate than; they would `o choose. ".- As a result, much of the new development excludes housing that could serve lower income persons. New rental construction approved at the local level tends to be small units that cater to older couples (who use relatively few local services); new housing for young families (who use many public services, particularly schools) is less widely available. The problems created by Proposition 13 are unresolved; in consequence, local governments continue to resist rapid development of housing. J 20 In addition to these direct policy driven causes of high 0 Monopoly power can be exercised by developers who benefit from restrictive land use regulation, which limits the amount of land available for development and makes controlling local land markets easier. Complex: administrative procedures, lengthy application -.periods,, and' other measures Ghat typify Bay Area hocal development polcies.,`can-:.induce .monopolistic „� control of Iocal land markets. Credible studies of development -have'. %und.that the `excess profits were largely attriliutable td -' housing supply and the lack of competition 16 In other suburban areas, the lack of developable land and high development fees have given dominant control of the housing market to a few large developers. 15 See David Dowall, The Suburban Squeeze, Berkeley CA: University of California Press, 1984. 16 See David Dowell, op Cit. 21 These indirect effects may exert a powerful influence over local land use and development- Any policies hoping to improve the present housing conditions in the Bay Area must recognize these important, though subtle consequences of such policies. Environmentalism and local land regulations preventing rapid growth are supported by many Bay Area residents, especially since limitations on property tax rates have made it more difficult for existing residents to "profit" from additional housing development- Attempts to change this pattern are not likely to be initiated by local governments or their constituents. The ultimate source of the problem is the balkanized pattern of bUilding permit and land use regulation. 22 THE IMPACT OF SUBURBAN GROWTH RESTRICTIONS ON U.S. HOUSING PRICE INFLATION, 1975-1978' David Segal and Philip Srinivasan Oxford University and Harvard University The paper estimates a simultaneous equations model of housing price inflation 1975-1978 for a cross-section of51 metropolitan areas. A two-stage least squares pro- cedure is used to estimate the demand-side and supply-side determinants of price changes. One --of the major sources of inflation is shown to be a variable reporting suburban grow;#opstepie:fnction.:of: potentiUly dey eloped land just beyond the inatgln=of tiriiait=settietnent: that;is.s tiestered from growth, Nearly two-thirds of. tht SMSAs'.in:atr sample tiad growthrestui tions 'wthan average of 1.2% of avail- able,suburbanjpisd:.sef:off timits:to growth- Som a cities barred growth from as much as 30 to 40% of the nearby surrounding land. Su ch controls were found to have con- tributed significantly..to iinftation., . The-growwth Controlled cities experienced an infla- tion` rate that wai �boiit 17"1�hiahef:than those that did notrestrict, ceteris paribus'= 125% instead of 10.8% annually, 1975-1978. As the purchase price of new and existing homes began to dip in early 1980, hous- ing prices nationwide had just completed a period of their most dramatic increase since the start of record keeping. During the period 1973-1979, home prices increased at an average annual rate just short of 10'/6; toward the end of the period, home inflation accelerated, reaching 13.41/6 in the last year. Such figures mask an enol`nious amount of geographic variation. In Denver and Phoenix, prices rose at an annual rate of 20 to 25%between 1977 and the spring of 1978; price increases were only slightly tower in Chicago, Dallas, and Houston. Some Phoenix builders reported selling out homes even before streets and curbs had been laid in. Boom conditions in some California cities had no parallel (Grebier and Mittel- bach, 1979). At the other extreme, home price increases in Jacksonville and Rich- mond averaged 5.3° annually, and in Milwaukee a mere 4L20/6. A literature is emerging on ?he forces behind the housing price inflation of the mid - to late -1970s, on why price increases in this sector have tended to exceed increases in the general price level by 2 percentage points or more in all but two of the years since 1974. Frieden, Solomon and Birch (1977), as well as Hendershott and Hu (1979), have stressed the role of inflationary expectations in inducing a higher demand for housing than would otherwise have been the case. Schwab (1979) beiieved capital market imperfections such as the institutional prevalence of the level -payment mort- gage forced many households to choose differently among three goods, present con- sumption, future consumption, and housing, in favor of the last. Remarkably little has appeared on the causes of variation in housing inflation rates across cities. We address this question by means of a straightforward comparative statics model, reporting the partial effects of demand and supply forces on housing price outcomes at two points in time, The model is tested using observations on 51 metropolitan areas for 1975.1978. 14 Urban Geography, 1985, 6. 1, pp, 1426. Copyright U 1985 by V H. Winston & Sons, Inc All riahts reservsvi SUBURBAN GROWTH AND HOUSING INFLATION 15 We find that dgmand-side factors—variations in the rates of income, population and mortgage rate changes within our sample—had a significant influence in housing price increases, No less important, however, is the role played by suburban growth restrict tions. They explain as much as 40% of the variation in urban housing price inflation unexplained by demand-side factors. As a class, growth -restricted cities entered the mid-1970s with a hatf percentage point higher inflation rate tha.. unrestricted cities, ceteris paribus. Moreover, every 10% of a city's potential suburban land that was set off-limits to growth during 1975-1978 contributed an additional 1.0 percentage point annually to its inflation mte in home prices. Taken together, these considerations meant that the average growth -restricted city, with more than 12% of its suburban land off-limits to growth, experienced an inflation rate in housing prices nearly 2 per- centage points above an average unrestricted city. The impact of growth restrictions, however, was not linear: Larger fractions of sub- urban land withdrawn from growth caused increasingly higher inflation rates. Cities reporting more than 2(?% of outlying land growth restricted added more than 6% to their annual housing price inflation rates, ceteris paribus. In the balance of the paper, we describe our underlying model, the data, and the empirical results. THE MODEL An appropriate means for analyzing the effects on housing prices of unanticipated exogenous demand shocks is the stock -adjustment model. Once we know the nature of the adjustment process and the magnitude of the adjustment parameter, it is possi- ble to identify short -run equilibrium prices for all points in time after a shock. If, on the other hand, there are no demand shifts that are not anticipated, then a statics model is quite appropriate for examining equilibria at different points in time. We do allow for the possibility of unanticipated demand shifts within the context of a comparative statics model by assuming that exogenous changes in demand cause equilibrium prices to move smoothly, i,e., we assume that the stream of exogenous demand increases is such as to cause equilibrium prices to grow smoothly. Mathemati- cally this requires that when we solve lagged -adjustment equations for short -run equ i- librium housing prices at t and t + 1, the actual values are ahead of the target values by the same percentage in both time periods. The above assumptions allow us to present a dynamics process as a comparative statics result. This is illustrated in Figure 1, in log -log, price -quantity space. Because slopes of demand and supply functions in such a diagram are elasticities, a demand shift between times t and t + 1 that leaves demand elasticity unchanged will cause D t+ 3 to be parallel to D,. There are two issues of particular interest in this paper. One is whether the supply function for growth -restricted cities, S or S', has the same y-oxls intercept as that for cities that do not restrict suburban growth, S. A second is whether the supply elastici- ties of housing are the same for both kinds of cities. This is a question of whether the dopes of S or S' and S are the same. Were this the case, cities with growth restrictions would have the same inflation rate as those without. This is because c is the same vertical distance above d as is b above a.' On the other hand, if supply elasticities differ between th � cases of growth restrict iV W U Ir Q_ 0 O J SEGAL AND SRINIVASAN In Pit+1[.............. N In Pt+1 In Pt S� LOG QUANTITY —S Dt S Dt+1 Fig. 1. Housing market inflation: A comparison of cities with and without growth restrictions. tions versus no -growth restrictions (S' versus S), the price inflation rates will alsodiffer. The mechanism causing this to happen is an interaction effect between the housing supply function and a variable representing growth restrictions: The fact of interaction causes the slopes of the supply function (the elasticity in Fig. 1) to be different be- tween growth -restricted and -unrestricted cities. We shall be interested to test whether and how growth restrictions affect housing price inflation, i.e., whether the interaction effect is statistically important; for the present, we suggest as a working hypothesis that such effects do matter. Specifically we would expect restrictions to increase the rate of inflation, ceteris paribus—that they render (c' - d') > (b - a). Also we shall be interested to see whether the two supply functions cross the y-axis at the same point On a priori grounds we might expect that the supply curve for growth -restricted cities would have a higher intercept, reflecting a higher initial price level for these cities. Before proceeding with a discussion of the functional form of equations to be esti mated, some comments are in order regarding a model structure that might justify the above hypotheses. Consider two cities of identical size and spatial structure except that one has growth restrictions while the other does not. Such restrictions might take the form of environmental ordinances withdrawing from development a portion of the annulus of open space just beyond currently outlying settlements. This is shown in the schematic diagram of Figure 2 where the land into which suburban settlement might ordinarily expand is shown as the outer ring or annulus, and the area that is sequestered from growth is represented by the shaded portion. The impact of an increased demand for housing on existing patterns of density and SUBURBAN GROWTH AND HOUSING INFLATION 17 ' Existing Settlement Fringe i Annulus cf Potential Housing Growth Fig. 2 Area of potential suburban land restricted from growth. prices (and hence on mead density or price) is well known (see Muth, 1969; Mills, 1972). Growth restrictions of the sort described, by limiting yrowth at the margin of settlement, thus upset the trade-off between travel time and for size. Households will pay higher prices for interior locations to avoid the extra commuting time of living b e yond the growth -restricted area Accordingly, lot prices at all interior locations, as well as at locations in the unrestricted portion of the annulus and beyond what would be the new margin of settlement in the absence of growth restrictions, will be bid to positions above their no -growth -restriction equilibrium levels. Moreover, in terms of a dynamic model of the statics version of one presented here, housing prices in the growth -restricted area will be higher than in its unrestricted counterpart, both initially and at all future points in time. Assuming the substitution elasticity between land and noniand factors of production is less than infinite, growth restrictions will lead to an unambiguous increase in average density and housing prices. What estimating procedure is suitable for testing the impact of suburban land -use restrictions? A model analogous to Figure 1 involves a pair of demand equations at two points in time and one or more supply equations, dependingupon whether growth• restricted and unrestricted cities have separate supply functions. Consider the following two -equation model: (1) Demand equation: In P t = ao - a j In Q t + a2 In Z t t Supply equation: In Pt = bo + bi In Qt (I+cGt) + bz In Z2 t + b sGt where Pt is the average price of housing in a city at time t, Qt is the size of the city's housing stock, Zjt is a set of demand-side variables other than Qt that vary across cities, ZZ t represents supply variables other than Qt, and Gt reports the fraction of a 18 SEGAL AND SRINIVASAN city's potential suburban land that is removed from growth—the shaded area in Figure 2 as a fraction of the area of the entire annulus. Our goal is to study neither the level of housing prices across cities nor the differ- ence in such levels at two points in time, but instead the percent change in prices. If the various demand and supply elasticities are unchanged over time, a reasonable as- sumption for the short run (and also a testable hypothesis), we c2n imagine a pair of difference equations, one demand and one supply, that take the difference of each equation of (1) at two points in time, t and t + 1: (2) p=ao-aiq+a2Zi p=bo+blq(1+cG)+b2z2 +b3G where p = (In Pt+1 - In Pt); q = (In Qt+I - In Qt); etc. The lower-case symbols of equa- tion (2) represent percent changes in the variables represented by capital letters in equation (1) (see note 2). The object of the analysis, as mentioned earlier, is two- fold: to learn whether the supply function for growth -restricted cities has a different elasticity from that for unrestricted cities (e ? 0); and to learn whether the y-axis in ter. cepts are the same for the supply functions for growth -restricted and unrestricted cities (b3 = 0). As we note in the next section, the scarcity of high quality time -series data on changes in the levels of suburban growth restrictions in different cities led us instead to consider a variable reporting the average percentage of developable suburban land put off-limits to growth during the inflationaryperiod under study, G. in the transi- tion from equation (1) to equation (2) this procedure clearly would cause the G in (1) to drop from the difference equation of (2) because it is a state variable equally present in both level equations. Above we argued that the presence of a variable representing growth restrictions affecting a pair of growing cities (a} that start with identical size and spatial structure and (b) that otherwise grow identically can be expected to increase housing prices in the restricted city at all periods of time subse- quent to the restriction, above their equilibrium levels in the absence of the restric- tion. Accordingly we include G in only one of the levels equations cf (1) so that it survives when the subtraction is performed. Including G in equation (2) rather than (inGtti -InGt) means that testing to ascertain whether supply elasticities differ between growth restricted and unrestricted cities (slope cf S'= slope of S in Fig. 1) cannot be done by examining the coefficients of G alone, when G is entered independently of the other variables ofequaiion (2), its coefficient would shift the location but not the slopes of the functions of Figure 1. When G appears in the supply equation cf equation (2), its coefficient reports whether the y intercept of S' differs from that of a. In addition we must took at the signs and significance of the interaction between G and q, in an equation that includes q sepa- rately. Estimating the parameters a, b and c of equation (2) in reduced form is not an attractive option. This is because one of the variables, G, remains imbedded in the parameters of the reduced form equation even after q, or p, is subtracted out We instead rearranged the variables of the supply equation of equation (2) so as to linearize it, and we then used the two-stage least squares (2SLS) estimating procedure: SUBURBAN GROWTH AND HOUSING INFLATION 19 a (2') p = ao = alq +a2zi p=bo+bigl+cG)+b2z2 +b3G=bo+biq+dx+b2z2 +b3G where x = Gq and d = b c. The coefficient cis estimated indirectly: c= d/bl. Fieller bounds may then be computed for c. The vector of exogenous demand variables, zI, has three components in the empiri- cal work below: percentage change in per capita urban income, y; percentage change in urban population, n ; and percentage change in mortgage interest rates, r. We hypothe- size that a, , 0. if changes in y and n are fully foreseen by housing suppliers, the first -order partial derivatives will be close to zero. Otherwise they will be greater than zero, as supplier quasi -rents accrue. Whether such changes are fully foreseen can be learned empirically. We can expect aP— tobe negative. Because mortgage money is a good complemerr, ar tary with housing purchases, decreases in r will tend to shift the demand function to the right. Again, whether we move along a shot cr long run supply curve depends upon whether the decrease or increase in r was anticipated. Z2, the set of exogenous supply variables, has but one component, construction costs, k The hypothesis of .3P. < 0 requires no elaboration. a THE DATA The empirical analysis of the next section is performed on observations on a cross- section of 51 metropolitan areas for which data were available during 1975-1978. The areas are listed in Appendix A along with observations on the 19751978home price increases and the growth restriction variable. The latter variable is the focal point of much attention in this paper and is discussed at length below. First we offer a brief description of the other variables and their sources. Home Prices Federal Home Loan Board data were used, reporting a weighted average of current doltar prices for new and existing single-family houses sold in each of the SMSAs in December 1975 (P7$) and in December 1978 (P78), Only conventionally financed homes are included by the FHLBB. Data on prices in the 30 largest SMSAs are to be found in the monthly publication, Terms on ConventionalHome Mortgages (Washing- ton: FHLBB, 1974 through 1978). Observations on additional SMSAs are made avail- able by the FHLBB for a small fee. We employed three demand-side and twc supply-side variables which are presented in sequence. Income Many housing market studies have concluded that permanent income is the best income measure as an argument of housing demand, Such data are not readily avail- 2O SEGAL AND SRINIVASAN able for metropolitan areas and we used current per capita personat income as a surro- gate. The data are drawn from the Survey of Current Business. We used average annual data, comparing 1975 personal income per capita (Y75) with that for 1978 (Y78) - Population The number of households is the measure best suited to study the population com- ponent of housing demand, particularly households in the demographic categories most likely to occupy single-family dwelling units. These data cannot be obtained annually for metropolitan areas so we tried several proxies. The most successful was straight population taken from the Census Bureau's P-26 Series, for ) my 1975 (N75) and the same month of 1978 (N7s). We also tried migration data from the same p u b lication, as well as numbers from annual issues of the Rand McNally Commercial Atlas. Mortgage Rates Our final demand-side variable relates the price of a good complementary with housing, mortgage money. We used FHLBB data from the same source as the price data reporting contract mortgage rate of interest plus lender fees and charges in December 1975 (1175) and December 1978 (R78). Although 19 states have usury laws, the post -1975 recovery saw a mortgage climate sufficiently mild to allow interest rates to move unconstrained.3 On the supply side our primary variable is the Boeckh index of nonland construc. tion costs, with Milwaukee, 1967 = 100 as the numeraire. The index captures unit labor and materials costs in single-family construction and allows for intercity cost comparisons over time. We used numbers for July 1975 (K7$) and )uly 1978(K78), published in Boeckh Modifier of Construction Cosmo by the American Appraisal Com- pany (Washington: 1975, 1978). We assume that the labor and material construction costs are horizontal with respect to the stock of housing in any particular city, imply- ing a horizontal supply curve for these costs in that city. Growth Restrictions By far the largest part of our effort in data collection involved gathering numbers on the fraction of otherwise available suburban land during 1975.1978 that was put off-limits to development. The early part of the decade saw a major increase in the number of communities employing growth management techniques based on environ- mental or fuel allotment considerations. In some cases, development was stalled be- cause of moratoria on water supply (particularly in the South or West) or on sewer connections. In other cases the unavailability of larger allotments of natural gas was a factor. (Chlyrecently has the moratorium on natural gas tie-in been lifted from some of the areas surrounding Baltimore.) In several metropolitan areas there has been public acquisition of open -space land through fee simple acquisition or annexation. in still others, rationing devices such as building permit restrictions and zoninghave been used to sequester some land from growth. The data were gathered through interviews with the staffs of Regional Councils of SUBURBAN GROWTH AND HOUSING INFLATION 21 Governments, or of rdgional and local planning agencies, in each of the 51 areas. The purpose of the interviews was to ascertain, for a given metropolitan area, the extent of land -use controls within the jurisdiction of various agencies prevailing during the 1975-1978 period. Data were collected both on the percentage of land in otherwise developable sub- urban land removed from growth in 1975 (+), in 1978 ((&), and on average throughout the period (G). As is shown in Appendix A values of this variable ranged from 0 (about a third of the areas) to 43.5 (Sacramento). The average percentage of land removed from growth in areas reporting growth restrictionswas 12.7%; for a fifth of the areas, land sequestered from growth ran over 15%throughoutl975-1978. For several specifications in the next section we use dummy variables for the per- centage of land that i s growth managed: 0-5% (Go) ,5-10% (G 1), 10-15% (G2 }, 15-20010 (GA and above 20% (G4)- Quantity G4). Quantity ofHousing Units As one of the two variabies, along with price, endogenous to our model, the Q (or q) variable plays a central role. As noted above, eliminating the quantity variable in a reduced -form equation poses a linearity problem, so we left the variable in the anal- YsIs• Unfortunately, annual data on the stock of housing units by metropolitan area are not available, so it was necessary to estimate them. The q variable has a flow term in its numerator, net additions to stock between 1975 and 1978, and astock term in its denominator, the quantity of units in place in 1975. Data were more readily avail- able for the numerator than the denominator; the latter had to be constructed from 1970data and figures on population shifts in the interim. The rather crude approach to estimating the denominator of q is mitigated by (a) the relatively higher quality cf numerator data and (b) the relative insensitivity of estimates of q to errors in the denominator. The mean value of q in our sample was .071. Accordingly, a IN measurement error in Q7S for a city which experienced a change in housing stock 1975-1978 close to the mean will result in a 0.7% error in the estimate ofQ for that city. In estimating Q75 we took 1970 Census figures for the total number of housing units in an area, augmented them by new construction figures (permit data coming from the Commerce Department), and decremented them by our estimates of with- drawals (equal to demolitions plus net change in vacancies) from the housing stock. Lacking good data for withdrawals, we used as a proxy for changes in vacancies 1970- 1975 figures on net household formation (net natural increase plus net migration), drawn from the Current Population Survey. The numerator data, Q78 - Q75, were largely drawn from the "Summary of Housing Characteristics" in the Annual Housing Survey. Results Table 1 presents the results of four models that were tested. The first two models suppress a separate constant term for the supply firnction for growth -restricted cities, forcing S' to intersect at the y-axis. The second pairof modelsallows for the inter- R Model Number: 1 TABLE 1.—TWO STAGE LEAST SQUARES REGRESSION RESULTS 2 3 4 Variable ,, .; Demand Equation Supply Equation Demand Equation Supply Equation Demand Equation Supply Equation Demand Equation Supply . Equation CONST - .084* ' .027t ' .084* - .074t ,016 .011 q. -1.171 .74t -1.031 1.981 - ,017 .7171 - -890• 2.T Y 1.68• 1.67► 1-58* 1.66• n, 2.821 2 54t .536 2.27* r 1.46• -1,47* -1.49* -1,47* G .005* X, .297• .069• k; 2.46** 252t 1.48** 1-44** Gs .009 G: -030 :Gs .091 :G4 .172 -14.6* -3.571 X1 -22.2* -3.31* Xs .18.3t -3.72* X♦ 62 -1.28 standard error of dependent -211 .211 ,211 -211 ,211 .211 .211 .211 variable standard error of regression ,. .198 .212 ,196 .421 .210 .148 .196 .154 t Coefficient greater than standard error. • Significant at 5% level. ** Significant at 1%levet. SUBURBAN GROWTH AND HOUSING INFLATION 23 L section to be other than at the y-axis (by including the growth management variable separately). Models 1 and 3 treat growth management as a continuous variable; mod- els 2 and 4, in the form of a dummy variable, as described in the previous section. The coefficients of the demand equations appear to shed some light on the question of anticipated versus unanticipated demand shifts. Income and mortgage rate changes have the right sign and are significantly different from zero in all models, suggesting the possibility that shifts in these variables were less thzn fully discounted in the mar- ket place. That is, when shifts are futly anticipated by the market, the new equilib- rium positions tend to be only slightly above the old ones as demand shifts along a highly elastic long -run supply function. The inflationary impact of such shifts under these circumstances would be minimal and might not differ from zero. While inflationary consequences of income and mortgage rate variation in our cross- sections sample are in evidence, suggesting less than full anticipation and some amount of supplier quasi rents, the impact of poputation shifts is more ambiguous, While the elasticity of the population variable, n, has the same order of magnitude in three of the four models, it is statistically significant in only the last. The coefficient has the right sign in all nnde]s. The sign and significance of the q variable follows much the same pattern as that of n. It is Interesting to note that the pattern of demand elasticities is very much what one might expect on the basis of the published literature (e.g., De Leeuw, 1971). Price elasticity is about unity and income and interest efasticities about 1.5. We found support for our assumption of no change in the demand elasticities be- tween t and t + I (parallel shift of DD in Fig. 1). We did this by comparing (a) the sum of squared residuals for the separate equations for 1975and 1978of equation (1) in which the equations were partitioned by time period with (b) the sum of squared residuals for the pooled data set and, employing an F test, could not reject the hypothesis of a constant elasticity at standard levels of significance.4 On the supply side, construction costs, k, have the right sign and magnitude in all of the models and a high tevel of significance in alt but one model. The housingsupply variable, q, has the right sign in all models, but the desired level of significance in only one. Jhe growth management variable, whether interacted with q (as in x, x, , ... 7x4) or standing on its own to report y intercept information (5,G 1, . .. ,G4), almost everywhere has the anticipated sign, magnitude and significance. Because of the sta- tistical significance of G, we learn that models 3 and 4 are better specified than the first two and that when G is specified as a continuous variable the y-axis intercept of S' is .005 above that of S. This suggests that growth -restricted cities have half a per- centage point "head start" on inflaticn over unrestricted cities, before the interaction or elasticity effect (S'versus S) is reckoned. What is the "elasticity effect"? This, we recall, is the tstimate of c, reckoned as the coefficient of x divided by thzt of q in the supply equation. When growth manage- ment is viewed as a continuous variabie—models i and 3—this computation yields 0.4 4 and 0.1, respectively. Becaux the former model forces the intercepts of S` and S to be equal it makes c higher than would otherwise be the case. Abdel 3 has the greater credibility. The estimate of c = 0.1 from this model implies that every 10%of potential sub- urban land sequestered from growth (as in Fig. 2) causes inflation to be one percent- 24 SEGAL AND SR INIV ASAN age point higher, ceteris paribus, or 1.5 percentage points higher when the force of the constant term is included. On average, growth -restricted cities are seen to have a 1.7 percentage point higher inflation rate than that of unrestricted cities when demand - and other supply-side factors are properly controlled for. On average, growth -con- trolled cities had an inflation rate In housing that was 3.0 percentage points higher than that for uncontrolled cities. Our estimation procedure, which brings this down to 1.7 points, ceteris paribus, suggests that the simultaneity problem, addressed by our technique, is not trivial. Fieller bounds for c were estimated at the 95% confidence level and found to be {.018. This suggests the possibility of a 20% error in our elasticity measure at that Ieve I of confidence, Models 2 and 4 suggest that the impact of G may not be linear and that in percent- age terms there may be a greater inflationary impact from growth restrictions at higher levels than a linear model would suggest. The percentage gaps in the coefficients of X1, ... , x4 increase more rapidty than the percentage of land sequestered from growth. As a result, cities having more than 20% of their potential suburban land re- moved from growth have a housing inflation rate about 6 percentage points above unrestricted cities, ceteris paribus. On the basis of the analysis here we cannot conclude that growth restrictions are bad on welfare grounds. Some would argue that there are benefits from controls such as social costs associated with growth that do not have to be borne. Certainly home- owners in communities that put controls in place gain from their capitalization effect on property values. Would-be owners who are priced out of such communities by the inflationary aspects of controls are losers. A careful study is needed before the wel- fare effects of suburban land -use restrictions can be fully assessed, One conclusion that foliows from the immediately previous comment is that there may be an overstatement in housing inflation rates such as are reported for the period 1975-1978. While it is true that growth restrictions lead to higher prices, it may well be that there is a quality differential between suburban housing located in or near growth -controlled communities and housing that is not The Tiebout model suggests that people with a choice who choose to migrate to communities in the former cate- gory and topay higher prices for the assurance of lesser crowding later OR are getting a product, or an attribute of one, not picked up in the CP1. NOTES 'The authors gratefully acknowledge helpful comments from Brian J.L. Berry, Adonis Yat. chew, and Thomas L Steinmeier. NSF Grant SOS 79-09370 helped support the research. 'That the vertical distance separating b and a is a good proxy for price inflation between t and t + 1 in the no -growth -restriction case can be easily shown by: Pt+1 Pt + Pt+1 - Pt 't+1 ` Pt Pt.�i - Pt In Pt+j . In Pt = to ( Pt ) = In ( Pt )=In (i + Pt`_'.j Pt "'Residenrial Construction: Three Years of Recovery," in Srivey of Cdjrrent Business, June 1978, pp. 18-28. 'In the case of the pooled regression for the two pairs of "levels" equations in (1)—for 1975 and 1378—we restricted variances of the two residuals to be the same. It should be pointed out >.hat this is not a necessary assumption. We might just as well have assumed different variances but SUBURBAN GROWTH AND HOUSING INFLATION 25 identical parameters. In point of fact, the assumption of common variances was not unreasonable —the sum of squared residuals was approximately the same for the separate equations of (1). LITERATURE CITED De Leeuw, Frank, 1971, February, The demand for housing: A review of cross-section evidence. Review of Economics and Statistics, Vol. 53. Frieden, Bernard, Solomon, A, and Birch, D., 1977, The Nation's Housing, 1975 to 7985. Cambridge. Grebler, Leo and Mittelbach, Frank G., 1979, The Inflation of House Prices. Lexing- ton, MA: Lexington Books. Hendershott, P. and Hu, S, 1979, Inflation and the benefits from owner occupied Housing,.. National Bureau cf Economic Research Working Paper No. 383. Cam- bridge. Mills, Edwin 5.,.1972, Studies in the Structure ofthe Urban Economy. Baltimore: The Johns Hopkins Press, for Resources for the Future, Inc. Muth, Richard, 1969, Cities and Housing. Chicago: The University of Chicago Press. Schwab, Robert M, 1979December, Inflation Expectations and the Demand for Hous- ing (Available from author.) Short Papers The Effect of Growth Control on t h e Production of Moderate -Priced "ousing SeymourL Schwartz, David E. Hansen, and Richard Green INTRODUCTION nity of Petaluma, California (Schwartz. Hansen. and Green 1981). detected statisti- cally significant price increases for new houses by comparing price changes for stand- ardized houses in Petaluma to price changes in two nearby communities.' This statistical analysis did not, however, tell the entire story about the effects of the program on housing production and OR housing opportunities, es- pecially formoderate-income h6mebuyers. To many policymakers the distributive consequences of growth control, especially those affecting lower-income households. are of great concern. The important questions these policymakers want answered are: What is the effect ofthe growth control program on the availability of lower-priced housing. and how are the housing prospects of moderate - income families affected? To provide the in- formation with which to answer these ques tions, vw examined the characteristics (price and floor area) of the houses actually built in Petaluma and a neighboring comparison city—Santa Rosa- between 1970 and 1976. In this note we present the results of this analysis and discuss the reasons for the ob- sen,ed differences between cities. First, we describe the characteristics of Petaluma's growth control program and discuss our methods.' Land Economics. Vol. 60. No. t. February 1983 0023.7639/94rtl0t-0110 S1.S0i0 L 1983 by the Board of Regents of the University of Wisconsin Svstcm PETALUMA'S GROWTH CONTROL PROGRAM Petaluma was a small agricultural trading center (14,035 population in 1560) until the mid-1960s. when rapid suburban growth from Sart Francisco (40 miles south) and Marin County. spread to Petaluma. This growth. which increased Petaluma's population to 24,870 in 1970, strained the capacity of the sewerage system and caused serious over- crowding in the schools. events which were largely responsible for Petaluma's adoption of a pioneering growth rate limitation pro- gram in 1972. Petaluma sought to limit its growth rate by establishing housing quota cf500 new units per year (single-family plus multi -family) from 1973 through 1977 (City of Petaluma. 1972). Developers competed in an allocation process in which a citizens review board eval- uated subdivision proposals according to two major sets of criteria: one to ensure that ade- quate public services were provided by the Pxe ::: - Schwartz is with the Division of Environ- mental Studies. University of California. Davis. and professors Hansen and Green arc with the Department of Agricultural Economics of that university. They thank Richard Belzer and Peter J. Hunter for gathering and processing the data. and Michael Johnson for help- ful suggestions. The research was supported by grants from the Public Service Research and Dissemination Program. and the Institute of Governmental Affas'rs. both at the University ofCalifornia at Davis. 'An hedonic model of house price was used for the comparison of price changes. The standardized houses were statistical composites. using the average of each of six house characteristics. Price changes wcre compared for several combinations of house and lot size. See Schwartz. Hansen, and Green (1981) fordetails. -A detailed discussion of the growth control program iscontained in Schwartz. Hansen. and Green i 1991). Schwartz, Hansen, and Green_ Growth Control developer, and the other to ensure that house and subdivision quality and other goals sought by the city were attained. Housing al- location and building permit data indicate that the growth control program reduced the number cf housing units built. In the first three years of the program only 37% of the single-family units proposed by developers received allocations (permissions to build). Also, during this peiiod the number of build- ing permits issuedwas 67% less than the num- ber issued during the three years before growth control. The comparison city of Santa Rosa is 15 miles north of Petaluma, which is at the outer limits for most commuters to San Francisco. Santa Rasa is larger than Petaluma (1960 population of 31.027) and has a considerable industrial and commercial base of employ- ment. Santa Rosa did not change its policy of encouraging growth during the period of this analysis n= did it experience any other changes that would rule it out as a suitable control for this comparison between cities. DATA AND METHODS Sales prices and physical characteristics cf new houses sold in Petaluma and Santa Rosa between 1970 and 1976 were obtained from the Society of Real Estate Appraisers. Our sample included approximately 75% of all sales Curing this period. The total number of cases was 597 for Petaluma and 784 for Santa Rosa. Sales prices were deflated to 1970 val- ues by means cf the Boeckh construction cost index.` which closely followed the consumer price index. We calculated the annual cumulative dis- tribution of sales prices and floor area for housessold in the two cities between 1970 and 1976. To determine what percentage of the houses could have been purchased b% moderare-income households (or lower). we calculated the maximurn price that such households could ha\ -e afforded to pay. Cali- fornia's Department of Housing and Corn rn u- nity Development defines the moderate - income range as between 801i� and 120% of rhe county's median income for it household of four people. To calculate the maximum III price that a household in this income range could have paid, we assume that the house- hold spends 30% of its gross income for hous- ing and that the buyer makes a 20% down paymeni and takes a 30 -year. constant - payment loan at the interest rate that pre- vailed in that year (in the range of 9.0% to 9.75% for FHA loans). Under these assump- tions the maximum price that a moderate - income household could have paid is approxi- matcly $25,000 in 1970dollars. Taking this as the cutoff (criterion) price. we compare the results 1r, Petaluma to those in Santa Rosa. Comparing Petaluina to another city is neces- sary to eliminate outside events (other than Petaluma's program) as possible explanations for the result.' If the pattern of changes in Petaluma is different fear that in Santa Rosa and is in the direction predicted by theory. we can conclude that the changes were due to growth control. The degree of confidence in such a conclusion will depend. of course. on the appropriateness of the comparison city. It is important to note that new houses built un- der the growth control program in Petaluma did not appear on the market until 1974, sowe consider the period 1970-1973 as pre -growth control and the period 197.1-1976 as post - growth control. RESULTS I n Petaluma the percentage of houses that sold for less than $25,000 ($1970) was be- tween 48.3%v and 56.7% before growth con- trol: aftergrowth control it dropped to 15.2% in 1974, 2.3% in 11, 75 and 3.3% in 1976(Ta- ble 1). In 1976. 68.2 % of Petaluma houses sold for more than $30.00() whereas before erowth control no more than 21.7% sold for more than $30,000 ($1970); in three of the four years before growth control fewer than 'The Boeckh index is published in U.S. Department of Commerce. Bureau of Industrial Economics, Con- struction Review. ` 'Sin -c a true experiment usine random a.sienment i�, impossible in this situation, %,.e use it quastexperimcnt. where the comparison city of Santa Rosa serve, a> a control. See Cook and Campbell t iu%U) for a detailed &-cu>sion of guusicxpcrime ntaI method, 112 Land Economics TABLE DISTRIBUTION OF SA,LiES PRICES OF NEW HOUSES :CUMULATIVE PERCENTAGE OF HOUSF, ,SOLD AT OR BELOW THE STATED PRICE' Cumulative Percentage Sale Price LessThan: Year: _ 1970 1971 1972 1973 1974 1975 1976 $20.000 9.1 12.5 25.000 52.I 54.2 30.000 97.8 92.5 35,000 99.9 100.0 520.000 26.2 21.5 25.000 43.1 38.7 Kom 66.2 68.8 35.000 78.5 95.7 PETALUMA than 1,400 square feet averaged 32.1 plc before $30.000 ($1970). The contrast to Santa Rosa growth control and 32.8% after growth con - 13.3 6.7 4.4 0.0 0.0 48.3 56.7 15.2 2.3 3.3 78.3 96.2 58.7 51.2 31.8 98.3 98.I 93.5 88.4 74.7 SANTA ROSA reasons why the disappearance of low-priced These data,prgytde strong evidence ,of the houses in Petaiuma can be attributed to its 7.1 5.1 10.4 5.8 10.7 32.9 36.5 39.9 37.4 37.5 78.6 69.4 66.4 67.8 59.8 94.3 88.3 85.3 87.9 74.I 'Prices are in constant 1970 dollars. °We d o not show the remaining price category which results in an ent ry of 100% in the last row because our interest is in the lower priced houses. The reader can easily calculate the remaining percentage of houses that sold for more than S35.000 (the difference between the last entry and 100%). 8% of Petaluma houses sold for more than than 1,400 square feet averaged 32.1 plc before $30.000 ($1970). The contrast to Santa Rosa growth control and 32.8% after growth con - is striking. There, between 32.9% and 43.1 % trol. The percentage of very small houses was of the houses sold for less than $25.000 over greater in 1975 and 1976 than in any previous the entire period (1970-1976). From i974 to year except 1970. 1976 (the post -control period). between It. Is clear that small. ,lower-priced new 37.4% and 39.9% of Santa Rosa houses sold .. houses nearly vanished from Petaluma after it for less than $25,000. Thus, the percentage of. ` Imsed growth control; but that did not hap- "affordable" housing ,,dropped from"-aDOUt`' pe^I,iti the comparison city of Santa Rosa dur- SO_o to Tess than 5% rt Petaiuma bclt`it re-, ing tilts. same. period:. -There are two major mained high -nearly 1t0%=tri Santa Rosa: reasons why the disappearance of low-priced These data,prgytde strong evidence ,of the houses in Petaiuma can be attributed to its shift to Petaturn 's housing away from the low growth control program. First. the criteria for end of ttie`tnarket after growth control: evaluating development proposals and The data for floor area documentthe dis- awarding housing permissions (atlocations) appearance of the small house in Petaluma were heavily weighted toward quality and after growth control. This is not surprising amenity items. More than 50% of the maxi - since hedonic price studies have repeatedly mum number cf points awarded in the rating shown :hat floor area is the most important of subdivision proposals were for such items determinant of variation in house price. If we as architectural design quality, site design consider 1,400 square feet to be a small quality, character of landscaping and screen - house, we see that the percentage cf small ing. provision of foot or bicycle paths and houses built in Petaluma dropped from about equestrian trails, and provision of usable open 39% in 1970 and 1971 to 11.0% in 1976. The space (City of Petaiuma 1972. General Plan. percentage of very small houses (below 1.200 Housing Element). Second. the city council square feet) dropped from about 20% in the made it clear to buildefs in the first year's 1970-1972 period to 1.1% in 1976. Again the allocation process that it wanted subdivisions results from Santa Rosa are in sharp contrast. cf high quality. Proposed subdivisions cf The percentage Of Santa Rosa houses smaller modest quality were rapidly e 1 i m i n a t ed from Schwartz, Hansen, and Green: Growth Control TABLE2 113 DISTRIBUTION OF FLOORAREA OF NEW HOUSES: CUMULATIVE PERCENTAGE OF HOUSES WHOSE FLOOR AREA IS AT OR BELOW THE STATED SIZE (SQUARE FEET) Floor Area Less Than: Year: 1970 1971 Cumulative Percenta_ee" 1972 1973 1974 1975 1976 PETALUMA 1200 sq. ft_ 20.4 22.5 20.0 10.6 6.5 2.3 1.1 1400 31L7 39.1 28.3 17.9 23.9 18.6 11.0 1600 58.4 56.7 43.3 13.3 28.3 25.6 23.1 I900 81.0 77.5 71.7 M.3 73.9 55.8 39.6 SANTA ROSA 1200 sq. ft. 4.6 6.5 10.0 11.0 14.0 17.2 16.1 1400 36.9 34.4 15.7 31.4 34.3 35.6 38.6 1600 52.3 57.0 51.4 54.0 62.9 57.5 55.4 1900 70.7 84.9 x2.9 86.9 86.7 81.0 913 consideration (Tarr 1978) We ,conclude ;, In Callforni therefore, that Petalutha's growth controi- .: pre.me cottrts'1 program' effectively eltmtnated'the prodnc for a gro%wth; tion �f dower pritgd housing to that retv�:, betzefits QnIy ti ..t-,•..,5 a... �':. n.. . -.».e, POLICY IMPLICATIONS Many local government decisionmakers perceive important benefits to their cont- munities from growth control, including en-- hanced environmental quality a rid amenities, maintenance cf "small town character." and better public services and fiscal status' (Ro- senbaum 1978;Johnston 1980). However, f& - cal decisionmakers may not be aware of or. concerned about. the costs of growth control because most of the costs fall upon individuals who live outside the growth control commu- nity or on renters in t he community. Since the losers are usually in lower income groups than the beneficiaries of growth control such pro- grams have potentially serious equity conse- quences (Schwartz 1982). To the extent that the losers lack political power to influence de- cisions within the growth control jurisdic- tions, the stage is set for confrontation be- tween state and focal pohcvmakers over the acceptability CC growth control programs. Re- cent actions by some local governments. as well as by state legislatures and state courts are evidence of growing concern for the eq- uitv consequences. w 3ersev. the state su ram: to provid4 keniiii;j. com in a;communit, negattve.regtona! mpacts on the suppir tower priced houstr g tf tt :cannot :da so th, .program: wtlY�be �•oosidered exclusionary Ii Cahforrtia: thy' Itigisiature mandatedthat to cai governments act:.af&rmatively to Inee their Nf shar't of regionalhoustng needs fo all income groups .(Chapter 1.141- Cali f orni Statutes• of tit The legtstature furthe placed het�iien of proof; -on local- govern meets t tacr.artt growth control ordinances tt, •'Less socially acceptable reasons for restricting growth may exist but are not usually expressed. For e\- ampie:. Ellicks64 (1979) -asserts that suburban growth controls a> a designed to enrich existing homeovrner> who, in effect, form a housing cartel to restrict the vtp- plv of nev►single=tataily houses. The targe literature on cxclusionarn land.use practices points to the protection of property values as :the primary motive for such prac- tices (Delafons 1969: Babcock and Bossetman 1973). "in California the relevant case is Associated Howe- hudders of Greater East gar r. Cin- of Livertrupre. 1 -- Cal 3d 582.557 P. 2d at 483. 135 Cal Reporter 41 (1976 t: in New JCTSCy the relevant case is Southern Burlington NAACP v. Township of.ttown Laurel, 161.1.J. Super. ct. Law Div. ?17. 391. A.2d 415 119781. 114 show,. in; any. court challenge, that the ordi- nance "is necessary for the protection of the public health, safety, or welfare of the popu- lation": (Chapter 1144, California Statutes cf 1%0).? Although California and New Jersey are at the forefront of efforts to eliminate exclusion- ary land development practices and provide affordable housing, other states seem likely to follrn" suit. Consequently. stringent growth control programs may not be able to withstand legal challenges unless the enacting communities also make special efforts to pro- vide affordable housing to lower-income households. Petaluma s^program withstood a legal;chalinge shgrtly after ti was adopted. based largely on the qty's goal of providing between 8% and=32% `of new,,hous ng in a pace range `affo'r'dable fo' low= or moderate - income householdsx Our=analvsis' shows, however; that Petaluma'failed>to achieve this goal:.Instead, its°`growth control program � nt+arlV"'eiiminatPti _nE+ir�'`inwPr nriri.�t-""c�noin_ family housing Petaluma's only incentive for encouraging affordable; housing was. by awarding, in;its evaluation of proposed devel- opments, up to 15 points out of. a total of 130 gotnts bu 6 pro. ion of affordable fitous- rriosv of the `remaining points ,were awarded for b.b sd§ dd subdivision'quality and arneni- ties. A stronger commitment must be made by the local government if affordable housing is to be built. Other jurisdictions—for exam- ple. Davis. California—have combined a stringent growth control program with stronger incentives for providing affordable housing (Schwartz and Johnston 1983). How- ever. the ability of even these stronger incen- tives to overcome the adverse impacts of growth control programs on affordable hous- ing is very much in doubt. Land Economics References Babcock. Richard R. and Bosselman. Fred P. I973. Exclusionan, Zoning: Land Use Regula- tion and Housing in the 1970s. New York: Praeger Publishers. Delafons, John. 1969. Land -Use Controls in the United States. Second Edition. Cambridge. Nlass.: The NIT Press. City of Petaluma. California. 1971. General Plan. Petaluma Housing Element (July 10, 1972). Cook. Thomas T., and Campbell. Donald T. 1979. Quasi -Experimentation: Design and Analysis Issuesfor Field Settings. Chicago. Ill.. - Rand McNally College Publishing Co. Ellickson, Robert. 1977. "Suburban Growth Controls: An Economic and Legal Analysis." The Yale Law Journal. 86 (Jan.: .385-511. Johnston. Robert A. 1980. "The Politics of Local Growth Control." Policy Studies Journal 9 (Special # 1): 427-39. Rosenbaum. Nelson. 1978. "Growth and Its Dis- contents: Origins of Locill Population Con- trols." In The Policy Cycle. eds. Judith May and Aaron Wildayskv. Beverly Hills. Calif-: Sage. Schwartz, Seymour L. and Johnston, Robert A. 1983. "Inclusionary Housing Programs." Journnl of the American Planning Association 49 (Jan.): 3-21. '1982. "Equity Implications of Local Growth Management." In Enrironmental Pol- icy Implementation. ed. Dean E. Mann. Lex- ington, Mass.: Lexington Books. Hansen, David E.: and Green. Richard. 1981. "Suburban Growth Controls and the Price of New Housing." Journal of Environ- mental Economics and Management 8(Dec. ): 303-20. Tarr. Fred. senior planner. City of Petaluma. 1978. Interview with Maureen Newby. Nov. 'Chapter 1143 in the Statutes of 19SO is at Culifi rn,a Government Code. sections 65580 ff.. Chaptcr 11.1.1 in the Statutes of 1980 is at California Evidence Cade, sec. tion 669.5 0 THE EFFECTS OF GROWTH MANAGEMENT ON THE HOUSING MARKET A Review of the Theoretical and Empirical Evidence JANE H. LILLYDAHL LARRY D. SINGELL University of Colorado at Boulder ABSTRt►CT. The no=growth or;slow growth poftctes that Itave.spad rapidly in rhe past 1 S years vary -an form surd Include ex»hcit population targets, rigid development controlad re to ptmvtde expanded public serytce' among other techniques The Purposeof thtsarttcle uto ret7ew tlusee foaxt The article fust presents a revww of the economtc :theor,� that tllumtncites the economic motrvwww", or such controls and idrnstfies some o the mcyor Impacts to tie gtnpncally assessed The next part of ar iele focuses on the hottsr�t8 price, productrorr, and equity effects of growth �ontm4 as documental alt the empiritaal kteiiature'The'article concludes witha s�ecdon devoted to INTRODUCTION Policymakers in American cities have significantly changed their attitudes regarding urban growth. For approximately 20 years after World War 11, strong pro -growth attitudes dominated. Growth was seen as necessary €orproviding expanded employ- ment opportunities and desirable because it would augment city budgets and provide greater social, cultural, and economic diversity. In the 1960s and 1970s the perception ofenvironmental deterioration as a result of urban growth emerged and cities began to search for ways to obtain higher quality residential environments. Initially, concerns were focused largely on physical features exemplified by air pollution and noise levels. However, in the mid-1970s. a number of suburban communities began to feel that JOURNAL OF URBAN AFFAIRS, Volume 9, Number I, pages 63-77. Copyright C 1987 by JAI Press, Inc. All rights of reproduction in any form reserved. ISSN: 0735-2166. continued growth threatened a wider range of amenities important to the quality of life. Limited or no -growth ordinances were seen as a means to maintain a pleasant, small- town atmosphere and lifestyle as well as to provide for open space, greenbelts, attrac- tive neighborhoods, n -h n i m al traffic congestion, and high quality public services at reasonable tax levels. Controls cf this type have now been ena•.ted in a large number of suburban communities in every region of the country. Indeed, the fervor to stop or slow growth became as strong in the 1970s and 1980s as that of the pro -growth movement in the 1960s. The no -growth or slow-growthpolicies that have spread rapidly in the past 15 years vary in form and include explicit population targets, rigid development controls, and refusal to provide expanded public services (e.g., schools, water supplies, or sewage treatment facilities), among other techniques. Such policies now find support in urban areas in every region of the United States, even in areas that previously encouraged rapid growth. Dowall (1982) identified 567 local governments across the country that had some form of growth control, and Segal and Srinivasan (1985) found that nearly two-thirds of a sample of 51 SMSAs from every region in the United States had growth restrictions. Although an average of 12% of the available suburban land was set off limits to growth in their sample, in some cities growth was barred from 30 to 40% of the surrounding Iand. Growth controls are, of course, not a new phenomenon. Local governments have long had the power to regulate new developments through zoning and other land use controls. What is new is the introduction of specific growth targets or limits and the pervasiveness of such policies. For example, a recent survey of 64 San Francisco Bay area jurisdictions showed that since 1970, approximately half of the jurisdictions had employed sane sort of moratorium on residential development for some significant period cftime (Gabriel, Katz, & Wolch, 1980). Although growth control policies are now widespread, Solomon (1976) was forced to conclude in his review of the literature that virtually nothing was known about the size or nature of the impacts of growth control. Since that review there have been a number of efforts to evaluate the impacts of such controls on the price, quantity, quality, and other aspects of housing from both theoretical and empirical perspectives. The purpose of .this article is to review these efforts. The article first presents a review of the economic theory that illuminates the economic motivations for such controls and identifies some of the major impacts to be empirically assessed. The next part of the article focuses on the housing price, production, and equity effects of growth control, as documented in the empirical literature. The article concludes with a section devoted to policy implications. THE THEORETICAL FOUNDATION FOR GROWTH CONTROL The reduction in social well-being that may result from unguided or uncontrolled urban growth, and therefore the justification for growth controls, is based on several eco- nomic arguments. For purposes of summary and analysis, these arguments may be grouped into three broad categories. First, secondary consequences or side effects of urban growth as exemplified by congestion occur and may be overlooked by decision - makers. Economists refer to these effects as externalities and demonstrate that if ( Effects of Growth Management on the Housing Market 173 has been a major source of increased personal wealth for middle-income Americans, resulting in a major wealth redistribution, between owners and renters (Sternlieb & Hughes, 1980). Renters may also incur significant losses in other more indirect ways. For example, rental units may be located in inferior school districts or in areas less accessibleto employment opportunities. Steger(1973) found, forexample, that limited housing choice of central city residents results in a 5% loss of income because of additional commuting costs. Growth control programs may also create importantjurisdictional inequities When growth is discouraged in one community, it may be shifted to srrrounding jurisdictions_ Even if housing prices are not increased. if public services are producsd under condi- tions of increasing costs, this will place a greater burden cal surrounding communities. In such cases, it may be in the best interest of surrounding communitiesto adopt growth " control measures also. Schwartz (19 82), €orexample, observed that "there is evidence to suggest a chain reaction of growth control adoptions in communities between San Francisco and Sacramento following several years after the pioneering programs of Petaluma and Davis" (p. 232). Growth control programs have pronounced effects on the availability of low- and moderait-income housing. In unique urban environments where permission to build is based cn amenity and design characteristics, affordable housing for low- and moderate -income families may vanish. In metropolitan regions where interconnected housing markets lead to similar types of controls in many adjacent communities, low - and moderate -income families maybe forcedby such controlsto live in neighborhoods that further disadvantage them with respect to job opportunities and education. A postponement or elimination ofthe opportunity for home ownership may also result in a significant redistribution of income. CONCLUSIONS AND POLICY IMPLICATIONS The evaluations of,the impact of growth controls suggest that while these efforts ma} have.help maintain the desired community character, 'housing pric es have increased and in manyses'the avatiabiility of low or moderate prtced housing has decline( substantially .or:evetF Ztsappeai<ed.- This combined. with national inflatioin and higt mortgage mteresrrates :has created a real crisis to kousing; markets. Currently, a verb large.percet►tage of families cannot:afford;to'purchasehousrng. r�►n'expanded numbei of -households hasbeen forced to rent and/or reduce housing..size or quality. Higher prig siin the Nl rental market also make. the, accumulation of:capital forfuture purchas( mare difficult The long=term viability of growth control programs cansequentl3 requires that some:means be developed tobffset the price, equity, and extrajurisdic tionat impacts of these programs, :,A number of policy recommendations have been advanced. For example, Ellicksor (1977) proposed a. judicial remedy recommending that. communities be, allowed tc nursue Qrowth control measures but that. state . courts use ttie taking clause in their constitutionsto entitle landowners and housing consumers to sue for damages. Certain landowners could recover damages for land value losses and home buyers or renters could bring class action suits to recover price increases brought about by growth 741 JOURNAL OF URBAN AFFAIRS 1 Vol. 9iNo. 1/1987 control. if the growkh control community could demonstrate that its program was both efficient and equitable to consumers of housing; it would not be subject to damages. As an alternative to Ellickson's judicial remedy, a number of local governments have sought to overcome the undesirable side effects by establishing inclusionary housing programs which either provide incentives for or mandate the construction of Iow- and moderate=cost housing In a recent review of inclusionary housing programs, Schwartz and Johnston (1983} :provided' evidence for the inadequacy of nationwide efforts to provide housing forjow or moderate income famines Fofsuch programs to be suc- cessful;Schwartz and Johnston Argue:that they must: (1) mandate moderate -cost units frpm medium to large scale developers i-ian requue fees from most other developers; (2) screen buyer and;renter applications to maxmuze social;benefits; (3) control resale price ip order to retain units' at`=moiierate prices; and t4) provide? substantial economic inceaitivesiio developers (p=19). As an alternative, or perhaps in conjunction with such programs, cities might sell local revenue bonds for the purpose of providing low-interest loans to moderate- inc.on ^ t t y : Below market tnterest rates are, hawever� in general inadequate in b ringing' housing costs down. to :the :point where they: are: affordable to low and moderate income families.' Unfortunately, many of the forces that underlie the inflation in housing costs are beyond the control cif local governments. Hence, if the provision of low- and moderate -cost housing is considered a desirable social goal, a national hous- ing policy committed to assisting local governments in this effort may be necessary. While:a combination of federal`effotts.and inclusionary hpusing programs by;iocal 3 governments might be workable, the larger problem of efficicntIy providing hottsirig in 4 o a `metropolitan 'region, still remauis, It6 y7 that aggregate housing, prove- s'ion wile match needs in' an optimal manner, when individaai political "subdivisions institute" gmw�lt eont�ols and when housing construction `is the primary ,source of contra!: ,Growth control and inclusionary prpgxams repress or distort ,normal free market pi�odetction a.reduce�the ttorrn�l l�eiefits of trickle down to Lower income buyers anii renters in:addition, these programs:are enacted by subinetropolitan com- tan community. While there is a real need for efforts to maintain the quality of local environments, metropolitan areawide governments are still too weak to ensure that such efforts will take cognizance of the benefits and costs to the larger society. All too often, a disproportionate portion cf the costs of such programs fall on the lowest income groups who can least afford tbEm. NOTES 1. The graphical analysis draws on Ellickson's 1977 article in the Journal of Law and Economics. 2. The actual increase could be more if the price were set by market forces that did not consider the congestion costs. 3. This raises one of the difficult theoretical and empirical issues in the growth control literature. If antigrowth policies raise housing prices within municipal boundaries, these higher prices will make housing in neighboring jurisdictions more attractive to consumers. The result would be an increase in the amount of housing built outside the boundaries, and potentially. an increase in housing price in these jurisdictions as well. From a resaarch point of view, this Effects of Growth Managemen! on the Housing Market 175 frequently causes difficulties since cities with unique environ men.al settings that impose controls can be compared most easily with cities in the same environment without controls. However, if the cities are close enough to have similar environments. then their proximity also integrates their housing markets so that little or no price difference can be observed over time. When compari- sons are made between cities that are distant enough to have separate housing markets. environ- mental amenities are less likely to be comparable. From a practical or policy point of view, the inp3at of growth control on surrounding communities means that one jurisdiction can impose costs on neighboringijurisdictions who are not free to express their preferences in the ballot box. 4, Some Boulder, Colorado, residents have claimed that due to growth controls, high-density condominiums and apartments have replaced single-family units with the result that this city is less family oriented and more oriented toward singles lifestyles. 5. Many of the empirical studies of growth control have used California data. Because of the atypical housing markets in that slate, it is possible that the results may not be generalizable. However, the articles which do focus on growth management in other parts of the U.S. (e.g., Butler& Myers's [ 19841 article on Austin, Texas, and Knaap's [ 19851 study of Portland, Oregon) present findings which are consistent with those cf the California studies 6. One externalityassociated with some growth control programs is delays resulting in higher construction costs. Although this issue is beyond the scope of this paper, it is worth noting that these costs may be substantial. The Rice Center for Community Design and Research (1979) estimated that government regulations in ftHouston housing market resulted in costs of $1,400 to $2,100 fora 13 -month delay. The ConstructionIndustry Research Board (1975) cameup with substantially higher estimates of a delay prior to construction of $1,027 per month. 7. Rosen and Ka tz (1981) criticized th is study because land price increases were completely a ttnbuted to growth control ignoring inflation and other market factors and because there was no control city to allow for a determination of increases in profits and costs resulting from growth management. 8. Some of the econometric problems include: autocorrelation, simultaneous equation bias, specificationerrar, and partial use of forecasted rather than actual data - 9. It is common for researchers to have concentrated on the price effects and ignored the production effects. More research needs to be done investigating the growth control effect on housing quantity and not just on housing price. REFERENCES Beadey, T. (1984, Autumn). Applying moral principles to growth management, Journal of the American Plonning Association, 459-469. Blewett, R (1983, January). Fiscal externalitiesand residential growth controls: A theory of chtbs perspective. Public Fumnce Quarteriy, I I(1), 3-20. Butler, K., & Nkms,D. (1984, Autumn). Boomdme in Austin, Texas: Negotiated growth man- agement. "Journal an- agement."Journal of the American Planning Association. 447-458. Construction Industry Research Board 0 975, April). Costs of delay prior to construction. Los Angeles: Author. Cooley, T., & LaCivita, CJ. (1982). A theory of growth controls. Journal of Urban Economics. 12,129-145. Correll, M., Lillydahl, J., & Singell, L. (1978, Mky) . The effects of greenbelts on residential property values: Some findings on the political economy of open space. Land Economics. 54(2),207-217. Dowall, D. (1982). An examination of population -growth -managing communities. In D. Mann (Ed.), Environmentalpolicy implementation. Lexington, MA: Lcxington Books, Dowall, D. (1984). The suburban squeeze. Berkeley: University of California Press. 761 JOURNAL OF URBAN AFFAIRS I Vot. 91No. 111987 Downing, P. (1977, June). Suburban nongrowth policies. Journal of Economics Issues, 11(2), 387-399, Elliott, M. (1981). The impact of growth control regulations on housing prices in California. American Real Estate and Urban Economic Association Journal 9.115-133. Ellickson, R. (1977, January). Suburban growth controls: An economic and legal analysis. The Yate LAw Journal, 86,395-5 1I. Gabriel, S., Katz L, & Wol ch, J. (1980, Spring). Local land -use regulation and Proposition 13. Taxing andSpending, 73-81. Gabtiel, S., & Wolch, J. (1980, November) -Local land vse regulation and urban housing values. (Center for Real Estate and Urban Economics Working Paper No. 80-18.) Berkeley: University of California Press. Gleeson, M. (1779). Effects of an urban growth management system on land values. Land Economics, 55, 350-65. Huszar, P. (1977, July). Equity and urban growth: Real property value appreciation in San Jose, California. American Journal of Economics and Sociology, 36.25 1-261. Janczyk, L & Constance, W. (1980). Impacts of building moratoria on housing markets within a region. Growth and Change, l I, I I -19. Johnston, R.A. (1980, December). The politics of local growth control. Policy Studies Journal, 9. (Symposium Issue on Environmental Policy.) Johnston, RA, Schwartz, S1, & Hartz, W. (1980). The effectiveness of the moderate cost housing program of Davis, California. Unpublished manuscript, Division ofEnvironmectal Studies, University of California Johnston, R.A., Schwartz, S.I., & Tracy, S. (1984, Autumn). Growthphasing and resistance to infill development in Sacramentocounty. Journal of theAmerican Planning Association, 434-446. Katz, L., & Rosen, K (19 8 0, September). The effects of land use control o n housing prices. (Center for Real Estate and Urban Economics Working Paper No. 80-I3.) Berkeley: University of California. Knapp, G. (1985, February). The price effects of urban growth boundaries in metropolitan Portland, Oregon. Land &ctnc nics,61(t ), 26-35. Landis, J. (1986. Winter). Land regulation and the price of new housing: Lessons from three California cities. Journal of ft -American Planning Association, 9-21. Rice Center for Community Design and Research. (1979). The cost of delay due to government regulation in the Houston housing market. (The Urban Land Institute Research Report 1128.) Houston, TX: Author. Richardson, H (1973). The economics cf urban size. Lexington, MA: Lexington Books. Rosen, K, & Katz, LF. (1981).Growthmanagement and land use controls:The San Francisco Bay Area experience. American Peal Estate and Urban Economics Association Journal, 9, 321-344. Schwartz, S. (1982). Equity implications of local growth management. In Dean Mann (Ed.), Environmental policy implementation (pp. 223-248). Lexington, M A Lexington Books. Schwartz, S., Hansen, D., & Green, R. (1981). Suburban growth controls and the price of new housing. Journal of Environmental Economics and Management, 8,303-320. Schwartz, S., Hansen, D., & Green, R (1984. February). The efffect of growth control on the production of moderate -priced housing. Land Economics, 60(1), 110-1 14. Schwartz, S., & Johnston, R.A. (1983). Inclusionary housing programs. Joumalof the American Planning Association, 49, 3-21. Segal, D., & Srinivasan,P. (1985). Suburban growth and housing inflation. Urban Geography. 60), 14-26. Singell, L.D. (1974). Optimum city size: Some thoughts on theory and policy. Land Economics, 50(3), 207-2;.2. } Ettects of Growth Management on the Housing Market 177 Solomon, A. (1976, February). The effect of land use and environmental controls on housing: A review. Cambridge, MA: Joint Center €orUrban Studies of MIT and Harvard University. Steger, W. (1973). Economic and social costs of residential segregation. In bWm Clawson (Ed.), 1 &xkmbV urban laP4 policy (pp. 83-113). Baltimore, M D The Johns Hopkins Press. Sternlieb, G., & Hughes, J. 0 980). Housing and sheltercosts: The schizoidproblem of the central city. In America's housing. prospects and problems. New Brunswick, NJ: Rutgers University, Cumty- for Urban Policy Research -Tiebout, C. (196). A pure theory of local public expenditure. Journal of Political Economy. 64, 416-424. Tolly, G. (1974). The welfare economicsofcity bigness. Joumal of Urban Ebx=zics. 324. Urban Land Institute, & Gruen, Gruen, and Associates. (1977). Effects of regulation on housing costs Twocase studies. Washington, DC: Urban Land Institute. LAND PRICE INFLATION AND AFFORDABLE HOUSING 'a$ J. Thomas Black and James E. Hoben U.S. Departmentof Housingand Urban Development Prices for standard lots and acreage were collected for 1975 and 1980 in 30 metropolitan areas and then analyzed, using multiple regression, to identify factr•-s which would explain variations among metropolitan areas. Extreme price variations were observed. For example, from 1975 to 1980, the price of a standard residential lot increased as little as 31% in one area, white the price r o x 176%in another. Over 80% of the variation in lot price increases was explainable by a model combining land supply and demand factors. Ik,thelrardero!`linpsnce ori, the,%ctors`were::(t) an Index of, regulatory resuiciiozi; (2j popu>:tion increases; (3j percapita Income in creases. and {4j job increases. The analysis suggests that pubI c regulatory, Inframc- mre and"fax pollclet 4m'- stgnificantbj affect land, supply and demand and, in turn, prices. Communities that clwose to manage growth must monitor land supply and demand' and,`adjvst their potkks,to ensurewnipetitive noninftationaiy land markets. r,' Otherwise' -major 0.cream, in land prltxs for Housing and- buss nesses may result. _ INTRODUCTION Land comprises a significant portion of the cost of housingand of many businesses. Over the decades, that proportion has risen and fallen depending upon the cost of land, the amounts utilized, and changes in other cost components. In 1980 it ap- peared that land costs had risen sharply in many growth markets. While the U.S. Cen- sus showed that land as a component of housing cost was around 21%nationwide, in some markets, such as the West Coast, FronVrange and Sunbelt, there were reports that land comprised 40% of single-family housing costs. What makes these increases so important is that they were in the areas of the nation where housing demand was greatest The reason for concern over rising urban land prices is simple: Housing quality and home ownership are threatened and business costs are increased. Besidescontributing to problems of housing affordability, a drop in housing production results in reduced business activity, and increases in housing costs can lead to increased vuage demands, thus significantly affecting the economy. To find out if land prices really were increasing, and , if so, why, the authors cal lected price data from 30 metropolitan areas and, through correlation and multiple - regression analysis, examined the reasons for variations in levels and trends in land prioes,' The principal research question was to what extent did demand versus supply factors explain changes in residential land prices. What follows is a brief three-part summary of the exploratory research. First, a description of the variation in land prices between metropolitan areas; second, a report on efforts to identify why prices vary between markets; and finally, some hypotheses are offered on the operation of land markets. The research methodology is dmribed in tie text and additional details are provided in the appendix. 27 Urban Geography, 1985, 6, 1, pp, 27-47. copyright 0 1985 by V.H. Winston & Sons, Inc. All rights reserved 28 BLACK AND HOBEN LAND PRICES A major difficulty in understanding land markets has been the absence of compara- ble land price data. Data collection has been hampered by the fact that no two parcels are exactly alike; each pie:e is unique in terms of its size, location, topography, sub soil conditions, public regulations, supporting services, ownership, and future utility. Pdcesvary with these attributes and with the conditionsof sale. Analysts have struggled to develop price data by manipulating sales data or prop- erty assessment records, performing residual cost calculations or projecOng future returns for raw lands. None of these efforts has produced a suitable record of com- parable prices for different areas of the country and for particular years. The US. Census Survey of Construction includesone item on residential land costs; however, it is not published because the response rate is lower than the Censusconsiders accepta- ble. The Federal Housing Administration (FHA) publishes land cost data for FHA - insured new and existing single-family properties but the sample is not representative of all single-family housing Finally, the U.S. Department of Agriculture publishes data on the value of farm acreage but this series excludes land at the urban fringe. Faced with poor data, the authors developed a survey methodology.2 It involved defining two types of standard land parcels and asking homebuilders and residential appraisers in 30 metropolitan statistical areas (SMAS) for estimated prices. Since the greatest interest was in land as a component of housing costs, the work focused on two types of parcels considered key to housing production: An improved single-family lot—defined as having 10,000 square feet, zoned for single-family, with utilities to the lot, located in an area attractive to buyers of mid -price, single-family homes and within 20 minutes of a major employment area. Unimproved acreage suitable for single-family use—defined as parcels of 20 to 100 acres, with utilities available to the site at negligible cost, at the urban fringe, within 20 minutes of a major employment center, without any adverse environmental conditions, and not in a prestigious area. The definition of a standard parcel est*lished a single image of a piece of land for which an expert could then estimate a probable price. By using the sane definitions, prices also could be compared from one market area to another. The SMAs3 were selected to represent all sections of the country and varying sizes and growth rates (see Fig, 1). The sample was not random. The nation's largest metropolitan areas were omitted because of the expectrd difficulties respondents might have with the survey. No special treatment was given to multi -centered areas. The authors feel, however, that the areas selected reasonably represent the universe of metropolitan market areas. For each metropolitan area, approximately 25 to 40 real estate experts were identi- fied from referrals and professional membership lists. The experts were mailed a sur- vey describing the standard parcels and asked to estimate prices for both 1975 and 1980. The average response rate was 10, or about 35%, with a high of 16 and low of 4. In general, individual price tstimates for an area were consistent The prices cited in this paper are the median values provided by the experts. Following are the findings on land price variations for residential lots and raw acre- age among 30 metropolitan markets. 40 BLACK AND HOBEN 1975-1980 Percent Increase in Raw Land Prices Explanation of Inter -Metropolitan Land Price Variations Independentvariables (in order of importance) Improvements in correlation with additional variables R2 Adj. RT 1. Percentage increase in jobs,1975-1980 .331 .302 2. Regulatory restriction rating, 1980 .496 .44S 3. Percentage increase in income, 1975-1979 .570 ,508 4. Physical restriction rating ,599 .519 5. Percentage increase population, 1975-1980 .632 .536 increase In raw land prices= 121.1 + 1.0 (jab I nc.) -24.3 (reg. rating) +4.0 (pay Inc.) - 26.8 (phys. raring) + 3.0 [popu. inc.) Standard deviation 48.9; cares 28, adtbed San Jose and Cincinnati, the first because of extreme values and the second for lack of 1975 price data. Supply and demand factors represented by regulatory restrictions and increases in population, jobs, and income were the major factors influencing the rate of increase and land prices. The explanatory power of the raw land model was lower than that for improved lots since it is believed that the raw land market is less directly influenced by current changes in supply -demand conditions and more influenced by ownership char- acteristics and longer term market expectations. It is important to stress, at the conclusion of this analysis, that the data collected and its manipulation were exploratory at best. Additional research is needed to con- firm or refine the observations. Areas for possible improvements include randomly selecting the sample of metropolitan areas, refining the standard parcel definitions, and testing the price survey methodology. More time points are needed. Finally, much work is needed on defining and testing additional explanatory factors. The regulatory restriction rating needs refinement, and other supply measures need to be developed. SOME SPECULATIONS In spite of the methodological limitations, the preceding results support two propos sitions regarding urban land policy. 7. PrevaZZYng Land frices May Be signlflcantly Affected by the Aggregate Effects of Govemment Policies It is well known mat the value of an individual parcel can be increased or decreased by a change in zoning which affects the economic rate of return from the parcel. Simi- larly the provision of a sewer, road, or transit line can impact a parcel's value. What this study suggests, and is not commonly recognized, is that at a much larger scale the combination of local policies (regulations, infrastructure investments, taxes) probably changes the overall balance of land supplies to demand and thereby raises er lowers the average price for land. LANDPRICE INFLATION 41 The recognition in the 1970s of strong connections between urban development policies and objectives such as encouraging revitalization, infill and compact develop- ment, reducing pollution, protecting agriculture, and balancing local budgets has re- sulted in subtle but far-reaching changes in land markets. Generally, local policieshave evolved from accommodating growth to the control of growth. New policies include trying to direct the amount, type, location, and timing of private development_ In locations where special efforts have been made to manage development, itis likely that the public sector has greatly reduced the supply of land available for development Especially serious land/housing price problems occur when local policies seek eco- nomic growth but limit population (residential) growth. Ifa community adds jobs, the demand for housing must increase. Ifthe amount of land for residences is limited and densities are kept low, there is bound to be increased land competifion and price inflation. Public officials, interest groups, and citizens need to balance their desire for com- pact orderly growth and the protection of the environmentand agricultural landswith policies which -ill also assure competitive land markets with well -located affordable sites for homes and businesses. For example, the San Francisco Bay area is now com- ing to this realization. The Bay Area Council, an association of 300 leading businesses, has issueo a strong call for balanced land policies. San Francisco and Sunnyvale, Cali- fornia have responded by adopting policies linking economic expansion and housing Much morethought needsto be given to how tvachieve and maintain these balances. 2 ere,.,. 1T Many., Lard 49 its public Policies and Economic Cycles Are Likely to i� ffecf Them O/ffe�rit/y � '- ` An interesting finding of the study was the difference in the absolute prices and rates of increase in prices for the two land types. The median price of improved lots increased an average of 66.41/6 while the price for raw acreage increased 92.7%. Why would one increase be so much more than the other? One explanation of the higher increase in raw land prices is that the price of raw land is only one of the factors affecting the price of an improved lot. There are other costs of developing land. The costs of converting raw land to the developable stage may not have risen as much as the cost of the raw land itself. if they did, the price of developable land would have risen as much or more than the price for raw land. Another explanation is that the owner -investors of improved land and the owner - investors of raw land are influenced by different forces. Each owner -investor group has different objectives and different financial capabilities and therefore public policies and economic cydes affect each differently. For example, developers specialize in purchasing raw land and, through a process of rezoning, subdivision, and the construc- tion of streets, water systems, etc., produce improved lots. Land improvement involves considerable equity investment plus funds at relatively high interest rates. The devel- oper is impatient, if not obligated, to seii his improved lots within a short term. His financial success i s dependent upon maintaining a Cash flow. I f there i s an excess of improved lots relative to demand, the deveioper is forced to cut his profit An excess of tots occurred in the late 1970s and was exacerbated by the onset of the housing A 42 BLACK AND HOBEN recession of the early 1980s. The 66% increase in lot prices probably only reflects the 52% inflation for the 5 years, the costs of increasing focal development fees, and a modest profit, if any. Data for price changes from 1980 to 1983 might reveal no changes or even declines where developers went bankrupt. In contrast, the owners of most raw devetopable land are more likely to be patient, long-term investors who are financially capable of withstanding short-term drops in demand and who look for long-term price appreciation. A recent study of fringe area land owners found tfiat as much as 50'% of the improved suburban and rural fringe land was held by the current owner for 8 years or more (Brown, Phillips and Roberts, 1981), and it was also estimated that only one quarter to half of the owners were interested in selling their land at any one time (Brown et al., 1981; Real Estate Re- search, Inc., 1982). The carrying costs for raw land are modest as there are usually few improvements. Taxes also tend to be modest Such owners don't have to sell to survive. 3 Thus while the price of improved lots is especially sensitive to national economic conditions, the price of raw land may be less so. As others have suggested, it appears that the price of land, especially that of raw land, may follow a ratchet pattern of moving upward but seldom downward since few parcels 2re subject to forced sales. Beyond these two explanations, there is the possibility that the imposition of restrictions on the raw land that may be developed can produce conditions of narrow or quasi -monopoly control of available raw land Only a relatively small proportion of raw land is available for sale at any given time, as noted above. if this pool isfurther diminished through growth controls, the number of*remaining possible sellers may be reduced enough to grant them significant marker power. We do not know whether such additional concentration of market power was 3 occurring between 1975 and 1980. Analysis of the types of growth controls imposed in specific areas and the course of raw land prices in dree areas might provide some evidence. These remain importantquestions for further research. CONCLUSION In many market areas, land for building i s becoming an ever larger cost in housing and other development when there is no general land shortage. if any one point stands out from, the findings of this research,' it isthat policy makersshould pay moreatten- tior.to }and cost''as a component of housing, shopping center, industrial park, and 'office costs;; We monitor` em ployment levels, money supplies, and interest rates nationally. The U.S. Department of Agriculture monitors farm and ranch land prices and quantities. But no one systematically monitors urban land prices, much less the factors which might explain urban land market behavior. Policy makers should know the prices for representative parcels and know more about how public decisions on zoning, the construction cE a sewer tine or mad, an employment expansion, or adoption cf growth limits can affect the development market Current trends suggest that the supply aE improved land for building has diminished considerably because of curtailed infrastructure investments, more regu- lations, and widespread enactment of agricultural land preservation programs. The increased costs in some localities are maskedby permitted increases in residential den- U\NOpn>Ce|mFLAT0m 43 -~`0-_g-_='housing � Without beftit knowled of ge Jim be be h i 'ex 'ch e�ffiainte eal es �.��,-nanc�-bt-h'ealthy.,comi�dtitive� an mar is red relit6dJ9 NOTES `Txb article is based upon the findings wfuresearch project jvindvsupported ur the Urban Land/nmioute (uU) andthoV'S. DoportmontofHousing and Urban Development <*VD>Office of Policy Development and noaoomh. Thomas Black iothe VLIStaff Vice PmaidontforResearch and )ume* I- Hoovn is the MVo Chief ofCommunity Planning and Design Research. Frank Donau, Jay Miller, and Thomas Richardson assisted with the analyses while serving as interns at uL| and HUD. Tbcviowaand conclusions expressed inthe article are those ofthe authors and not necessarily those oftheir au*^cim 7 The survey method was first suggested b y Professor J amcs; Brown of the Department of City and Regional Planning otHarvard University. 'Ono cfthe sample areas, 8ou|do,. Colorado, was notmmetropolitan statistical area atthe time ofthe survey. "Twelve areas womaumoyodbm 1900amd lV in 1981. 44 BLACK AND HOSEN APPENDIX Land Price Estimates The !and price data for the 30 metropolitan areas were obtained by requesting rea! estate experts in each area to estimate the price for prescribed standard parcels in ,heir area. Standard parcels were defined tD minimize qualitative and quantitative differences among properties and to permit comparisons across regions. Two standard parcels were specified: (1) an improved single family tot, and- (2) raw acreage suitable for single family development; The survey instrument with a description of each standard parcel is reproduced below. URBAN LAND INSTITUTE RESIDENTIAL LAND PRICE SURVEY Insirurtions We are requesting your best estimates of typical prices for improved single family lots and raw single family acreage as described below. Your answers need not reveal anything about your own business dealings. Please record two prices for each of the two types of property. The first should be a recent estimate based on 197980 transactions. The second should reflect the price of a comparable piece of property in 1975. The latter can be based on either an actual transaction or a published average. If information is not readily available for 1975, feel free to substitute another year between 1970 and 1975, but be sure to note that year in the space provided. It is very imporunt that your estimates refer to residential land which reasonably meets the stated criteria for size, location, and other characteristics Estimates based on land which significantly deviates from the standardized characteristics described belowwIll reduce the accuracy of the survey results. Do not feel obligated to provide all four estimates if you do notfeel qualified to d oso. The attached form is provided for your answers. PROPERTY TYPE ONE: Improved Single Family Residential Lot Character/st/u Size: approximately 10,000 square feet (:1000) Zoning: single family detached Location: suburban fringe within 15-20 minutes driving time of a major employment center (not necessarily the central business district) ' within 2 miles of an existing grade school or bus zone Development Rights : No restrictions other than toning and building require- ments Utilities to Lot: sewer, water, electricity, telephone Neighborhood : nota prestige ::rea (homeprices within $60,000 to $90,000 range) . area at least 50% developed • no unusual conditions which might impact the land price such as: a significant pollution (air, water, noise) o environmental hazards (floods, etc.) LAND PRICE INFLATION 45 o close proximity to amenities such as major parks or shopping areas Financing: The price should reflect normal financing terms for your area. PROPERTY TYPE TWO: Unimproved Acreage Suitable for Single Family Resi- dential Use Charocterlstics Size: 20-100 acres Zoning: Residential, suitable forsingtesamily detached development Location: - developing fringe area - within 15-20 minutes cE a major employment center (not necessarily the central business district) Development Rights: No restrictions other than zoning, subdivision and build- ing requirements Utilities to Property: Connections to network available at negligible cost Other Characteristics: - nota prestigearea (home priceswithin $60,000 to $90,000 range) - no unusual physics! attributes such as slope or soil conditions which would increase the cost of development - no unusual environmental conditions, e.g. significant pollution or hazards - no unusual amenities such as extremely close proximity to a major shopping or a recreational area Financing: The price should reflect normal financing terms for your area. Approximately 25 to 40 real estate experts in each area were identified from refer- rals and professional membership lists. They included land developers, builders, pri- vate appraisers, public assessors, and lenders. The experts were mailed the price survey and asked to estimate prices for 1975 and 1980. The average response rate from each region was 10 or about 35% with a high of 16 and low of A In general, individual price estimates for an area were consistent However, to minimize the distortion of unusually high or low price estimates, the median price estimate was selected as the representative price for a standard parcel. No actual sales were analyzed as part of the price survey cr for validation purposes. Analysis oMice Variations The 1975-1980 land prices were analyzed by regressing them against a number of factors which might approximate land supply and demand forces in each SMA. A stepwise multiple -regression program was used from the Statistical Package for Social Sciences (SPSS). The selection of factors was limited to readily. available SMA data, except in two cases where f?ctors were developed by the researchers. The independent factors analyzed were: A. Supply Factors o Physical restriction rating, 1980. (U Ll research staff estimate—see article for ex- planation.) C Is 46 BLACK AND HOBEN 0 1980 regulatory restriction rating, 1980. (ULI survey—see article for explana- tion.) o Single family building permits per 1,000 population, 19751980. (U.S. Bureau of Census) o Ratio of employment income increase to issued singie family building permits, 1975-1980. (U.S. Bureau of Labor and Census) o Ratio of new to existing home loans, 1975-1980. (U.S. Federal Home Loan Bank Board) B. Demand Factors o Population, 1980. (U.S. Bureau of Census) o Percentage increase in population, 1975-1980. (U.S. Bureau of Census) o Employment, percentage increase in 1975-1980. (U.S. Bureau of Labor) o Income per -capita, 1979. (U.S. Bureaucf Economic Analysis) o Percentage increase in per -capital income, 19751979. (U.S. Bureau of Eco- nom*.c Analysis) C Baseline Factors 0 1975 improved lot price. (ULI Survey) 0 1975 raw land price. (ULI Survey) o Median new single family home price, 1980. (US .Federal Home Loan Bank Board) LITERATURE CITED Barnard, C and Botcher, VIS 1981, Landowner characteristics as determinants of developerlocational decislons� Paper presented at AAEA meeting, Clemson, SD Beaton, Q R., 1982, A n Examination of Relationships Between Land Use Planning and Housing Costs in Oregon, 1970-1980: Focus on the Urban Growth Boundary. Salem, CR: Willamette University. Black, T. and Hoben, j., editors, 1980, Urban Land Markets: Price Indices Supply Measures, and Public Policy Effects. Washington, DC: The Urban Land Institute. Brown, J., Phillips, R. and Roberts, N, 1981, Land marketsat the urbanfringe—new insights for policymakers. Journal of the American Planning Association, Vol. 47, No. 2 131.144. Coughlin, R., Klein, T. and William, M., 1983, Changes in patterns of land ownership in use during the early stages of urban development: a case study analysis. Re- search Report series. Philadelphia, PA: University of Pennsylvania. Hoben, J., 1981, Monitoring land prices in the United States. WoridCongmss on Land Policy 1980. Cambridge, MA: Lexington Press. Landis, J.D., 1984, Land regulation, market structure and housing price inflation: lessons from three California housing markets. Doctoral dissertation, University of California, Berkeley Planning School. LAND PRICE INFLATION 47 Peiser, R, 1981, Land development regulations: a case study of Dallas and Houston, TX. journal of the American Real Estate and Urban Economics Association, Vol. 9, No. 4,397-417. Real Estate Research Corporation, 1982, Infill Development Strategies. Chicago, 1L: American Planning Association. " Segal, D. and Srinivasan, P., 1980, The impact of suburban growth restrictions on US housing price inflation 1975-1978. Unpublished paper for the National Science Foundation. Land -Use Controls and Housing Costs: An Examination of San Francisco Bay Area Communities* David E Dowall ** and John D. Landis*** 3 This paper reports on our efforts to gauge the effects of Iand use controls OR housing markets, We discuss how land use controls affect land and housing assts and explain why communities use such con- tmis to restrict development. We present the results of an econometric model created to assess the inflationary effects of land use controls on housing costs. The model is based on data assembled in the San Fran- cisco Bay Area. The model results indicate that density controls and land availability do systematically affect the price of new housing units. INITRODUMON Confronted with continuing increases in the cost of new housing, city planners and urban economistsnow find themselves re-examining the validity of local land use and development controls. In California, the controversy over land LM con- trols has raged for some five years, sparked to no small extent by housing prices Financial support for this paper came from the California Department of Real Estate, the California Air Resources Board and the Construction Industry Advancement Fund. Associate Professor of City and Regional Planning, Center for Real Estate and Urban Economics and Institute of Urban and Regional Devebprnent, University ofiCf�+a, Berkeley, California, 94720. *** Department of City and Regional Planning, University of California, Berkeley, Cali- fornia, 94720. 67 rvt111ial l Vol. I which remain the nation's highest. In his widely -quoted 1979 work, The Environ- mental Protection Hustle, Bernard Frieden [1 ] concludes that in the case of the San Francisco Bay Area, unnecessary growth controls are adding thousands of doLars to the cost of constructing new housing. Perhaps a more important effect, Frieden notes, is that by constricting supply in the face of burgeoning demand, local land use controls fuel the flames of housing inflation. Moreover, those responsible for such restrictive controls often act entirely out of self-interest, for as the price of new housing rises, so too do the prices of existing housing—in the process providing existing landowners with windfall profits. Similar conclusions -' have been voiced by Gruen and Gruen Associates 121 :n a study of the appli- cation of growth controls in Petaluma, California, and more recently, by Dowall (31 in a study of land use controls as administered in six representative Bay Area suburban communities. Academics and consultants, however, are not the only observers to express alarm at the current situation; the State of California has also voiced concern. A recent survey of local land use planning in California, undertaken by the Gover- nor's Office of Planning and Research [41, revealed that over 5017o of the 93 cities in the San Francisco Bay Area were actively limiting population growth. To reduce excessive project approval times, the California Legislature in 1977 enacted AB 884, a bill requiring local governments to approve (or reject) major residential projects within one year of the initial submission date. But, because of a variety of loopholes, AB 884 has not been effectively enforced.' Most re- ,� cently, in an attempt to break the housing supply deadlock, Sacramento legis- lators have moved to require that individual communities accept their "fair share" of new housing supply, and identify and remove local roadblocks to new construction.Z Despite the flurry of legislative activity to promote housing construction and facilitate the availability of affordable housing, little effort has been made to assess the cost and price -push effects oz local land use controls on housing. The barriers to such a complete assessment are substantial and center on the lack of ' good quality land supply and price data, difficulties in delineating meaningful housing submarkets, and separating demand side forces from supply side con- , straints. This paper reports on our efforts to overcome these barriers and provide some further insights into the relationship between local land use controls and the operation of urban housing markets. The study region is the nine -county San Francisco Bay Area—a metropolitan area containing over 100 independent units of local government with widely divergent approaches to controlling development. t The second part of this paper presents a simple typology for understanding and organizing the effects of land use/development controls on housing markets, and within such a context, provides a further review of recent empirical work. Part three provides soine additional insights into the various rationales behind the adoption of land use controls, and reasons why the Bay Area's housing supply i i 19821 Land Use Controls and Housing Costs [ 69 f i crunch has now reached crisis proportions. Part four presents the results of our empirical analysis of the effects of land use and development controls on new housing prices. i THE DIRECT AND INDIRECT EFFECTS OF DEVELOPMENT ! CONTROLS ON HOUSING PRICES - California cities and counties have a variety of tecluliques at their disposal for regulating the type, quality, and timing of new development. In addition to such traditional controls as zoning and subdivision restriction, cities may: establish urban limit lines, or borders beyond which new development will not be allowed to occur—effectively setting vacant land supplies. routinely bargain with builders to reduce densities and mitigate negative environmental impacts. use slope-based/zoning, a technique for reducing hillside development den- sities. together with willing farmland owners, establish 20 -year agricultural pre- serves. levy development fees and charges to pay for the construction of needed on-site infrastructure (sewer and water ones), and also to subsidize the purchase of parkland and the maintenance of local schools. Lm growth management programs to airectly limit developmsnt. To summarize the range of effects of land use and development controls on new housing prices, we have identified four direct and two indirect sous of ef- fects. Presented in Exhibit 1, these effects are best considered generic; that is, their relative impact on housing prices will depend largely on the level and char- acter of housing demand, the proximity of regulated communities to other juris- dictions, and the cummulative degree of restrictiveness generated by local land use policies. Direct Effects The most common way in which land use and development restrictions af- fect housing prices is by directly increasing builder costs—increases which under most circumstances are passed on to homebuyers in the form of higher prices. 701 AREUEA Journal [ Vol. 10 EXHIBIT 1 DIRECT AND INDIRECT EFFECTS OF ENVIRONMENTAL CONTROLS ON HOUSING COSTS Effect on Housing Costs Land -Use Controls Most Likely to Generate Effect Direct Effects: Increase in Land costs Zoning, urban limit lines, density constraints, growth - management timing ordinances and permit programs Increase in lot•prepara- Subdivision requirements, growth -management timing tion costs ordinances, dedication requirements Shifting development Capital -budgeting programs, fees and development costs from public to charges, dedication requirement subdivision require - project ments Administrative and All land -use controls to some extent. Costs increase with delay costs the relative complexity of the regulation Indirect Effects: Facilitating monopoly Controls that restrict the number of developers operat- power Ing in communities will allow builders to charge excess housing prices. Regulations act as barriers to market entry, reducing competition in the housing market Market Reorientation Restrictions on devetopment often force developers to reorient their projects to higher -income customers Reducing supplies of vacant land, or restricting the permissible intensity of residential development can substantially affect land costs. As Ohls, Weisberg and White [5J have illustrated, zcning regulations which restrict vacant land sup- plies below the levels which would normally be exchanged in the market tend to increase land costs. Stull [6) has shown that communities adopting policies which shift land away from residential uses and toward employment -generating uses can expect residential land prices to increase as employment growth accele- rates the demand for housing. Unfortunately, direct empirical estimates of the inflationary effects of zoning on housing prices are not widely available. Accordingly, Davies [7] has used a simulation model of the London, Ontario area to examine how municipal actions increasing the supply of developable lots would affect housing costs. For 1967, Davies tested an increase in the supply of lots equivalent to 50 units, or about 11%. The simulation results suggested a corresponding lot price reduction of about $200 per lot (4.5%). A second simulation experiment for 1973 revealed that the same absolute increase in land supply (250lots) would generate a $244 - per -lot decrease in price (3.4%). But although the effect on lot prices is signifi- cant in both cases, the effect on house prices is less than 1%. ibir g housing price and land use data for suburban Boston communities, Stull [8) tested the relationship between land use controls and housing prices. )82 ] Land Use Controls and Housing Costs [ 71 fter controlling for accessibility, housing stock characteristics and the quality public services, Stull found that housing prices were lower in communities ith greater proportions of vacant land. In addition to reducing the supply of residentially developable land, zoning :ts to restrict development intensity. While large -lot zoning, on the one hand, •nds to reduce the per -acre price of raw land, such reductions in price may be ffset by higher land requirements. Interpolating from Peterson's empirically etermined land -price gradients for Fairfax County, [9] Virginia, illustrates the otential cost effect of large -lot zoning. At a distance of ten miles from the ur- an center, Peterson found that large parcels zoned 1h, 1, 2 and 10units per acre ,ere selling for $5,800,$7,900, $13,700 and $32,000per acre, respectively. Con- erting these acre prices to per -lot values, the prices implied by the %, 1, 2 and 0 units per acre zoning are $11,600,$7,900, $6,850 and $3,200, respectively. A second type of direct cost effect of development controls is the increase in he cost of lot preparation and home construction. Numerous estimates of the )rice effects of subdivision and building code requirements are available, some )etter than others. In a 1976 study of Jacksonville, Florida, the Urban Land nstitute [ l 0] found that locally mandated changes in water system design and street width requirements added $830 to the cost of producing a "finished" lot 1976 dollars). In an earlier study for the Kaiser Commission, Burns and Mittel- )ach [ 11] estimated the inflationary impacts of excessive subdivision and zoning -equirements at 2 to 4% of the price of new housing. While it is difficult to differ- -ntiate between necessary and excessive requirements, available evidence sug- gests that some cost reductions might be attainable by reducing subdivision standards. Short-term cost reductions are unlikely in the case of building codes, however, as most builders have fully integrated building code requirements into their production technologies.3 A third direct cost effect of development regulation is generated when local governments attempt to shift the public service costs to new development back to private builders. Traditionally, municipal governments have shouldered the pub- lic service costs of new construction. Recently, however, an increasing number of communities have begun to impose additional fees, taxes and land dedication requirements on project developers. This trend was particularly noticeable in California following the 1978 passage of Proposition 13, which limits home- owner property tax payments to 1% of assessed value. A recent survey of Bay Area communities (I2] revealed substantial increases in development fees and taxes following Proposition 13. For example, in 1976, prior to the passage of Proposition 13, development, utility and impact fees averaged $ l 121 for the standard single-family home. By 1979, mean fee levels, in nominal terms, had risen to $ 1907, an increase of 70%. After discounting for inflation, fees rose by some 35%. As in the case of subdivision requirements it is difficult to determine when fees and charges are excessive. Charges that reflect the actual costs of providing public services to new development are reasonable, equitable and j uurnai i Vol. 10 desirable. However, in some instances, cost -shifting appears excessive—when revenues are used to provide services that benefit the general community. The fourth direct effect of development regulation results from adminis- trative delays associated with regulatory compliance. With the widespread use of fiscal and environmental assessments, particularly in California, the time re- quired to obtain development approval has increased tremendously over the past decade. A national survey [ 13] of the time Isngth of the development ap- proval process found that in 1970, 72% of the developers interviewed obtained project approval in seven months or less. 97% obtained approval in on year or -1 less. By 1975, only 15%of the survey respondents had obtained permits in less than Seven months, and only 42% obtained development approval in one year or less. To be sure, the blame for such delay rests as much with builders who pre- pare improper submissions, as with overzealous reviewers. Regardless of fault, however, delays in the development review process generate economic costs - 7) costs generally borne by consumers. Delays result in increased land -holding and overhead costs, development loan interest costs, exposure to inflation and op- portunity costs of tying up capital. In a 1978 Rice University study of Houston area builders, [ 14] the overall costs of approval delays were estimated to add between $388 and $596 to the per unit cost of new housing. In California, delays associated with California Environmental Quality Act compliance were estimated to add between 4 and 7% to the selling price of new units (1974) [15] . IndirectEffects In addition to the direct cost effects of land use regulations, community con- trol over residential development often confers significant monopoly power on developers and alters marketing and pricing decisi=. Monopoly power allcm builders to charge excess prices for housing and increased production costs may force developers to reorient residential projects to high-income markets. Regulation establishes monopoly power in a variety of ways. By restricting developable land supplies, the potential for market entry and the possibility of increased competition are reduced. Studying the Edmonton, Canada, housing market, Cook f 16] found substantial concentration among developers operating in the city's six restricted development areas, with the four largest builders supplying 64$ of the single-family lots produced between September 1973 and August 1976. A closer inspection of the six citydesignated development areas indicates that each is controlled by one developer. Land use regulations which rely on complex administrative procedures act as barriers to market entry. If the level of complexity is great, potential developers may be reluctant to enter a local market, particularly if they perceive that tbose builders who have already established good working relationships with local planners arc more likely to obtain development permission. A study of two San J 19821 Land Use Controls and Housing Costs [ ?3 Jose, California builders [ 17] found that between 1967 and 1976, after holding land and materials cost increases constant, and discounting for mandated in- creases in housing quality, the two builders had increased their profit margins by between 15 8 and 231 % (constant 1976 dollars). The researchers concluded that such excess profits were partly the result of reduced competition. The second type of indirect effect of land use regulation is the reorientation of residential projects. The previous discussion of monopoly suggeststhat devel- opers may increase prices to merely match cost increases. However, there is an- other reason for rising prices. Often, builders find that project marketability declines as prices rise. By changing the product only slightly, many builders are able to reorient their projects toward higher -income consumers, a reorientation which increases profits. Development restrictions which limit residential den- sities and increase production costs often force builders to scrap plans for high- volume, affordable housing in favor of a more limited number of higher -priced units [18]. THE BAY AREA PERSPECTIVE The production cost side is only half the development control story. The other half, as noted above, concerns the interplay of supply and demand; whether in the face of rising housing demand, development controls restrict the aggregate supply response, and in doing so, push housing prices upward. Although such a market dynamic is difficult to verify empirically (in part because of the afore- mentioned problems with identifying submarkets and pure -demand side effects), with respect to the San Francisco Bay Area, there is substantial evidence that such supply constraints are in fact contributing to the continuing chnib of new and existing housing prices. In 1978, the San Francisco Bay Area edged Washington, D.C. for the dubious distinction of having the highest housing prices of any metropolis in the United States. In 1981, the median sales price of a new home in the Bay Area stood at $1 14,000—a figure more than $30,000 above the national average, and repre- senting a 269% increase over a ten-year period. [19] At the Same time that prices have been rising, vacancy rates have been falling. According to a recent estimaie from the Association of Bay Area Governments, in 1981, the average Bay Area vacancy rate stands at roughly 2%—down from the 5% vacancy rate of only five years k.;o. [201 What combination of market factors explains these trends? Like many grow- ing metropolitan regions, the Bay Area has faced considerable demand pressure from growiryg households and industries. The so-called baby boom generation has now reached prime house -buying age. Changing cultural and social values }lave dramatically increased the formation of households as more individuals seek separate residences. Employment growth and increased immigration to California and the Bay Area have further accelerated household growth. The demand for housing is strong in the Bay Area, but it only partially explains rapid housing price inflation and high prices. The other key ingredient is insuf- ficient supply. Unlike other high growth Sunbelt regions, the Bay Area is, relatively speaking, in short supply of developable land. Extensive land development since World War II, the increasing use of growth management controls, more restrictive land use and environmental regulations, and the "go-slow" development posture created by the passage of Propositions 4 and 13 have significantly affected land conversion in the region. Despite the existence of what in absolute terms is an enormous supply of vacant land, much of this total cannot be developed because of rugged topography or enviromental fragility. Other vacant lands that could be developed are restricted from um by local land use controls. A 1975 inventory of land gee recorded that of the region's 4.5 million acres, only 350,000 acres were vacant and "developable." [21 ] If only the acreage zoned for residential development and likely to be serviced is considered, the total shrinks to 161,800 acres, a number which at prevailing densities (8.7 units per acre), would accomo- date some 1.4 million additional housing units. But unfortunately, as ABAG points out, residential densities are falling, and rather rapidly. If present land conversion trends continue, the projected 1990 regional housing demand will not be accomodated— there simply will not be enough land for residential devel- opment. [221 Are such trends truly worrisome? After all, as vacant land supplies diminish, land prices can be expected to rise, signaling to developers and planners that land : use intensities must also rise. Unfortunately, there now appear to be substantial and permanent forces within the San Francisco Bay Area which will prevent a move to higher housing densities. Among these forces: The Rise of Local Growth Controls: The pro -growth attitude of most Bay Area communities in past decades has been replaced by a slow -growth posture, brought on by rising fiscal worries generated by Proposition 13. Cities that once relished being regional growth centers now view growth with much skepticism. With the new fiscal calculus of Proposition 13, single-family development usually generates higher public sector costs th-�n revenues." 2ns fact, in conjunction with a greater recognition of the environmental impacts of development, has led some communities to reduce the amount of land available for residential development. Coupled with the lack of developable land in older Bay Area cities, development opportunities are becoming scarce, and some builders are leapfrogging out to exurban agricultural areas. lobs, But Ab Housing: Another result of Proposition 13 has been that numerous communutes have altered their approach to land um pianning and zoning. Caught in a fiscal squeeze, many towns have stepped up efforts V 19821 Land Use Controls and Housing Costs f 75 4 to increase their tax base by attracting more commercial, office and light industrial development. But while attempting to attract economic develop- ment, most communities have not concomitantly adjusted their zoning to provide housing for additional employees. As a result, new employees, particularly those just migrating into the region, may find it increasingly difficult to purchase their own homes, even assuming that currently high mortgage interest rates abate somewhat. Increased Development Fees and Charges. In addition to limiting devel- opment of fiscally "unprofitable" housing, most Bay Area communities have dramatically increased the fees and charges they levy an developers. A recent ABAG study found that total development fees for single-family homes range from $800 to nearly $6,000 per unit. [23) Crucially, the ABAG study reveals that the twenty-two communities charging the high- est fees (ranging from $3,000 to $6,000) are all located in the developing suburban reaches of the Bay Area—communities, which not coincidentally, boast some of the region's more affordable housing. How have these various factors affected housing supply? Exhibit 2 shows that region -wide, new housing production dropped from 46,235 units in 1977, to 38,472 units in 1978, to 33,763 units in 1979. In contrast to the Bay Area ex- perience, statewide housing starts in 1978and 1979were well above 1976levels. What is perhaps more alarming than the decline in regional housing starts is the fact that the greatest absolute 1977-1979 declines occurred in the three counties --Contra Costa, Santa Clara, and Sonoma—which together contain the majority of the region's developable land supply. Falling residential densities, growth controls, job growth at the expense of housing development, and the "obstructionist" attitudes of citizens and com- munities toward new development—all of these factors are acting to reduce the supply of vacant land available to housing development. How are these trends likely to affect housing costs? Economic theory suggests that if the current reg- ime of local land use and development control policies remains unchanged, land and housing prices alike will continue in an upward spiral, placing extreme bur- dens on low and moderate -income households and ultimately slowing the region's economic growth rate. Higher land and housing costs may also act to push up wage rates, as workers struggle to pay higher housing costs. Carried to their logical extreme, higher land and wage costs may reduce the Bay Area's attractiveness to business and industry. Such concerns, which may seam exces- sively "long run" to economists, are nonetheless very real to Bay Area civic and business leaders. 88 ] AREUEA Journal [ Vol. 10 CONCLUDING COMMENTS While it is widely agreed that land use regulations contribute to housing price inflation, little supporting empirical evidence is available. Although tenta- tive, the results of this paper illustrate that density controls and land availability do systematically affect the price of new housing units. However, our research also indicates that the direct cost effects are not as great as some critics of land use controls allege. For example, according to econometic estimates of new housing prices, the combined effect of increasing development densities by one unit per acre, reducing development fees by 50%, and doubling supplies of va- cant land—all drastic steps—would be to lower the sales price of a new home by $6,000. This estimate amounts to roughly 6% of the average price of a new Bay Area home in 1979. New homes prices in growing suburban communities are less sensitive to limits on development densities and vacant land supplies and slightly more sensitive to increasing development fees. Our results also support the con- tentions of suburban builders who report that planning and development fees are added to the price of new housing on a one-to-one basis. In other words, for every one dollar increase in fees, the list price of a new home increases by one dollar. However, we note that land costs are more important to builders, and ac- cordingly, density limits become critically important in determining project selling prices and profit. To the extent that builders can distribute higher land costs, as well as infrastructure costs, over a greater number of constructed units, higher single-family housing densities are crucial for holding down selling prices while maintaining profit levels, The importance of the suburban housing market in acting as a relief valve for Bay Area housing demand is implied by Exhibit 5. Although changes in sup- ply do not greatly affect housing prices in the region as a whole, the flow Cf new units onto the market is a major determinant of housing prices in expanding suburban markets. For example, a 500 -unit increase in the flow of new homes into a suburban market would imply a decline in all suburban home prices of nearly $6,000. Thus policies which greatly restrict new construction and/or den- sities in active suburban communities are found to be inflationary. A logical extension of this finding is that if local governments in the San Francisco Bay Area are committed to reducing housing costs, they should consider loosening density restriction or other controls which inhibit the flow of new housing onto the market. It is important to be careful in drawing rigorous conclusions from the results of partially specified econometric models, particularly when the observation set consists of city averages instead of well-defined economic agents. And because the significance levels of the estimated development policy coefficients vary de- pending on how the models are specified, the link between development controls and higher home prices must still be regarded as unproven. Nonetheiess, the results presented here are surprising for their consistency, and their agreement 19821 Land Use Controls and Housing Costs [ 89 with expectations. They suggest that the housing price effects of pursuing re- strictive growth.control policies in expanding urban areas, far from being smU and localized, are significant and widespread. NOTES 1. Frequently, communities circumvent the intent of AB 884 by requiring builders to sign a waiver exempting the city from complying with AB 884. 2. Senate BLU (SB) 2853, was enacted by the legislature in 1980. It mandates that re- gional councils of government (COGS) allocate new housing construction to cities on the basis of projected growth and land availability. Initial allocations for the San Francisco Bay Area were to be published during the summer of 1981. 3. According to a recent survey by the National Association of Homebuilders, [ 37] when asked to list major construction problems, responding homebuilders listed building codes eighth—well behind increased labor, materials and land costs. Moreover, only 1.31/6 of those responding listed building codes as their "most Significant" construction -related problem. 4. Whether or not new residential construction does or does not "pay its own way" will depend on a number of factors, including the existence of excess capacity in public services, the level of demand for such services by new residents and prevailing tax rates. In California, where Proposition 13 limits yearly property tax payments to 1% of assessed value, muni- cipalities continue to experience substantial revenue si=Lf0s. In the San Francisco Bay Area, Proposition 13 has led numerous communities to examine the net fiscal costs of new development through the use of C RIS (Cost -Revenue Impact Study), a municipal finance model developed by the Association of Bay Area Governments. In most of the cities in which CRIS has beenused, new single-family development was found to generate insufficient tax revenues to cover the accompanying increase in public service costs. 5. The extent to which households trade between access to employment and land and housing prices is the basis of much of contemporary urban economics. For the basic excep- tion, see Edwin S. MMs [33). 6. This argument, first offered by Tiebout (1956). is best presented in Wallace Oates [36). 7. Data were provided by Professor Kenneth Rosen, Director of the Center for Real Estate and Urban Economics, University of California, Berkeley. 8. The use of time -series dummy variables is not without drawback. Although we argue that the variation in new home prices attributable to increasingly stringent (across time) land use controls is far less than the price variation attributable to cross -community land policy differences, the time -series effect is not insignificant. By using single -time dummy variables (and thus lumping together the time -series variance cf all the independent varia- bles), we sacrifice the ability to identify precisely these time -series effccts and bias the re- sulting yearly price indices. An alternate method for estimating time -series hedonic price indices, one which parti- ally obviates the problem of time -series dummy variables, has been suggested by Grtiiches ( 381, and more recently by Palmquist [391. In the present case, the technique consists of- Estimating £ Estimating the New Housing Price Model for each sample year (1977, 1978, 1979) in both linear and logarithmic form. Inserting the mean values of the independent variables for the base year ( 19 7 7) into the separate equations estimated frr 1978 and 1979. E W, J 901 AREUEA Journal [ Vol. 10 Comparing the estimated values of the dependent variable (NPRICE) for 1978 and 1979 to the base year 1977. The resulting ratios are quality -controlled tithe -series indices. In the table below, the indices derived using the Griliehes/Palmquist technique are com- pared with the hedonic indices derived through the use of time -series dummy variables. In neither the linear case cr the logarithmic case does the Gri'liches/Palmquist index seem reliable. We offer two reasons for the discrepancy. First, the observationsare not individual home transactions (as Palmquist suggests is appropriate), but instead are city-wide averages. Second, the estimated coefficients vary widely when separate year models are attempted, in part because we lack a sufficient number of yearly observations. Year Time -Series Quality -Controlled Dummy Variable Hedonic Index Linear Log Linear' Log 1977 1.00 1.00 1.00 1.00 1978 1.116 1.26 1.33 1.98 1979 1.28 1.38 1.31 1.19 REFERENCES (11 Bernard Frieden, The Environmental Protection Hustle (Cambridge, MA: MIT Press, 1979). [2) Claude Gruen, "The Economics of Petaluma: Unconstitutional Regional Socio - Economic Impacts," in R. Scott, ed., The Management and Control of Growth, Vol. II (Washington, D.C.: The Urban Land Institute, 1975), pp. 173-186. 131 David Dowall, The Suburban Squeeze: An Examination of Suburban Land t:om,er. cion and Faqulaticn in the San Francisco Bay Area. Draft (Institute for Udm and Regional Development, 1981). (4) Office of Planning and Research, Local Government Planning Summary: 1975 (Sac- xamento, CA: State Office of Planning and Research. 1975). 151 James Ohls, Richard Weisberg and Michelle White, "The Effects of Zoning on Land Value," Journal of Urban Economics, 1(1974): 428-444. [6) Wiliam J. Stull, "Land Use and Zoning in an Urban Economy," American Economic Review, 64 (1974): 337-347. [7) Gordon Davies, "A Model of the Urban Residential Land and Housing Markets," Canadian Journal ofEconomics, 10(1977) : 393-410. f 8 William J. Stull, "Community Environment, Zoning and the Value of Sitgle Family Homes," Journal ofLaw and Economics, 18 (1975): S3S-SS7. (9) (barge Peterson, "Land Prices and Factor Substitution in the Metropolitan Housing Market." Working Paper No. 0875-02-01 Washington, D.C.: The Urban Institute, 1974). (10) The Urban Land Institute, "Effects of Regulation on Housing Costs: Two Case Stu- dies." UM Research Report No. 27 (Washington, D.C., 1977). (11) Leland S. Burns and Frank Mittelbach, "Efficiency in the Housing Industry," in Re- port cf the President's Committee on Urban Housing,A Decent Home, Vol. Il (Washington, D.C.: U.S. Government Printing Office, 1968),p. 133. [12) Stuart Gabriel, Jennifer Watch and Lawrence Kau, "Land use Regulation and R - position 13," IBER Working Paper No. 84.4 (Berkeley, CA: University of Cali- fornia —Berkeley ali.fornia—Berkeley,1980). 1982 J Land Use Controls and Housing Casts 1 9 1 [131 Srephen Seidel, Housing Costs and Government Regalatim (New Brunswick, NJ: Center for Urban Policy Research, L978). 1141 Rice Center for Community Design and Research, The Delay Costs of Government Regulation in the Houston Housing Market (Houston, TX: Rice Center, 1M). 1151 Environmental Analysis Systems, Inc., "The California Environmental Quality Act," prepared for the State Assembly Committee on Local Government, 1975. [161 Richard Cook, "Lot Prices and the Land Development Industry in Edmonton, Can- ada" (Berkeley, GA: Department of City and Regional Planning, University of California—Berkeley, 1977)_ 117] The Urban Land Institute, Effects ofRegulation on Housing Costs, op. cit. 1181 Bernard Frieden, The Environmental Protection Hustle, op. cit. 1191 Real Estate Research Council of Northern lif=da,Northern California Real Estate Report (San Francisco, CA: 1980) . (201Association of Bay Area Governments, 1981. 1211 Association of Bay Area Govetnments, Summary Report: Provisional Series 3 Projec- tions (Berkeley, CA: ABAG. 1977) . 1221 Theresa Hughes and Associates, from ABAG, San Francisco Bay Area HousingActi- vity Report, No. 2, 1980. (231 Association of Bay Area Governments, Development Fees in the San Francisco Bay Area: A Survey (Berkeley, CA: ABAG, 1980) . [241 Ronald Lafferty and H. E. Frech, "Community Environment and the Market Values of Single Family Homes: The Effect of the Dispersion of Land Uses,"Journal of Law and Economics, 21 (L978): 381-394- 1251 Stull, op. cit. 1261' Frederick Reuter, "Externalities in Urban Property Markets: Arl. Empirical. Test of the Zoning Ordinance of Pittsburgh," .'ournal ofLaw and Economics, 16 (1973): 313-349. 1271 J. P. Crecine, 0. A. Davis and j. E. Jackson, "Urban Property Markets: Some Empiri- cal Results and Their Implications for Municipal Zoning," Journal of Law and Economics, 10 a 967) : 79-99. 1281 Association of Bay Area Governments, Local Land Poky Survey (Berkeley, CA: A - BAG, 1977) . [291 Ibid. [301 Gabriel, op. cit. [311 California Department of Finance, Yearl} CensusReports (Sacramento, CA: Depart- ment of Finance, 1980). (321 Ibid. 1331 Edwin S. Mills, Urban Economics (Chicago, IL: Scott Foresman, Inc., 1M). [34) Metropolitan Transportation Commissin, "1978 Travel Demand Study" (Berkeley, CA: MTC, 1M). 1351 Urban Decision Systems, "Updated Household Income" (Los Angeles, CA: Urban Decision Systems, 1980). [361 Wallace Oates, Fiscal Federalism (New York, NY. Harcourt, Brace and Jovanovich, 1972) . [371 Michael Sumichrart, et al., Profile of the Builderand His Industry (Washington, D.C.; National Association ofHomebuilders, 1979). [381 Zvi Griliches, Price Indexes and Quality Change (Cambridge. MA: Harvard University Press, 197 1. (39) Raymon(' Palmquist, "Alternative Techniques for Developing Real Estate Price In- dexes," Review of Econcadcsand Statistics, 62 (August 1980):442-448. BIBLIOGRAPHY FOR FURTHER RESEARCH Avrin, M. E. "Some economic effects of residential zoning in San Francisco." In Residential Location and Urban Housing Markets, edited by G. X, Ingram. Cambridge: Ballinger, 1977. Babcock, R. F. and F. P. Bosselman. Exclusionary zoning: Land -use regulation and housing in the 1970s. New York: Prager, 1973. Bergman, E. M., et al. Internal validity of policy related research on development controls and housing costs. Chapel Hill: Center for Urban and Regional Studies, University of North Carolina, August, 1974. Bergman, E. M., et al. A Policy guide to evaluations of policy related research on development controls and housing costs. Chapel Hill: Center for Urban and Regional Studies, University of North Carolina, August, 1974. Bergman, E. M., et al. External validity of policy related research on developnent controls and housing costs. Chapel Hill= Center for Urban and Regional Studies, University of North Carolina, August, 1974. B.ickert, Browne, Cuddington & Associates. An analysis of the impact of state and local government intervention on the home ,. building process in Colorado 1970-1975. Denver: Colorado Association for Housing and building, April, 1976. Brooks, M. E. HousS.ng equity and environmental protection: The Needless conflict. Washington, D.C.: American Institute of Planners, 1976. Brooks, M. E. "Housing trust funds: Lessons from inclusionary zoning." In Inclusionary zoning moves downtown, edited by D. Merriam, D. 3. Brower and P. D. Tegeler, 7-22. Washington, D.C. = Planners, 1985. California Planning Roundtable. "'Welcometo California 1990s, jobs, housing and transportation. ..the great balancing act." Sacramento, California, October 1988. Center for Real Estate and Urban Economics. "San Disgo County's housing cap and growth debates," Quarterly Report. Berkeley: University of California, 1968. Chambers, D. N., and D. B. Diamond, Jr. "Regulation and land prices." Unpublished manuscript, National Association of Realtors and National Association of Homebuilders, June 1, 1988. Dowall, D. E. "Reducing the costs effects of local land -use controls." Journal of the American Planning Association (April 1981): 145-153. Dowall, D. E. "Local -regional planning conflicts: ASAG's compact growth plan and its effect on the metropolitan housing market," Presented at the Annual Conference of the American Institute of Planners, New Orleans, September 28, 1378. Downs, A. The Costs of sprawl, Washington, D.C.: The Real Estate Research Corporation, 1974. Evans, M. S. "Costs of regulation." Fort Lauderdale Sun Sentinel (December 31, 1975) : 6A, col. 4. Feagin, J. R. "Tallying the social costs of urban growth under capitalism: The Case of Houston," in Business Elites and Urban Development, edited by S. Cummings, 205-234. Albany: State University of New York Press, 1988. Fischel, W. E. "Does zoning matter? Empirical evidence on zoning, externalities, and housing costs." In the Economics of Zoning Laws: A Property rights approach to American land -use controls, chapter 11, 231-251. Baltimore: Johns Hopkins University Press, 1985. Frech, H. E., III. "The California Coastal Commission- Economic impacts." In Resolving the housing crisis, edited by M. Bruce Johnson. San Francisco: Pacific Institute for Public Policy Research, 1982. Frieden, B. J. and A. P. Solomon. "The Controversy over home ownership and affordability," AREUEA Journal 5 (1977). Greiner, A, "'The Housing affordability gap and Boston's economic growth: Potential for crisis." M.I.T., Department of Urban Studies and Planning (October 1987). Draft for discussion only. Johnson, R. A., S. I. Schwartz, G. Wandesford- Smith, and K. E. Savage. "Mandating affordable housing by the use of state permit powers." Unpublished paper, University of California, Davis, 1986. Katz, L. and K. Rosen. "The Interjurisdictional effects of growth controls on housing prices." Journal of Law and Economics 30 (April 1987): 149-160. Kroll, C. A. and J. D. Landis. "San Diego County's housing cap and growth debates.'" Quarterly Report, Center for Real Estate and Urban Economics, vol. 3, 1988. Schwartz, S. I., D. E. Hanson, and R. Green. "The Effect of growth control on the production of moderate -priced housing." In Growth management: Keeping on target? edited by D. R. Porter, 15-20. Washington, D.C.: Urban Land Institute, 1986. Seidel, S. Housing costs and government regulation: confronting the regulatory maze. New Jersey: Center for Urban Policy Research, 1978. Seidel, S. Government regulations and housing costs. New Brunswick. Center for Urban Policy Research, 1977. Sidor, J. "Affordable housing, land supply, and land development: California." Council of State Comnunity Affairs Agencies, Washington, June 1983. The Benefits and costs of metropolitan growth: A Research approach. The Institute for Policy Studies, Johns Hopkins University, Baltimore, Maryland, 1989. The Costs and benefits of local growth control ordinances: A Preliminary assessment. Proposal to the California Policy Seminar. Turner, M. A. Housing needs to the year 2000. Washington, D.C.: National Association of Housing and Redevelopment Officials, 1987. Urian land Institute. The cost of delay due to government regulation in the Houston housing market. Washington, D.C.: Urban Land Institute, 1979. Urban Land Institute. "The Effects of growth management on housing costs in San Jose, " ULI Research Report #2? , Washington, D.C., 1977. U.S. Department of Housing and Urban Development. Final Report of the task force on housing costs. Washington, D.C.: HUD, 1978. U.S. Department of Housing and Urban Development. The President's National Urban Policy Report. Washington, D.C.: HUD, 1978-88. White, J. R. "Large lot zoning and subdivision costs: A Test." Journal of Urban Economics (May 1988): 370-84.