Journalof PublicTransportation 103

California'sTransit Village Movement

RobertCervera Universityof ,Berkeley

Abstract Transitvillages-dense, mixed-usecommunities near rail stops-could increase rail ridershipand reduceautomobile dependency; however, few good examplesexist in the U.S.today. Barriersto buildingtransit villages include questionable market viabil­ ity, conservativelending practices, and neighborhoodopposition to multi1amilyhous­ ing. Thispaper shows,however, that there is a reasonablystrong marketdemand for well-designedtransit-oriented neighborhoods. After viewingvisual images of simulated transitvillages, more respondents from theSan FranciscoBay Area expresseda willing­ nessto live in a moderatelydense community with nice amenities than in one with a third lowerdensities but little neighborhoodopen space or consumerservices. Many current occupantsof transit-basedhousing in Californiaare young professionals living in one or twoperson householdswith just one car. Whatmost distinguishes tenants of transit­ basedhousing is theirtendency to workin downtownsand otherlocations well-served by rail transit.The demand for goodquality housing near rail has allowedsome rail-served apartmentsin the Bay Area to commandrent premiums. Strong market interest in rail­ based housing,coupled with recentstate enablinglegislation, bode favorably for the future of transitvillages in California.

Fall /996 104 Journalof Public Transportation

Introduction In California,a movementis currentlyunder way to createtransit villages around urban rail stations.In October1993, Governor Pete Wilsonsigned the TransitVillage Act, Assembly Bill 3152,which encourages cities and countiesto build higher densityhousing and more concentrateddevelopment around the state's rail stops. Californiahas investedover $10 billion in urban rail transit infrastructureover the past 20 yearsand is poisedto spendupwards of $60 bil­ lionmore over the next 30 years(mainly in SouthernCalifornia). Yet most devel­ opmentin recentyears has turnedits backon transit,focused on freeway-served suburbancorridors instead. Since the 1972opening of the BayArea Rapid Tran­ sit (BART)system, 35 millionsquare feet of privateoffice space has beenbuilt in parts of Alamedaand ContraCosta Counties unserved by BART,compared to only9 millionsquare feet withinone-half mile of BARTstations in the two East Bay counties(mostly limited to downtownOakland, Walnut Creek, and Con­ cord)(Cervera and Landis1995). One consequenceof growthoccurring away from transit stops has been mass transit's decliningmarket share of metropolitantrips. Whiletransit jour­ neys rose in absolutenumbers in Californiaduring the 1980s( one of the few stateswhere this was the case),transit's share of commutetrips fell in the state's four largestmetropolitan areas, despite their new rail systems:greater Los Ange­ les-5 .4 to 4.8 percent;San FranciscoBay Area-11.9 to IOpercent; San Di­ ego-3. 7 to 3.6percent; and Sacramento-3.7to 2.5percent. Nor do thesetrends appearto be slowing.A "Stateof the Commute"report by the CommuterTrans­ portationServices (1994)-the annualtracking study of commuterbehavior in the greaterLos Angelesregion-showed SouthernCalifornia's drive-alone rate increasedfrom 77 percentin 1992to 79 percentin 1993. Giventhe tremendoussunk investment states like Californiahave in urban rail transit, these ridershiptrends are worrisome.Transit villages, proponents argue, will help reverse,or at least stave off, the trend toward growingauto­ dependencyand shrinkingtransit market share. Besides capitalizing on expen­ sive public investmentsin rail, proponentsargue that focusingfuture develop­ ment aroundrail stopswill produceother social benefits: increased regional ac-

Fall 1996 Journalof PublicTransportation 105 cessibilityand reducedtraffic congestion along rail-served corridors; a morecom­ pact, sustainableurban form that conservesenergy and reduces pollution; in­ creasesin affordablehousing; more choices on whereto live and how to travel; increasedmobility for the transportationdisadvantaged; and the creationof vil­ lage environmentswhere people fromall walks of life come into daily contact with each other,similar to America'sstreetcar communities of yesteryear. Thisarticle examines recent progress toward creating transit villages around California's urban rail stations. Examplesof housing developmentclustered around Californiarail stations are described,followed by a discussionof the opportunitiesand barriersto transitvillages. The marketpotential for large-scale transitvillage development is then assessedusing visual simulationtechniques. Characteristicsof California'sexisting rail-based housing projects are later pro­ filed in terms of tenant composition,ridership levels, and rent premiums.The article concludeswith a discussionof California'stransit village legislationand other public policy initiativesthat might be pursued in promotingfuture rail­ orienteddevelopment.

DefiningTransit Villages The somewhatnostalgic-sounding name of"transit villages" has gainedcur­ rency in recent years to describeplaces conduciveto transit riding-compact, mixed-usecommunities that, by design,invite residents, workers, and shoppers to drive their cars less and use transitmore. Under California'sTransit Village Act, transit villagesextend roughlya quartermile from a transit station, a dis­ tance that can be coveredby foot in about five minutes;beyond this distance, suburbanitesare far more likelyto driveto their destinationsrather than walk to a stationto accessa train.The centerpieceof the transitvillage is the stationitself and the civic and publicspaces that surroundit. The transit station is what con­ nects village residentsand workersto the rest of the region, providingconve­ nient and readyaccess to downtown,major activity centers ( e.g., sportsstadium, collegecampuses), and otherpopular destinations. The surrounding public square or open area servesthe very importantfunction of being a communalgathering place and a site of specialevents and celebrations-a modern-dayagora. In the

Fall /996 106 Journalof Public Transportation mornings,the publicsquare might be convertedto an outdoorfarmer's market, populatedby flowerstalls, fruit stands,and food vendors.On weekends,con­ certsmight be heldthere. What is importantis that the transitstation functions as a window,or gateway,to the rest of the regionand is physicallytied to and asso­ ciatedwith the village'smajor gathering place. Such settings are commonat rail stationsthroughout Europe. Residents are drawnto transitnodes by the attrac­ tivenessand vibrancyof the surroundingcivic areas. And concessionaires, street artists,and neighborhoodmerchants are drawnto these settingsbecause of the heavywalk-on traffic. It's a win-winproposition. Transitvillages are hardlynew ideas. They borrow from the visionsof early city plannerslike EbenezerHoward in Englandand FredericLaw Olmstead and EdwardBellany in America,who advancedthe idea of buildingpedestrian-ori­ ented gardencities. Howard'svision was to build self-sufficientsatellite com­ munitiesof around 30,000inhabitants that would orbit London,separated by protectedgreenbelts and connectedby inter-municipalrailways. Some vestiges of transitvillages survive in the formerstreetcar suburbs of tum-of-the-century America,such as ShakerHeights in Cleveland,Chestnut Hill in Boston,River­ side nearChicago, and RolandPark in Baltimore.Streetcar suburbs depended on pedestrianaccess to transitto reachdowntown jobs and neighborhoodcenters, since manywere built prior to the inventionof the automobile.America's early rail-servedneighborhoods featured a rangeof housingfrom large estates to small cottagehouses, had distinctivegridiron street patterns, and focusedon a promi­ nent civic space near the rail stop to instilla sense of community.In order to attract early residentsto distant suburbs,these early transit villageswere de­ signedas safe, secure,and attractiveplaces-notably with the placementof the traindepot and public square in the heartof the communityand the use of restric­ tive covenantsand otherdevelopment standards to controlthe physicalenviron­ ment. In recentyears, the terms"neo-traditional" development and "new urban­ ism"have gainedcurrency to describeplaces that are compact,"quaint"-feeling, and rich in land-usemixture, and as a result,are moreconducive to walkingand transitriding. New urbanists,like Miami-basedAndres Duany and Californian

Fall 1996 Journalof PublicTransportation 107

PeterCalthorpe, borrow many of the successfulelements from traditional Ameri­ can townslike Princeton, and Annapolis, Maryland. Among the hall­ marksof neo-traditionaldesigns are a commercialcore withinwalking distance of severalthousand residents, a well-connectedgrid-like street network,narrow roads with curbsideparking (to bufferpedestrians), back-lot alleys, diverse land uses, and varyingstyles and densitiesof housing. Outsideof a few tum-of-the-centuryneighborhoods, few good examplesof transitvillage development can be foundin the U.S. today.Of course,there are high-riseapartment towers near subwaysin big cities like New Yorkand Chi­ cago and some recent mixed-useconcentrations near suburbanrail stations in metropolitanWashington, D.C. (e.g., Ballston,Bethesda) and San Francisco's East Bay (e.g., WalnutCreek, Pleasant Hill); however, few such places could be characterizedas "villages."Europe perhaps offers the best modern-dayexamples of transit village development,where dozens of compact,mixed-use satellite communitiesare interconnectedby regionalrail systems in metropoliseslike Stockholmand Copenhagen.Europe's transit villagesare built on a scale that encouragespedestrian circulation. Most rail stops focus on town centerswith a publicsquare and an outdoormarketplace. The accenton livabilityis showcased by pedestrianamenities-park benches,newspaper kiosks, bus shelters, side­ walk cafes, open-airmarkets, and arcadesdesigned to protect pedestriansfrom the elements.In Vallingby,one of Stockholm'srail-served satellites, the rail sta­ tion shares space with a super market,where returningcustomers can do their daily shoppingon the way home. The station is adjacent to a car-free village squarelined with more shops and service establishments, including several daycare centers(Figure I). Morethan 50 percentof Vallingby'semployed residents com­ mute by transit-despite the fact that Swedenhas one of the highestper capita car ownershiprates in Europe(Cervera 1995). It is importantto recognizethat transitvillages are notjust physicalentities. There are importantsocial and economicobjectives behind the transit village conceptas well. One objectiveis to create an urban milieu that brings people from manywalks of life into daily,face-to-face contact. Early streetcarvillages had these qualities.Today's auto-oriented suburbs, in contrast,have segregated

Fall /996 108 Journal of Public Transportation

Figure 1. Vallingby stationar ea,a pedestrian-friendly, car-freet ownce nter. Entry to rail station via shopping center and surrounding civic space.H igh-density mixed land usesn ear station taper off to lower densities farther away. Photo by Jeff Kenworthy. cultures and isolated people by age, class, and race-o ld from young, rich from poor, whites from blacks.Soc ial commentators like Anthony Downs ( 1994) blame low-density and class-segregated growth for creating deep divisions in American society and for isolating, both physically and socioeconomically, many blacks, Hispanics, and recent immigrants. Social integration is extremely difficult to achieve in a laissez-faire society with high levels of automobility and personal freedom. Transit villages are just one of many ways of building new kinds of communities that offer wider lifestyle choices. By creating an attractive, lively, but safe neighborhood environment, it is likely that a subpopulationof people from different social backgrounds and income levels will be drawn to these set­ tings. While these ideals are admittedly steeped in beliefs of physical determin­ ism, experience shows, both historically and internationally, that transit-oriented

Fall 1996 Journalof PublicTransportation 109 settings can impart a sense of belongingand an attachmentto place, besides inducingpeople to ride transit. Transitvillages must also be economicallyviable and financiallyself-sus­ taining.Creating attractive urban environments that have good transit accessto the rest of the region should,by definition,produce economic benefits. Fore­ most,the advantagesof beingnear rail in an attractiveurban environment should translateinto higher propertyvalues and commercialrents. To the degree that governmentscan recapturesome of these economicbenefits, such as through propertytax proceeds or special benefit assessments,then transit villages, in theory,can become economicallyself-supporting. Transit villages might also spin-offsecondary economic benefits-such as providingopportunities for joint development(e.g., building a retailstore adjacent to a rail stationand generating leaserevenues for a transitauthority), station-area concessions (e.g., food kiosks), and community-basedservices (e.g., operatingjitney connectionsbetween a neighborhoodand the transitstop). Transitvillages could also serveas catalyststo economicdevelopment and communityrebuilding. Recently, the FederalTransit Administration and Depart­ ment of Housing and Urban Developmentjoined forces to create a "Livable Communities"initiative that aims to empowerdistressed inner-city neighbor­ hoodsacross the U.S.by makingthem eligible for specialgrants and tax credits. Somelivable communities, like the Fruitvaledistrict in Oakland,California, re­ ceive urban rail services.The hope is that by creatingbetter quality neighbor­ hoods in areas with superiortransit services, private capital will return to these areas, puttingthem on a road to financialrecovery. In the case of the Oakland's Fruitvaleneighborhood, community leaders hope to one day createa transitvil­ lage focusedon the BARTstation (Knack 1995).Plans call for buildingattrac­ tive apartments,creating a publicsquare, and sitinga child care centernear the station, as well as transformingthe BARTstation itself into a true intermodal transfercenter. The neighborhoodalso hopesto createa mobilityenterprise that wouldprovide neighborhood jitney servicesand reverse-commuteruns to sub­ urbanjob centers,with localresidents in chargeof operating,dispatching, main­ taining,and servicingthe shuttlevans.

Fa/11996 110 Journalof PublicTransportation

Rail-FocusedDevelopment in California Californiais a naturalbreeding grounds for a transitvillage movement in that it is the nation'smost urbanized and transit-oriented state. California has the mosturban rail transitsystems-at currentcount, two heavyrail, five lightrail, and three commuterail services-and the highestmetropolitan population den­ sitiesin the nation(Larson 1993). Whilemodem-day transit villages remain merely a concepttoday, inroads have been made in recentyears in focusinghousing development near rail sta­ tions in Californiathat couldform the buildingblocks of futuretransit villages. Table1 lists26 largehousing projects built within one-quarter mile of California urbanrail stations between 1985 and 1994;collectively, these projects have added over 6,500 housingunits withineasy walkingdistance of rail stops. Most are rental apartmentcomplexes with densitiesof 20 to 60 dwellingunits (du) per acre,well above the 12-15du per acrebenchmark used by plannersas minimum thresholdsnecessary to supportrail in the suburbs(Puskarev and Zupan 1977). Presently,both SantaClara County Light Rail and BARTare in the processof convertingsurface parking lots at severalstations into residential/retail projects. Developershave been attracted to thesesites since, by buildingon existingpark­ ing lots, they do not bearthe risk of negotiatingland purchases among multiple propertyowners, any one of whomcan holdout, therebystalling a project.Bay Area plannershope that buildinghousing atop formerpark-and-ride lots will eventuallylead to mini-communitiesmushrooming around dozens ofrail stations, as wasenvisaged when BART was conceived over 40 yearsago. Localgovernments are doinga lotto promotetransit-oriented development in California.In the BayArea, the citiesof Hayward,Union City, El Cerrito,and PleasantHill have recentlyformed redevelopment districts around BART sta­ tions for the verypurpose of jump-startingnew development(see Figure 2 for a map of the BARTsystem and its stations).El Cerrito'sredevelopment authority has used tax-exemptfinancing to help underwritethe cost of assemblingland and financingnearly $10 millionof the $14 millionin infrastructureimprove­ mentsnecessary to supportseveral housing projects near the Del Norte BART station.The city workedclosely with a developerto createthe Del Norte Place

Fall /996 ~ ~ ::::: Table1. MajorHousing Projects Near California Rail Stations, Built Between 1985 and 1994 s:: \C) Rail System StationName PropertyName City Units YearBuilt Density* Unit Type s -\C) a. BART El CerritoDel Norte DelNorte Place El Cerrito 135 1992 30 rental °' ~ BART PleasantHill Park Regency PleasantHill 892 1992 43 rental "'t:l s:: BART PleasantHill TreatCommons PleasantHill 510 1987 40 rental c::,.. :::::- BART PleasantHill Bay Landing PleasantHill 360 1986-1988 37 rental C") WaysidePlaza: Phase l PleasantHill 36 1985-1986 24 ownership =:;i BART PleasantHill Cl BART PleasantHill WaysidePlaza: Phase 2 PleasantHill 60 1986-1987 60 ownership ::s i3 BART PleasantHill WaysidePlaza: Phase 3 PleasantHill 60 1987-1988 60 rental C BART UnionCity Verandas UnionCity 360 1988-1989 36 rental ~ BART SouthHayward The Foothills Hayward 188 1986-1987 33 rental 5·::s BART Fremont MissionWells Fremont 392 1989-1991 35 rental SCCLRT Almaden The Homesat AlmadenLake San Jose 84 1993 12 ownership SCCLRT Almaden The Apts.at AlmadenLake San Jose 144 1994 37 rental SCCLRT Almaden ParkAlmaden San Jose 370 1989-1994 40 ownership SCCLRT CivicCenter RylandMews SanJose 132 1993 33 rental SCCLRT RiverOaks Villagio SantaClara 273 1989 25 ownership SCCLRT RiverOaks Elan SantaClara 941 1991 25 rental SCCLRT RiverOaks The Fountains SantaClara 226 1993 NA rental CalTrain Mt.View Park Place Mt. View 370 1989 49 rental CalTrain Mt. View VillaMariposa Mt. View 248 1985-1986 28 rental CalTrain CaliforniaAve. PaloAlto Central PaloAlto 74 1985 18 ownership CalTrain CaliforniaAve. CaliforniaPark Apartments PaloAlto 45 1989 NA rental SD Trolley Amaya Villagesof La Mesa La Mesa 384 1989 20 rental SD Trolley LaMesa La MesaVillage Plaza La Mesa 95 1991 NA ownership SD Trolley BarrioLogan Mercadodel Barrio San Diego 144 1994 NA rental SD Trolley 47th Street CreeksideVillas San Diego 144 1989 NA rental SRT Butterfield WindsorRidge Sacramento 112 1988 NA rental LA-BlueLine Pacific@5th Bellamar LongBeach 160 1990 NA rental LA-BlueLine TransitMall PacificCourt LongBeach 142 1992 NA rental Note:BART= ; SCCLRT = SantaClara County light Rail Transit;Ca/Train= Ca/Train Commuter Rail Service; SD Trolley=San DiegoTrolley; SRT = SacramentoRegional Transit; LA-Blue line = LosAngeles Metrorail Blue line light Rail Transit;NA = 1101available or not known. *Numberof dwellingunits per acre. Source:NTRAC Project Database, 1994. 112 Journalof PublicTransportation

··f ~Uit~alc

Collseum .....~

sin ~_eandro FREMON T '·,. COR RID OR

San ;;,,<

Franc i sco Hir':"ord

Bay Sou~-~~rd uri;o~.~'.'.~\ 0 F~cmom

• - IO m1lcs

10 km

Figure 2. BARTsystem map.

Fall /996 Journal of Public Tra11sportatio11 113 project, a 135-unitapartment comp lex with 19,000 square feet of ground-floor retail; 27 of the units are priced below market as set asides for low-and moderate­ incomefa milies (Figure 3). To make the projectwork, the redevelopment author­ ity becamean equity partner, leasing land to the developerfo r $1 per year and 15- 20 percent of cash flow. Del Norte Place has leased rapidly. It opened in mid- 1992a nd by mid-1993, 97 percent of its apartments were occupied. The Bay Area's best example of suburban transit-oriented developmenten ­ circles the Pleasant HiII B ART station. Between 1988a nd 1993, over 1,800 hous­ ing units and 1.5 million square feet of class A office space was built within a quarter mile of the Pleasant Hill station (Figure 4). Pleasant Hill's success is attributable to three key factors: one, the creation of specific plan in the early 1980s that served as a blueprint for targeting growthnear the rail station over the ensuing 15 years; second, the existence of a proactive redevelopment authority

Figure3. Del Norte Placemi xed-useproject at El Cerritodel Norte Station.Th e project abuts the BARTstation and is separatedfrom the aerial BARTtrack by a linear park. Photo by Robert Cervero.

Fall /996 114 Journalof PublicTra11sporlalion whose staff aggressively sought to implement the plan by assembling irregular parcels into developable parcels and issuing tax-exempt bond financing for pub­ lic and private improvements; and third, having a local elected official who be­ came the project's "political champion," working tirelessly and participatingin innumerablepublic hearingsto shepherd the project throughto implementation (Cervero, Bernick, and Gilbert 1994).C urrentplans cal l for convertingtwo BART parking lots at the Pleasant Hill station into structured replacement parking to open up land for restaurants, retail shops, and a regional cultural-entertainment complex, activities that are currently missingbut are widely viewed as vital to­ ward creating a more village-like atmosphere. Plenty of building activity can also be found around other rail stations in California. In Santa Clara County, over 2,500apartment a nd condominium units

Figure4. PleasantHill BARTstation area. Some 1 ,800 housing units and 1.5 million square feet of office and retail building spaces urround the Pleasant Hill station. Photo from BARTfil es.

Fall /996 Joumal of Public Tramportation 115 have been built near light rail stops in the past five years. A recently completed 250-unit project, Almaden Lake Village,was built on the transit district's park­ and-ride lot adjacent to the Almaden station. As part of the County's Housing Initiative Program, plans are under way to build an additional 1,700 units of moderate-density housing (at 12 to 40 du per acre) near light rail stations over the next five years. Sacramento's updated General Plan calls for using an array of development incentives at 13 light rail stations, including higher allowabled en­ sities, lower minimum parking requirements, tax increment financing, and in­ dustrial developmentb onds. The City of San Diegoh as perhapsdone the most in recentyears to embrace transit-orientedd esign concepts, adopting a fonnal policy "to direct growth into compact neighborhood patterns of development, where living and working environmentsare within walkable distances of transit sys-

Figure5. AmayaStation ar ea on the ElC ajon line. More than 300 apartment units of the Villages of La Mesaabu t San Diego Trolley's Amaya Station. Photo by Robert Cervera.

Fa/11996 I 16 Journalof PublicTransportation terns" (CityCouncil Policy 600-39). Since 1990,more than 380 modemapart­ mentunits have been built adjacent to the La Mesa-Amayalight rail station(Fig­ ure 5). Currentlyunder constructionis OtayRanch, a master-plannedcommu­ nity adjacentto the cities of San Diegoand ChulaVista, that will featurefive villageclusters, at blendeddensities of 18du per acre,and, developers hope, will be serveddirectly by an extensionof the trolleyline.

Opportunitiesand Barriers to TransitVillages MarketOpportunities Threetypes of opportunitiesare workingin favorof transitvillage develop­ mentin stateslike California.One opportunityhas beendemographic growth in populationgroups that are primecandidates for transit-orientedliving: young households,retirees, childless households, and in-migrantsfrom foreigncoun­ tries. In the San FranciscoBay Area, for instance,the shareof populationin the 25-to-34and 65-and-overage groupsincreased from 23.5 percent in 1980to 30.8 percentin 1990.These households tend to be small,and for financialand conveniencereasons, require less space and are moreinclined to live in attached housingunits. In greaterLos Angeles,30 percentof householdsin 1990con­ tainedno children;in the innersuburbs, two-thirds of householdswere childless. In addition,immigration added over 2 millionto the populationof the Los Ange­ les-AnaheimMSA and nearly600,000 to the SanFrancisco-Oakland MSA dur­ ing the 1980s(Speare 1993). Because many new arrivalsto the U.S. migrateto urbancenters and seekaffordable housing, more compact communities near rail stopsmight appeal to many. A secondtrend that favorstransit villages is the growingwillingness of transitagencies and localgovernments to leverageprivate investments near rail stations.Specifically, the abilityto assembleland-such as througheminent do­ main,condemnation, or redevelopmenttakings-and thushelp write down costs appealsto manydevelopers (Bernick 1993). For manytransit agencies, surface parkinglots surroundingstations are their biggestdevelopment asset. Parking lots representlarge tracts of pre-assembled,cleared land that are relativelycheap to buildupon. Converting park-and-ride lots to housingconstitutes de facto land

Fall /996 Journalof PublicTransportation 117 banking.One of the reasonswhy so muchurban growth has clusteredaround rail stationsin cities like Torontoand Stockholmis that localgovernments were able to acquireland over and beyondwhat was necessaryto build rail stations,allow­ ing them to leaseor sell extraland to real estatedevelopers. In the U.S.,eminent domainlaws prohibit excess land acquisitions. Reusing park-and-ride lots achieves similarresults to land banking,however. Such was the case at the Ballstonsta­ tion in Arlington,Virginia, when its statuschanged from a tenninal to an inter­ mediatestation followingthe extensionof Metrorail's OrangeLine to Vienna, Virginia.The relocationof park-and-ridespaces and a bus transferfacility to the new tenninal freed up land, helpingto triggera massiveredevelopment of the Ballstonstation area, includingthe constructionof a 28-storyoffice-residential­ retail complexabove the station. A thirdopportunity for transit village development is today'sreceptive policy and legislativeenvironment for coordinatingtransit and land use decisions.Re­ cent federal initiatives such as the 1991 national surface transportation act (ISTEA),clean air act amendments(CAAA), and EmpowennentZone/Enter­ prise Communities(EZ/EC) programs provide funding sources and a legislative contextfor promotingtransit-oriented development. ISTEA explicity calls for a close coordinationof transportationprojects and urban development.Clean air laws encouragetransit initiatives,such as transit-supportivedevelopment, as a possibletransportation control measure (TCM) in non-attainmentareas. The EZ/ EC programpromotes such neighborhoodtransportation strategies as mobility enterprisesand neighborhood intermodal travel centers. Transit villages are clearly consonantwith these legislativeinitiatives. Barriersto TransitVillages Workingagainst transit village developmentin Californiaand elsewhere loom two significantbarriers: ( 1) fiscal: factorsthat detract from the financial feasibilityof transit-orientedprojects, such as questionablemarket viability and lack of conventionalfinancing; and (2) political:land-use policies and NIMBY forcesthat impedemulti-family housing development.

Fall 1996 118 Journalof PublicTransportation

Americans'preference for low-densityliving is finnlyrooted. A 1993sur­ vey by the BuildingIndustry Association of NorthernCalifornia found that 82 percentof surveyedhouseholds preferred a single-familyhome over all housing types.It is a fundamentalrule, accordingto one northernCalifornia developer, that "as densitygoes up, the generalinterest from the consumergoes down" (Bookout1992, 15). In part beca~seof the questionablemarketability of denser housing,coupled with the softnessof today'sreal estatemarkets and the fallout from the savingsand loans crisis of the late-l 980s, banks are understandably hestitantto providepennanent financing for largelyuntested products like tran­ sit-basedhousing. The higherconstruction costs, developmentfees, and risks associatedwith higher densityhousing are also major financialobstacles. As multi-unitbuildings become taller, costs for design,construction, and liability insuranceincrease commensurately. Beyond 40 du per acre, podiumor other expensiveparking structures become necessary. Once construction goes above four stories,the more expensivesteel-frame construction, elevators, and lobby areasdrive up unit costs.While, in theory,denser housing near rail stopsshould produceless trafficthan if the samenumber of unitswere built as single-family homes,in practicedenser projects pay relativelyhigher impact fees. A seriesof recent lawsuitsholding condominium builders liable for faultyconstruction as late as 10 years after project completionhas also frightenedsome California developersaway from the high-densityhousing market. A pair of"isms"-localism andNIMBYism-stand as the biggestpolitical hurdlesto transit villagedevelopment. In California,Proposition 13, the 1978 initiativethat reduced local governments' capacities to generaterevenues through propertytaxes, is oftenblamed for promptingcommunities to be morecompeti­ tive than cooperative.Some jurisdictions keep high-density housing out through fiscal zoning-"zoning in" high tax-yieldingland uses, like office parks, and "zoningout" service-demandingactivities, notably apartments (that burdenal­ readyoverburdened schools and city services).To many,transit-based housing carrieswith it the specterof morecrowded schools and congestion,the stigmaof low-incomeprojects, and the prospectof tarnishingthe characterof an estab-

Fa/11996 Journal of Public Transportation 119 lishedneighborhood, thus loweringproperty values. NIMBY opposition to apart­ mentproposals resulted in restrictiveland-use policies and the passageof build­ ing moratoriain several neighborhoodssurrounding BART stations that were prime for more intensiveredevelopment, including Rockridge, North Berkeley, WalnutCreek, and the MissionDistrict in San Francisco.In Hunt Valley,Mary­ land,a majoremployment hub northof Baltimorethat recentlyreceived light rail services,NIMBY pressures resulted in the rezoningof prime land that was pro­ posed for some 1,500apartment units to a rural-conservationdesignation, de­ spitethe presenceoflight rail and an imbalanceof morethan threejobs for every availablehousing unit in the area.

The Marketfor TransitVillages Relativelylittle is knownabout the marketpotential oftransit villagedevel­ opment,in largepart becauselittle has been built to date, mainlydue to the bar­ riers cited above. Transit-orientedcommunities such as the celebratedLaguna Westdevelopment south of Sacramento,designed by architectPeter Calthorpe (1993),have struggledfinancially and for the mostpart incorporatemodest tran­ sit provisions.Presently, the entiretransit village movement seems caught in a "Catch-22":there are few examples,in part,because of questionablemarket fea­ sibility,and the marketpotential of transitvillages is questionablebecause there are few examples. In the absenceof good U.S. examplesof transitvillages, researchers with the NationalTransit Access Center (NTRAC) at the Universityof Californiare­ cently attemptedto dynamicallysimulate them using computer-generatedim­ ages (Cerveraand Bosselmann1994). The main objectivewas to gaugethe de­ gree to which peoplemight be willingto accepthigher densities needed to sus­ tain rail transit servicesin exchangefor more public amenities,like neighbor­ hood parks and close-byretail shops and eateries.Nine photoslideimages that simulateda "walk" throughfour neighborhoodswith differentdensity/amenity mixeswere presentedto residentsof the San FranciscoBay Area in the springof 1994.Each simulated"walk" began by showinga view out the rear and front windowsof a hypotheticalhouse located three blocks from a BARTstation, pro-

Fall 1996 120 Journalof Public Tramportafion

ceeded along two residential streets towarda neighborhood retail plaza, and ended at a nearby public square fronting a BARTst ation. As the densities of the four photosimulated neighborhoods increased from 12 to 24 to 36 to 48 du per acre, so did the acreage of public parks, number of shops, and amount of landscaping in the neighborhoods increase. These densi­ ties span the minimumn ecessary to support rail transit (12 du per acre), as estab­ lished by Pushkarev and Zupan ( 1977) as well as the upper boundary (48 du per acre) of what can be built without going to more expensive steel-framed struc­ tures with elevators, lobby space, and structured parking. Four photoslide im-

3

Figure6. Fourslide images of a computer-simulatedtransit village designed at 12 dwellingunits per acre.To e first image shows a view out of a second-story window into the rear yard of a house in the village.T he second shows a view out the front door looking down the street. Toe third shows housesa t the end of the street.To e fourth depicts a modest retail plaza that leads to the nearby rail station.

Fall /996 Journalof Public Tra11sportatio11 121 ages created for two of the simulated transit villages- the 12 du per acre one with the fewest neighborhood amenities and the 48 du per acre one with the most amenities-a re shown in Figures6 and 7. All images were generated using three­ dimensional computer modelingand animation techniques. Factors such as build­ ing style and newness, the amount of sunlight, and street widths were controlled so that only densities and amenities varied across the neighborhoods. Based on the survey responses of 170 Bay Area residents who viewed the slides, the lowest density neighborhood was the most preferred-58 percent of the respondents ranked it as the most desirable. However, far more respondents

2

Figure7. Fourslide images of a computer-simulatedtransit village designed at 48 dwellingunits per acre . The first image shows a view out of a second-story window into a courtyard. The second shows a view out the front door looking down the street. Tile third shows houses at til e end of the street (that did not exist in the lower-densityn eighborhoods).T ile fourth depictsa retail plazawith more activities.

Fa/11996 122 Journalof PublicTransportation liked the transitvillage built at 36 du per acre with nicer publicamenities than the transitvillage designed at 24 du per acrebut withfewer community services or amenities.Notably, people preferred tightly spaced two-and-a-half story row houses with modest backyardslocated near a public park and retail shops, to one-totwo-story row houseswith larger rear yards and morestreet frontage, but withno nearbypark and fewer local services. Those most receptive toward higher­ densitytransit villages were young adults with moderate incomes who currently reside in apartments.

ProfilingResidents of Transit-BasedHousing Of course,the limitationof visualsimulations, however attractive or fanci­ ful they mightbe, is that theyare nonetheless"make-believe." Many developers and lendersare unlikelyto investin transit-orientedprojects until a clear con­ sumerdemand can be demonstrated.While no true modem-daytransit villages exist in the U.S. today,there is plentyof transit-basedhousing from whichone can begin to infer the likelymarket profiles of transitvillage residents. We re­ centlysurveyed the residentsof 28 large-scalehousing projects near California rail stations(Cervera and Menotti 1994 ). Tenantstended to be youngprofession­ als, singles,and empty-nesters,with typically just one car per household.In 12 housingprojects near BART,for instance,there was an averageof 1.66people and 1.26 vehiclesper household,compared to an averageof 2.40 people and 1.64vehicles for all otherhouseholds in the samecensus tracts (Table2). More than 90 percentof transit-basedhouseholds had just one or two occupants,com­ paredto 58 percentof householdsin surroundingtracts. Fewer than 8 percentof transit-basedhouseholds had children.More than 70 percentof surveyedhouse­ holdsnear BARThad one or no vehicles,compared to 48 percentof households in the same censustracts. While tenant characteristics of transit-basedhousing were not statisticallydifferent from characteristics of surroundingcensus tracts, ba-sedon mean statisticsfrom Table2, it is clear that those choosingto live in apartmentsand condominiumsnear rail stopslive in comparativelysmall house­ holdswith relativelylow automobileownership rates.

Fall 1996 Journal of Public Transportation 123

Table2 Matched-PairComparisons of Householdand Occupant Characteristics of ltansit-BasedHousing and Surrounding Census Tracts

Transit Surrounding Based-Housing CensusTract Matched Std. Std. Pair Prob- Mea11 Dev. Mea11 Dev. t Statistic ability HouseholdCharacteristics

Persons/household 1.66 0.81 2.39 1.37 1.90 .091 No. of vehicles available 1.26 0.68 1.61 I.II 1.56 .165 OccupantCharacteristic

Age (17+ years) 36.3 14.7 42.1 17.7 1.38 .196

Note:The "Surrounding Census Tract" consists ofthe censustract that encompasses the housingproject, with the estimatedpopulation for the transit-basedhousing projects netted from census tract data.

Whatmost distinguishesresidents of housingnear Californiarail stationsis theirtendency to workdowntown and in otherlocations well servedby transit.In the case of five apartmentand condo complexesnear the Hayward and San LeandroBART stations, 43 percentof employedresidents worked in downtown San Franciscoor Oakland,compared to just 13percent of employedresidents in the surroundingcensus tracts. And an estimatedone-half of the residentsof 1,800 apartmentunits near the PleasantHill BARTstation workedin downtownSan Franciscoor Oakland,compared to a citywideaverage of just 10 percent. In a study of residentiallocation choice in greaterPhiladelphia, Voith ( 1991)found similarexamples of residentialsorting, wherein people gravitated toward loca­ tions with comparativeaccessibility advantages to job sites. Censustracts with commuterrail servicenearby had 12 percentmore of their residentsworking in downtown Philadelphia than did surrounding census tracts. Like BART, Philadelphia'srail systemradially connects suburban communities to the CBD. Buildersare startingto realizethat a numberof downtownworkers, many of whomare youngprofessionals earning good wages, are attractedto rail-based

Fall /996 Journal of Public Transportation housing.Projects with nice amenitiesand whichcater to the tastesof youngpro­ fessionalsseem to appealto manychildless households seeking condominiums and apartmentsnear rail. One exampleis the Park Regencyapartment complex near the PleasantHill BARTstation, an upmarketcomplex complete with a pool, spa/sauna,and recreationalbuilding that has a waitinglist to move in. Three­ quartersof the Park Regency'soccupants are in the 18-34year age group, and more than 50 percentearn more than $40,000annually. Another high-amenity projectis DelNorte Place near the El Cerritodel Norte BART station; its market­ ing brochuresemphasizes the project'sfireplaces, bay views, ground-floor retail, and proximityto BART.In an interviewwith TheNew YorkTimes, the project developerstated that he aggressivelyput in a bid to the El Cerritoredevelopment authorityto build on the site becausehe believesliving near rail stationswill becomeincreasingly attractive as regionaltraffic congestion worsens (McCloud 1992). With so many residentsof transit-basedhousing working downtown and other rail-serveddestinations, these projectsshould generate high rates of rail commuting.Recent surveys show that Californiansliving within a quartermile of an urbanrail systemare aroundthree times as likelyto commuteby rail transit as the averageworker living in the samecity (Cervero1994). One-third of em­ ployed residentsliving in apartmentsand condominiumsnear BARTstations commuteby rail, comparedto 8 percentof all commutersliving in the three BART-servedcounties (San Francisco,Alameda, and Contra Costa).The two most importantdeterminants of rail usageare trip destinationand availabilityof free parking.Among those living in multi-familyprojects near BARTstations and headingto San Franciscojob sites with no free parking,nearly 9 out of 10 worktrips are by BART.If theycan park free in downtownSan Francisco, around 60 percentcommute by rail. For commutesto secondaryurban centers like Oak­ land and Berkeley,around half are by BART.For all other destinations(where oftenworkers park free),on averageonly 6 percentof commutetrips by station­ area residentsare by rail. Clearly,clustering housing around rail stops will do little good if, as duringmuch of the 1980s,job growthoccurs mainly along sub­ urban freewaycorridors. Both ends of worktrips-housing andjob sites-must

Fall /996 Journal of Public Transportation 125 be within reasonableproximity of stationsif clusteredgrowth is to pay signifi­ canttransportation and environmental dividends-in short,more mixed-use transit villagedevelopment is necessary. Becauserail-based households own relativelyfew cars and frequentlypa­ tronizetransit, zoning standards should be relaxedto allowjust one parkingspace per unit at complexesnear rail stations.This wouldlower construction costs by an estimated$12,000 per unit in the Bay Area (the typicalcost of a tuck-under, podiumparking space), and also createa morepedestrian-oriented environment. Tenantswith more than one car might be given the option of leasing a second space. Another novel idea suggestedby Holtzclaw( 1994) would have banks grant those living in rail-basedcondominiums an "efficient-location"loan for homepurchases. If rail-basedhousing lowers transportation costs (mainlyin the formof onlyhaving to ownone car),then these savings might be subtractedfrom principal, interest, taxes, and insuranceexpenses when calculatingmortgage qualifications.This acknowledgesthat lower transportationcosts frees more moneyfor housingconsumption. Such loan adjustments could further attract pro­ spectivehomebuyers to transitvillage locations.

Rail-BasedHousing and Rents If rail-basedhousing projects are becoming increasingly desirable addresses, this shouldbe reflectedin rent levels.Comparisons were recently made between 1994rents at multi-unitprojects within a quartermile of the PleasantHill BART stationversus otherwise similar projects in PleasantHill and the nearbycities of WalnutCreek and Concordthat were beyondwalking distance of a rail stop (Bernick,Cervero, and Menotti 1994). Rentsper square foot for one bedroom/ one bathroomunits near the PleasantHill stationwere $1.20, comparedto an averageof $1.09 for similarprojects (in termsof size, age, and amenities)in the same geographicsubmarket but awayfrom BART. Two bedroom/two bathroom unitsnear the PleasantHill stationsleased for around$1.09 per squarefoot com­ pared to around $0.94 per square foot for comparableunits away from BART. These findingstranslate into a 10 to 15 percentrent premiumassociated with being near BART.

Fall /996 126 Journalof PublicTransportation

A hedonicprice modelwas also estimatedfor measuringthe rent premium commandedby rail-basedhousing. Using multiple regression analysis, a hedonic pricemodel does what matched-pair comparisons are unableto: statisticallycon­ trol for a largenumber of attributesof the "housingbundle," allowing the unique effectsof each attribute(including proximity to BART)to be parcelledout. Table3 presentsthe findingsof the estimatedmodel. Units within a quarter­ mile of the PleasantHill BARTstation rented for around$34 more per month than otherwisecomparable units fartheraway from BART,controlling for the influenceof unit size, amenities,and otherfactors. More bathrooms, bedrooms, andamenities like playgrounds and weight rooms likewise increase monthly rents. Table3 also revealsthat unitsin morecompact projects rent for morethan com­ parableunits in lower-densityones. Project density, it shouldbe noted,reflects units per acre within a complexas opposedto the densityof the surrounding neighborhood.The rentalpremium associated with compactprojects could re­ flectthe benefitsof tenantsbeing closer to pools,playgrounds, and otherameni­ ties, as well as livingin a communalsetting. The rail-based projects used in this analysis,moreover, were comparativelydense, suggesting some interactionbe­ tweenthese two factors-closenessto stationsand projectdensity. The finding that bothproximity to transitand project compactness get capitalizedinto higher rents bodeswell for the futureof transitvillage development in the Bay Area.

Stimulatingthe Marketfor TransitVillages Perhapsthe mostpromising recent development in California'stransit vil­ lage movementwas the passageof the TransitVillage Act, AB 3152. The Act stipulatesthat no public works projects,tentative subdivision maps, or parcel maps may be approved,nor zoningordinances adopted or amended,within an area coveredby a transit villageplan unlessthe map, project,or ordinanceis consistentwith the adoptedtransit village plan. This was a small but important step toward bringingthe transitvillage idea to fruition.The bill, as originally drafted,would have allowedmunicipalities to designatea "transitvillage dis­ trict," similarto a redevelopmentdistrict, with specialland assemblageand tax incrementfinancing privileges. The originalbill also stipulatedthat developers

Fall /996 Journalof PublicTransportation 127

Table3 HedonicPrice Model for Multi-FamilyRental Units in the PleasantHill StationArea and Surrounding Submarket, 1994 Dependentvariable = rent per month, in dollars Variable Coefficient T-statistic Sig11ijica11ce BARTstation withinone-quarter mile ( I=yes, O=no) 34.101 1.526 .133 Size of unit (sq.ft.) .427 6.497 .000 No. of bedrooms 29.488 1.497 .141 No. of bathrooms 42.039 2.657 .Oil Playgroundon site ( I=yes, O=no) 30.461 1.689 .097 Weightroom on site ( I =yes, O=no) 66.544 4.721 .000 Projectdensity (units/acre) .397 1.380 .174 Projectage (in years,from 1991) -10.971 -6.200 .000 Project in Concord(I =yes, O=no) -129.842 -8.878 .000 Proportionof total units in project of unit type -44.545 -1.567 .124 Laundryroom on site ( I=yes, O=no) -21.221 -1.105 .275 SummaryStatistics: Numberof observations 60 R-Squared .919 F statistic 49.331 SignificanceF .000 buildingwithin the district be granteddensity bonuses of at least 50 percent. Becauseof stiff oppositionfrom fiscal conservatives,most of these provisions were later strippedfrom the bill. Regardless,the Act gave newfoundlegitimacy to California'stransit village movement. As passed,AB 3152 is a voluntarystatute encouraging cities and counties to plan more intensivedevelopment around rail stations,though it providesfew fiscal powers or special authorityto do so. Sponsorshope the bill will be ex­ panded in comingyears to providemore financialincentives, perhaps granting transitvillage districtspriority access to discretionarystate funds, such as from the nationaltransportation act (ISTEA)and fuel price rebate programs. California's transit village movementsuffered a recent setback,however, when Governor Wilsonvetoed an AssemblyBill (AB 1338) in the spring of 1995 that would haveestablished local revolving funds (from state and federaltransportation plan-

Fall /996 128 Journalof PublicTransportation ningmonies) and providedloans to citiesand countiesto enablethem to prepare specifictransit village plans. The veto, most observersagree, had more to do withthe generallyconservative fiscal mood of the timesthan an oppositionto the principleof transit-orienteddevelopment. Still, the veto underscoresthe reality that transit-orienteddevelopment is not highon the prioritylists of manypoliti­ cians, and that transit villagesface an uphillstruggle in becominga realityin stateslike California. Notwithstandingsuch political setbacks, it is encouragingthat some hous­ ing projectsnear rail stationsare leasingquickly, commanding rent premiums, and attractingresidents who patronizetransit. Local governments can leverage transit-orienteddevelopment by emulatingwhat was done in PleasantHill and El Cerrito-namely,by creatingspecific plans to guidedevelopment and usingtax incrementfinancing and other tools to assistwith land assemblageand absorb some of the risks of projectdevelopment. Given some of the doubts over the marketabilityof higherdensity housing and today'sconservative lending prac­ tices,some degree of risk-sharingbetween the publicand privatesectors will be necessaryif transitvillages are ever to take form.Relaxing zoning standards to allow fewer parkingspaces at rail-basedprojects and rewardingthose buying condominiumsnear rail stopswith "efficient-location" loans would further pro­ mote transit-orientedgrowth. Together, strong market interest,public-private cooperation,and a conducivepublic policy environment would prove a powerful combinationin takingthe transitvillage movement from idea to implementation.•:•

Acknowledgements This researchwas supportedby a grantfrom the CaliforniaDepartment of Trans­ portationand the Universityof CaliforniaTransportation Center. Many people contrib­ uted to the work reportedin this paper,including Michael Bernick, Peter Bosselmann, ValMenotti, and Jill Gilbert.The viewsexpressed are thoseof the authorand not neces­ sarilythose of the sponsors.

Fall 1996 Journalof PublicTransportation 129

References Bernick,M. 1993.The Bay Area's EmergingTransit-Based Housing. Urban Land Vol. 52, No. 7: 38-41. Bernick,M., R. Cervero,and V. Menotti.1994. Comparison of Rents at Transit-Based HousingProjects in NorthernCalifornia. Berkeley: Institute of Urbanand Regional Development,Working Paper 624. Bookout,L. 1992.The Futureof Higher-DensityHousing. Urban Land Vol.51, No. 9: 14-18. BuildingIndustry Association of NorthernCalifornia. 1993. Survey of HousingPrefer­ encesin NorthernCalifornia. San Francisco: Building Industry Association ofNorth­ ern California,unpublished report. Calthrope,P. 1993. TheNext AmericanMetropolis. Princeton: Princeton Architectural Press. Cervero,R. 1994.Transit-Based Housing in California:Evidence on RidershipImpacts. TransportPolicy, Vol. I, No. 3: 174-183. Cervero,R. 1995.Sustainable New Towns: Stockholm's Rail-Served Satellites. Cities 12 1: 41-52. Cervero,R., M. Bernick,and G. Gilbert. 1994.Market Opportunities and Barriersto Transit-BasedDevelopment in California.Berkeley: Institute of Urbanand Regional Development,Working Paper 621. Cervero,R., and P.Bosselmann. 1994. An Evaluationof theMarket Potential/or Transit­ OrientedDevelopment Using Visual Simulation Techniques. Berkeley: Institute of Urbanand RegionalDevelopment, Monograph 47. Cervero,R., and V. Menotti. 1994.Market Profiles of Rail-BasedHousing Projects in California.Berkeley: Institute of Urbanand RegionalDevelopment, Working Paper 622. Cervero,R., and J. Landis.1995. Development Impacts of UrbanTransport: A U.S. Per­ spective.In Transportand UrbanDevelopment, D. Banister,ed. London:Chapmand and Hall. CommuterTransportation Services. 1993. The State of the Commute.Los Angeles: Com­ muterTransportation Services, Inc., unpublished report. Downs,A. 1994.New Visions/orMetropolitan America. Washington, D.C.: The Brookings Institution.

Fall /996 130 Journalof PublicTransportation

Holtzclaw,J. 1994.Residential Patterns and Transit,Auto Dependence, and Costs.San Francisco:Resources Defense Council. Larson,J. 1993.Densi_ty is Destiny.American Demographics Vol. 15, No. 2: 38-42. McCloud,J. 1992.High-Density Housing Near San Francisco: Builder Betting Proxim­ ity to CommuterLine. New YorkTimes Vol. 141, Section 1, July5: 23. Pushkarev,B., and J. Zupan.1977. Public Transportation and Land Use.Bloomington: IndianaUniversity Press. Speare,A. 1993.Changes in UrbanGrowth Patterns 1980-90. Cambridge, Massachu­ setts:Lincoln Institute of LandPolicy, Working Paper. Voith,R. 1991.Transportation, Sorting and HouseValues. American Real Estate and UrbanEconomics Association Journal Vol. 19, No. 2: 117-137.

Aboutthe Author ROBERT CERVERO is Professorin the Departmentof Cityand RegionalPlanning at the Universityof California,Berkeley.

Fall /996