Urban Studies, Vol. 38, No. 11, 1973–1987, 2001 Historic Preservation and Residential Property Values: An Analysis of Texas Cities Robin M. Leichenko, N. Edward Coulson and David Listokin [Paper(cid:142)rstreceived, November 1999; in(cid:142)nalform,January 2000] Summary. Designation of historic districts is increasingly used as a tool to revive or halt the deterioration of central-city neighbourhoods. While historic designation is generally thought to have a positive impact on property values, evidence on this issue is mixed. One limitation of previousresearchisthatittypicallyfocusesonhistoricneighbourhoods inonecityandthusbases its conclusions on a very limited sample. This study expands upon previous work by examining the effects of designation on property values across a larger set of cities. The study employs hedonic regression models to estimate housing prices in historic districts and comparable neighbourhoods in nine Texas cities. Results suggest that, in most cases, historic designation is associated with higher property values. 1. Introduction Historic designation has become an import- designate properties as historically ant tool in efforts to preserve central-city signi(cid:142)cant. Although establishment of many neighbourhoods and to promote urban econ- local historic districts preceded NHPA—for omic development (Listokin et al., 1998; example, Charleston, South Carolina, estab- Slaughter, 1997; Rypkema, 1995; Wojno, lished historic district zoning in 1931 (Lock- 1991). Designation of historic districts has hard and Hinds, 1983)—the rate of been employed on a broad basis in the US establishment of local registers dramatically since the 1960s, following legal decisions accelerated after 1966 (Listokin, 1986). In thatupheld landmarking andpassage in 1966 1966, there were approximately 100 local of the National Historic Preservation Act historic district commissions in the US. (NHPA) (Listokin, 1986). The act gave the Presently, there are more than 2000 such Secretary of the Interior the authority to commissions (Listokin et al., 1998). maintain a National Register of Historic One of the main justi(cid:142)cations for desig- Places, comprising districts, sites, buildings nation of a historic district within a city is and objects of local, state or national historic that it provides a means to protect a historic signi(cid:142)cance (Wojno, 1991, p.297). In ad- neighbourhood from physical deterioration. dition, many municipalities have established With regard to property values, however, local historic registers that allow local gov- designation of a historic district may be ernments to establish historic districts and to either value-enhancing or value-detracting. RobinM.LeichenkoandDavidListokinareintheCenterforUrbanPolicyResearch,Rutgers,TheStateUniversityofNewJersey, 33 Livingston Avenue, Suite 400, New Brunswick, NJ 08901, USA.Fax: 732 932 2363. E-mail: [email protected] and listokin@rci.rutgers.edu.N.EdwardCoulsonisintheDepartmentofEconomics,PennStateUniversity,604KernBuilding,University Park,PA16802-2206, USA.Fax:8148634775.E-mail:[email protected]. 0042-0980 Print/1360-063X On-line/01/111973-15 Ó2001TheEditorsofUrbanStudies DOI:10.1080/00420980120080880 1974 ROBINM.LEICHENKOETAL. Historic designation is thought to have a fac¸ade treatments over and above those re- positive impact on property values by pro- quired in the jurisdiction’s general mainte- viding a form of insurance of future neigh- nance code. For example, in the city of bourhood quality. The prestige of of(cid:142)cial Abilene, owners of designated properties landmark designation in conjunction with as- must apply for a ‘certi(cid:142)cate of appropriate- surance that its desirable historic amenities ness’ (C of A) prior to performing any type will be fostered into the future by public of work on the property’s exterior (Coulson regulation, may make property-owners in and Leichenko, 2001). A ‘C of A’ is, in fact, historic districts more willing to invest in a requirement in many of the 2000 or so rehabilitation and maintenance of their communities with local landmarking. Fur- properties. One study of New York City, for thermore, maintenance work on the historic example, concluded that historic district des- property is often more expensive than it ignation, by fostering neighbourhood pride might otherwise bebecause ithas to conform and other attributes, “serves to strengthen to fairly rigorous guidelines (for example, both property values and social fabric”(New only certain types of paint may be allowed). York Landmarks Conservancy, 1977, p.2). Theselandmarkrestrictions anddemandscan In addition to direct effects on property exert a downward pressure on prices. values in a district, historic designation is In addition to control over a property’s also thought to have positive spillovers for appearance, designation may also detract neighbouring areas,wherebydesignation of a from a property’s value by prohibiting the district leads to a ripple effect of rehabili- conversion to other uses or to a more inten- tation and upgrading of properties in sur- sive use. This type of argument would sug- rounding neighbourhoods (Listokin et al., gest that, in some instances, designation of 1998; Rypkema, 1994; Coulson and Le- historic districtsmightnotre(cid:143)ectthe‘highest ichenko, 2001). Thus, historic designation is andbest’use of land—i.e.themostpro(cid:142)table seen as more than just a way to preserve use incorporating those uses that are legally historic buildings; it is also increasingly re- permissible, physically possible and gardedasboth acommunitypreservationand (cid:142)nancially or economically feasible (Kin- an economic development strategy. A recent nard, 1971, p.39). article noted that economics and revitalisa- The practice of historic designation also tion have taken their rightful places as the raises a number of broader legal and equity- pillars upon which the preservation ethic is related issues. These issues have been based (Rypkema, 1995). A prime exampleof addressed in literature on preservation and the growing recognition of the linkages be- property rights and on urban renewal tween preservation and local development and gentri(cid:142)cation (see, for example, Smith can be seen in the Community Partners Pro- and Williams, 1986; Smith, 1996; Schuler et gram, a new initiative of the National Trust al., 1992) and therefore will be only brie(cid:143)y for Historic Preservation, which aims to described. With regardto the legal aspects of demonstrate the “effectiveness of preser- designation, debate continues to surround the vation-based community development” (Na- issue of whether designation is a ‘taking of tional Trust for Historic Preservation, 1998, property’. The courts have overwhelmingly p.1). decided that designation is not a ‘taking’ but There are, however, a number of potential rather a police power regulation that value-detracting aspects of historic desig- justi(cid:142)ably furthers the public’s health, safety nation. Designation of a historic district may and welfare while recognising the rights of impose restrictions on alterations and demo- private property-owners (see, for example, lition (or it may at least require administrat- Penn Central Transportation Company v. ive review and/or some delay of such New York City, 438 vs. 104 [1978]); yet actions) and it may require maintenance of designation’s property value impact contin- exterior ornamentation and other historic uestobediscussed (asdoes the moregeneral HISTORICPRESERVATIONANDPROPERTYVALUES 1975 issue of public land-use regulations) in both values (for example, Scribner, 1976; legal and non-legal forums (Duerksen, 1983; Rackham, 1977; US Advisory Panel on His- Rypkema, 1994; Miller, 1998). toric Preservation, 1979). Other difference- Other issues raised include equity consid- in-difference studies found, however, that erations. For example, how should the bur- designation has a neutral or negative effect den of a public good—in this instance, on property values (for example, Heudorfer, preservation—be borne and shared between 1975; New York Landmarks Conservancy, the affected private property-owner and the 1977; Samuels, 1981; Gale, 1991). public at large? Another equity issue is the One important limitation of the above possibility of displacement of low-income studies of historic designation and property residents who can no longer afford to live in values is that they rely solely on comparing historic neighbourhoods (Smith, 1998). Ac- sample averages of the growth rate in prop- cording to this argument, higher property erty values in historic areas with those in values as the result of historic designation non-historic areas. Typically, no other vari- lead to increased rental prices and higher ables (for example, property characteristics) property taxes, and these, in turn, may dis- are controlled and, to the extent that there place low- to moderate-income residents may be variables independent of designation (Wojno, 1991). Although designation of his- that explain the changes in property values, toric districts cannot be equated with urban theresults maybebiased and inconsistent. In redevelopment and gentri(cid:142)cation, which an effort to rectify the above limitations, have been associated in many cases with the mostof the more recent studies of the effects attraction of higher-income residents and in- of historic designation employ hedonic re- creased housing prices, the potential for dis- gression models. This method of analysis placement of low- to moderate-income provides a means to assess the implicit value residents continues to be an important con- of the structural characteristics of a house.2 sideration. For this reason, the potential Use of a hedonic approach enables assess- bene(cid:142)ts of designation in terms of higher ment of the effect of historic designation on property values and increased tax revenues housing values while holding constant prop- must be weighed against the possibility of erty and neighbourhood characteristics. displacement of lower-incomerenters,partic- A number of studies employing hedonic ularly in cities with very limited low-income methods have concluded that designated his- housing supplies. toric properties and properties located within historic districts typically sell for a premium when compared with similar, non-designated properties (for example, Ford, 1989; Asabere 2. Empirical Studies of Historic Desig- and Huffman, 1994a; Clark and Herrin, nation and Property Values 1997; Coulson and Leichenko, 2001). Other The question of the effects of historic desig- hedonic studies, however, have found mixed nation on property values has been explored or negative results (for example, Schaeffer in the empirical literature for more than 20 and Millerick, 1991; Asabere and Huffman, years (Table 1). Many studies employ a dif- 1994b; and Asabereetal.,1994). Inaccount- ference-in-difference methodology whereby ing for their mixed results, Schaeffer and the changes in property values of houses Millerick (1991) note that the effect of his- within a district and houses outside a district toric designation on price may depend upon are compared.1 If prices increase (decrease) whether a property is locally or nationally more within the designated district, then des- designated. Their study found a positive im- ignation is inferred to have a signi(cid:142)cant and pact on values with national designation but positive (negative) effect. A number of dif- a negative impact with local designation. ference-in-difference studies have found that This difference, according to the authors, designation has a positive effect on property resulted from more stringent controls in the 1976 ROBINM.LEICHENKOETAL. onues Impactofdesignationpropertyval NeutralPositivePositiveNeutralPositive NeutralPositiveNeutralMixedPositiveNegativeNegativePositivePositive eeeee e e ccccc c c nnnnn n n eeeee e e rrrrr r r eeeee e e fffff f f od -dif-dif-dif-dif-dif -difnic-difnicnicnicnicnicnic h nnnnn nonoooooo Met ce-oce-oce-oce-oce-o ce-oHedce-oHedHedHedHedHedHed nnnnn n n eeeee e e rrrrr r r eeeee e e dies DiffDiffDiffDiffDiff Diff Diff u st mpirical n,TX;WA Resultsofpreviouse Location NewYorkCityAlexander,VAWashington,DCNewYorkCityandria,VA;Galvestovannah,GA;Seattle,Washington,DCBaltimore,MDWashington,DCChicago,ILPhiladephia,PAPhiladephia,PAPhiladephia,PASacramento,CAAbilene,TX xa 1. eS e Al bl a T ) 7 7 9 1 ( y c n a ConservHistoric k(1991)(1994a)(1994b) 7)o(2001) Study Heudorfer(1975)Scribner(1976)Rackham(1977)NewYorkLandmarksUSAdvisoryPanelonPreservation(1979)Samuels(1981)Ford(1989)Gale(1991)SchaefferandMillericAsabereandHuffmanAsabereandHuffmanetal.Asabere(1994)ClarkandHerrin(199CoulsonandLeichenk HISTORICPRESERVATIONANDPROPERTYVALUES 1977 local districts and from the prestige associ- included similarity in general characteristics ated with location in a national district. of thehousing (forexample, ageof thebuild- Overall, the more recent, hedonic studies ings, size and architectural style), similarity represent an important improvement over the in income levels and similarity of demo- earlier difference-in-difference studies. How- graphic characteristics. City planners and/or ever, one limitation of the multivariate stud- historic preservation of(cid:142)cers selected the ies—and one possible reason for their mixed comparable neighbourhoods in each city. (cid:142)ndings—is that they typically look at a In the cities of Abilene and Nacogdoches, small number of historic neighbourhoods in historic properties are designated individu- one city and thus base their conclusions on a ally; the cities do not have designated his- relatively limited sample within a single toric districts. Comparable properties in each housing market. This study expands upon city were selected based on location in the previous hedonic studies by examining the same neighbourhood or in neighbourhoods effects of historic designation on residential similar to those where the designated houses property values across a larger sample of were located. In Fort Worth, historic proper- cities. The advantage of our approach is that ties included properties located within his- we employ a roughly common econometric toric districts as well as a large number of framework across the different models (al- properties(93)with individual historic desig- though there are some differences in the nation that were not located in a historic various city-models) and this facilitates com- district. In order to take into account both parison across a large pool of cities—a com- types of historic properties, the Fort Worth parison which is not otherwise available analysis used property value data for the given the disparate models that previous re- entire city. Designated properties were com- search has provided. Nine Texas cities— pared with all other residential properties in Abilene, Dallas, Fort Worth, Grapevine, the city. Laredo, Lubbock, Nacogdoches, San Anto- nioandSanMarcosareincluded inthehedo- 3.2 Type of Historic Designation nic analysis. In several of the cities, we were able to distinguish betweendifferenttypes of histori- 3. Data cal designation. In the cities of Abilene and San Marcos, we were able to differentiate 3.1 Selection of Historic and Comparable between nationally and locally designated Properties historic properties or historic districts, while Prior to the estimation of the hedonic mod- in the city of Lubbock, we were able to els, it was necessary to select historic and differentiatebetween national, State of Texas comparable properties for inclusion in the and local historic designation. National des- analysis of each city. A complete list of ignation means that a property or district is designated historic properties was obtained included on the National Register of Historic from city-planning and/or historic preser- Places. State of Texas designation is a his- vation of(cid:142)cials in each city.3 In six of the toric designation category that has been cities (Dallas, Grapevine, Lubbock, Laredo, granted at the state level. Local historic des- San Antonio and San Marcos), all of the ignation may include designation of a local historic properties included in the analysis historic district, designation of individual are located within designated historic dis- properties as historically signi(cid:142)cant, or in- tricts. In these cases, residential properties clusion on special listings of historic local within the historic neighbourhoods were properties. compared with properties located in compar- Because national or state designation able neighbourhoods in the city. Criteria for seems likely to convey more prestige to an the selection of comparable neighbourhoods individual property or historic district and 1978 ROBINM.LEICHENKOETAL. Table 2. Datasources Numberof Average historic properties City DataSource Sample size property value in thesample Abilene Appraisal 7620 $39160 222 Dallas Appraisal 4920 $64838 2200 Fort Worth Appraisal 102948 $54519 1338 Grapevine Appraisal 59 $44673 27 Laredo Appraisal 338 $45396 177 Lubbock Appraisal 1922 $30471 440 Nacogdoches MLS 30 $93130 15 San Antonio Appraisal 3806 $47970 1912 San Marcos MLS 80 $94920 34 may therefore make the property or district Antonio. Appraisal district data were se- more desirable, we expect that, all other lected as our primary data source because things being equal, nationally or state-desig- these dataare comprehensive, covering allof nated properties will have higher values than the historic properties in anentireneighbour- will properties that carry only local desig- hood and all properties in comparable neigh- nation. Inaddition to conveying greater pres- bourhoods. While appraisal data have been tige than that conveyed by local designation, used in other recent studies of the property national and state designations are typically value impacts of historic preservation (see, less restrictive (Schaeffer and Millerick, for example, Gale, 1991; Coulson and Le- 1991). If there is no federal or state funding ichenko, 2001), potential limitations of ap- orother involvement (for example,federalor praisal data include possible in(cid:143)ation or state rehabilitation grants or licenses), then reduction of housing values by appraisers the owner of a federal or state landmark can, due to historic status. In each city where by and large, make alterations without his- appraisal data were used, we enlisted the aid toric ‘C of A’ approval. Inthe samevein, the of city planners in compilation of the data- owner can demolish the federal/state land- sets in order to ensure that the historic and mark and replace it with a ‘highest and best comparison properties (neighbourhoods) in- use’ structure. It isonly with local landmark- cluded in the sample had been recently ap- ing that signi(cid:142)cant restrictions on alterations praised based on a consistent method. and demolishing are sometimes triggered. In two cases, San Marcos and Nacog- These differences should further contribute doches, where appraisal data were not avail- to the more pronounced value-enhancing ef- able or were not consistent, property values fectof national orstatedesignation. Wewere were obtained from Real Estate Multiple able to test this hypothesis in Abilene, Lub- Listing services. Data from Real Estate Mul- bock and San Marcos. tiple Listings, which include the actual price at which a property sold, provide anaccurate re(cid:143)ection of the marketvalue of a home. The keyproblem with these data,however, isthat 3.3 Data Sources the sample sizes tend to be smaller because For the majority of the cities, data on resi- the data are based on actual sales. In the city dential property values were obtained from of Nacogdoches, for example, there were county appraisal district databases (Table 2). only 15 sales of designated historic proper- These cities include Abilene, Dallas, Fort ties during the study period. Smaller sample Worth, Grapevine, Laredo, Lubbock and San sizes limit the accuracy and reliability of the HISTORICPRESERVATIONANDPROPERTYVALUES 1979 hedonic analysis of the impact of historic explanatory variables are speci(cid:142)ed in linear designation.4 units (for example, bath is simply the num- ber of bathrooms in the house). With the semi-logarithmic form, the coef(cid:142)cient on 4. Modelling Approach each explanatory variable (square footage, The property value analysis involved the ap- number of baths, etc.) is interpreted as the plication of multivariateregression models to percentage change in thehouse’spricethatis assess the impact of historic designation on associated with a one-unit increase in the residential property values. The model form explanatory variable. For example, a used in the study involves estimation of coef(cid:142)cient of 0.07 on the variable bath im- house price as a function of property charac- plies that the addition of one bathroom is teristics, neighbourhood location and historic associated with an increase in house price of status. Since we are primarily interested in approximately 7 per cent. determining whether historic status exerts a As is typical in hedonic studies of this statistically signi(cid:142)cant effect on housing type, it is important to control for covariates price, and whether this effect is positive or of historical designation inour speci(cid:142)cations, negative, the key variable of interest is his- as this variable can be correlated to some toric status. degree with other attributes. To address this The basic form of the hedonic model is as issue, we examined bivariate correlations be- follows tween designation and the other housing at- tributes in each sample. Designation is lnPrice5 f(structural characteristics, obviously correlated with the year built in neighbourhood, historic) (1) each case, but in a number of our samples it where, lnPrice isthe natural logarithm of the is also (positively) correlated with land or assessed total value (or sale price) of the interior area at least as strongly as it is with house; structural characteristics of the house yearbuilt.Henceinclusion of these andother include variables such as square footage, attributes is appropriate, as omission of them year built, number of bathrooms, number of would bias upwards our measurement of the bedrooms; neighbourhood indicates the pricedifferencebetweendesignated andnon- neighbourhood in whichthehouse islocated; designated properties. and historic indicates whether or not the house is individually designated as historic 5. Empirical Results or is located in a historic district. De(cid:142)nitions of all of the variables used in Detailed results of the hedonic models for the analyses are presented in Table 3. To each city are presented in Table 4. Interpret- ensure as much comparability as possible ation of the individual estimated values in across the cities, each model started with a each city model may be illustrated through similar set of basic explanatory variables, the example of Abilene. For houses in the such as square footage, year built and his- Abilene area, other things being equal, an toric status. For most of the cities, we were increase in size of 1 square foot is associated also able to add additional explanatory vari- with an increase in property value of 0.059 ables such as number of garage spaces or per cent; based on the average house value presence of central air-conditioning.5 Several ($39160), each additional square foot in- models (Abilene, Lubbock and San Marcos) creases house value by $23. Similarly, an include variables designating type of historic increase of 1 square foot in land area is district,6 and the larger city models include associated with an increase in property value variables designating neighbourhood type.7 of 0.0091 per cent, implying that each addi- The hedonic models are speci(cid:142)ed in semi- tional square foot of land area increases logarithmic form, meaning that the house property value by $0.36. An additional bath- price is speci(cid:142)ed as the natural log and the room adds 16 per cent to the value of the 1980 ROBINM.LEICHENKOETAL. Table 3. Variable de(cid:142)nitions Variable name Variable de(cid:142)nition Housing characteristics Bath Numberof bathrooms (full and half) Fullbath Numberoffull bathrooms Halfbath Numberof half bathrooms Yearbuilt Yearthe housewasbuilt Squarefoot Squarefootage of the house Lotsize Squarefootage of the house’slot Bedroom Numberof bedrooms Heatac Presence of central heating and central air-conditioning Numstory Numberof storeys Numporch Numberof porches Garagesp Numberof garage spaces Structure Numberof buildings onthe property Condition Condition of the house Depreciation Depreciation of the house (alternative indicator of housing condition) Yearsold Yearinwhichthe housewassold Historic designation Historic Located inahistoric district and/ordesignated asa historic home National Located inanationally designated district oronthe National Register Texas Designated asa Texas historic property Noncontrib Located inahistoric district butnotcontributing to the district (Lubbock) Neighborhoodcontrolsa Abilene Censustrack Censustrack inwhich the property is located (13tracks intotal) Dallas RosemontCrest–Sunset Hills Historic District location–comparison area Winnetka Heights–South Winnetka Historic District location–comparison area TenthStreet–Bottoms Historic District location–comparison area Munger Place–Junius Heights Historic District location–comparison area Queen City–Charles Rice Historic District location–comparison area SouthBlvd/ParkRw–comparison area Historic District location–comparison area Colonial Hill–Saint Phillips Historic District location–comparison area Kessler Park–EastKessler Historic District location–comparison area Miller-Stemmons–Kidd Springs Historic District location–comparison area KingsHighway–Dallas Land and Loan Historic District location–comparison area LakeCliff–SouthLakeCliff Historic District location–comparison area Peak’sSuburban–Mill Creek Historic District location–comparison area FortWorth Elizabeth Located inElizabeth Ave.Historic District Grand Located inthe Grand Ave.Historic District Fairmont Located inthe Fairmont Historic District Isolated Historically designated property, butis notadistrict School District Schooldistrict inwhichthe property is located (12districts in total) SanAntonio Dignowity Hill–comparison Historic District location–comparison area KingWilliam–comparison Historic District location–comparison area Monticello Park–comparison Historic District location–comparison area Note: A selection of these variables were included inthe individual models for eachcity. aNotallcities required neighbourhood variables. HISTORICPRESERVATIONANDPROPERTYVALUES 1981 house, an increase of $6268. On average, square footage. Given the control for square houses with central heating and air-con- footage, the negative sign on number of ditioning have values that are 45 per cent storeys in Abilene simply implies that a greater($17628)than similarhouses without 2500-square-foot ranch-style house would this amenity. (While it seems unlikely that have a higher value than a 2500-square-foot central heating and air-conditioning alone 2-storey house. Similarly, in the Nacog- would have such a large effect on housing doches andSanAntonio models,thenegative values, the presence of central heating and sign on bedroom tells us that a 2500-square- air-conditioning is likely to be associated foot house with 2 (large) bedrooms is worth with other amenities that raise the value of a more than a 2500-square-foot house with 3 house—for example, higher-quality roo(cid:142)ng, (small)bedrooms. Theindividual coef(cid:142)cients carpeting and so forth.) With regard to year for theneighbourhood controls(not reported) built, more recently constructed houses have were generally found to be statistically higher values; each additional year of age signi(cid:142)cant.10 decreases the house’s value by 1.4 per cent. Concerning the interpretation of the All of the above estimates are statistically coef(cid:142)cients on historic designation, we again signi(cid:142)cant at standard levels of con(cid:142)dence use anillustration fromAbilene. The historic and all of the coef(cid:142)cients are of magnitudes coef(cid:142)cient of 0.19 (Table 4) suggests that similar to those found in other studies of this values for designated historic houses are type. approximately 19 per cent higher than for The housing characteristic coef(cid:142)cients in similar, non-designated properties. The the other city models may be interpreted in a coef(cid:142)cient on national indicates that nation- similar fashion. In general, the housing ally designated historic properties sell for characteristic variables tend to have the ex- approximately 5 per cent more than locally pected signs and are generally statistically designated historic properties. However, the signi(cid:142)cant. Basic characteristics, including effect of national designation is not statisti- numbers of bathrooms, square footage and cally signi(cid:142)cant; we therefore cannot state lot size generally have the expected, positive that national designation has a positive im- signs8 and are statistically signi(cid:142)cant in al- pact above and beyond that of local desig- most all cases. A positive coef(cid:142)cient on nation within the city. yearbuilt indicates that older houses gener- In general, the results indicate that historic allyhave lowervalues than donewer houses. designation has a positive effect on property Although the sign pattern on the yearbuilt values in all of the cities. The positive effect variables is generally as expected, the of historic preservation is statistically coef(cid:142)cients are not statistically signi(cid:142)cant in signi(cid:142)cant in Abilene, Dallas, Fort Worth, all cases.9 Grapevine, Lubbock, Nacogdoches and San Most of the additional structural variables, Antonio. The effect of historic preservation including presence of central heating and air is negative in San Marcos, but it is not conditioning (Abilene, Fort Worth, statistically signi(cid:142)cant. The (positive) effect Grapevine), number of garage spaces (Fort of historic preservation is also not signi(cid:142)cant Worth, Grapevine), number of porches in Laredo. Among those cities where historic (Laredo) and number of structures on the designation has a statistically signi(cid:142)cant ef- property(FortWorth, SanAntonio), have the fect on property values, historic designation expected(positive) sign and most arestatisti- is associated with average property value cally signi(cid:142)cant. While the negative effects increases ranging between approximately 5 of number of storeys (Abilene) and number per cent and 20 per cent of the total property of bedrooms (Nacogdoches and San Anto- value. Inpercentage terms,thesmallestaver- nio) seem to be counterintuitive, the reason age increases in property values occur in for these negative results becomes clear if Dallas, where the value of historic properties one keeps in mind that we are controlling for is 4.9 per cent higher than the value of 1982 ROBINM.LEICHENKOETAL. * SanMarcos 8.43E-4(0.013)——0.00251(1.47)5.95E-4(8.92)*—— ————0.156(3.24)**—0.0639(4.14)** 0.124(1.11)20.222(1.80)*—— — 0.70080 2 SanAntonio —0.0714(5.64)**0.109(4.85)**—3.79E-4(30.4)**3.31E-5(18.1)**0.00237(0.76)2————0.0737(12.8)**——— 0.186(16.9)**——— Included 0.7223806 2 Nacogdoches 0.174(1.85)*——0.00397(1.4)4.83E-4(4.92)**—0.174(3.51)**2———————0.00156(0.05) 0.201(2.06)*——— — 0.83930 * 2 * Lubbock ———0.0175(13.3)**5.1E-05(2.98)2—————————— 0.064(2.01)**1.17(6.71)**0.782(2.26)**0.0503(1.01)2 — 0.1141922 es 2 2 ultsforallciti Laredo ———7.5E-07(0.64)21.84E-4(9.84)**7.43E-5(11.27)**———8.515(3.05)**–————— 0.044(1.19)——— — 0.603338 0percentlevel. Res 2**** the1 Table4. Grapevine 0.0628(0.75)——0.00156(1.0)3.27E-4(5.84)9.85E-06(7.6)—0.157(1.8)*——0.0436(1.62)———— 0.191(3.03)**——— — 0.81959 ysigni(cid:142)cantat FortWorth 0.14(37.2)**——0.00671(88.7)**6.07E-4(257)**2.35E-06(21.5)**—0.409(110)**——0.103(67.2)**0.185(24.5)**——— 0.0882(8.24)*——— Included 0.805102948 indicatesstatisticall * Dallas —0.0253(4.23)**20.0673(4.75)**0.00347(12.1)**24.17E-4(60.7)**————————0.0215(63.7)**2— 0.0495(6.08)**——— Included 0.914920 the5percentlevel. Abilene VariableBath0.160(13.1)**Fullbath—2Halfbath—Yearbuilt0.0144(33.7)**2Squarefoot5.86E-4(46.9)**Lotsize9.1E-5(9.0)**Bedroom—Heatac0.452(41.6)**Numstory0.144(4.67)**22Numporch—Garagespace—Structure—Condition—Depreciation—2Yearsold— HistoricStatusHistoric0.191(4.79)**National0.0516(0.834)Texas—Noncontrib— NeighbourhoodIncludedcontrols 2R0.77n7620 **indicatesstatisticallysigni(cid:142)cantat