Copyright Appraisal Institute Jul 1994In this article, a standard regression model is used to perform an analysis that suggests that residential property located in a federally certified historic district sells at a 26% premium compared with similar property outside of the district. Because owner-occupied residential property does not qualify for rehabilitation investment tax credits, the premium is attributed to the "externality" effects that theoretically flow from a location in the historic district. In effect, federal district designation seems to create a protective "zone" that benefits the residential properties in the district.
To the extent that these results are valid for other residential properties in other districts, appraisers should be made aware of the potential adjustments necessary when appraising residential property in federal historic districts relative to potential comparables outside of districts or when using a historic property as a comparable for a nonhistoric property. To assist appraisers in these cases, we offer the following recommendations based on our findings.
* Maintain a list of federally designated property or obtain maps of federal historic districts in the area. Distinguish between federally certified and uncertified property. As an additional precaution, distinguish between properties in federal historic districts and properties certified only by local authorities.(1)
* Stay current with respect to federal legislation affecting federal historic property in order to identify the potential benefits of designation versus its costs or restrictions.
* Perform matched-pair analyses or even simple regression studies (this is possible given the increased computational power of personal computers) to give some indication of the net influence of historic certification.
* Create and continuously update sale databases for historic properties versus nonhistoric properties. Without suitable sale data, the most sophisticated analyses are useless.
Only when appraisers fully understand the market for historically certified property will they be able to accurately assess the impact of historic certification on market values.
The case study presented here offers some insights to appraisers searching for the impact of federal historic district designation. This analysis is of owner-occupied residential property in and out of federal historic districts in the city of Philadelphia.
HISTORIC OVERVIEW
Historic designation can be described as a land-use planning tool used to preserve the history and architecture of groups of buildings, structures, or sites. Preservation advocates see historic designation as a preventive measure used to maintain the status quo while providing a sense of unity with the past. Historic designation also remedies the inability of those who encounter historic properties in less direct ways (e.g., visitation) to register their preferences. Thus the primary benefit of historic designation is believed to be its protection of historical properties in an inefficient market where developers and users are indifferent or oblivious to historical significance.
Opponents see preservation as a bundle of negative restrictions, which, because of the requirements associated with meeting historic standards, must be viewed as a set of encumbrances on the bundle of rights usually associated with real property ownership. The more restrictive the standards, the greater the loss in property rights. Of course, the true impact of historic designation on property values depends on the expected benefits to owners of historic properties as opposed to the loss in property rights perceived to be associated with any restrictions imposed.
The direct benefits of the federal preservation law are essentially limited to the possibility of receiving an investment tax credit (ITC) equal to 20% of qualified rehabilitation expenditures incurred in the renovation of federally certified investment property. Owner-occupied housing, however, does not qualify for the federal tax credits. Consequently, there are no federal ITC benefits associated with the rehabilitation of owner-occupied residential property in federal historic districts. It should be noted, however, that federal standards must be met only if the owner decides to rehabilitate according to federal standards (and receive the ITC). Thus while federal law offers no specific tax benefits to owner-occupants of residential property, there are also few restrictions.
Benefits of federal historic districting may, however, be associated with either the perceived benefits of living in a federally certified historic district or from actual positive "externality" effects derived from the stabilizing, via historic standards, of the surrounding neighborhood. The limited restrictions on owner-occupied homes and the possibility of positive externality benefits from the historic district may cause owner-occupied homes to be sold at a premium compared with property outside of such districts. There is some evidence for this view. A recent study by Asabere and Huffman (1992) found that residentially zoned vacant lots located in federal historic districts attracted a price premium of as much as 131%.(2)
The net impact of historic districts on owner-occupied homes, however, is still far from certain. Historic structures by their very nature may be physically deteriorated or functionally obsolete. The potential presence of substantial obsolescence in historic structures may cause owners to discount partially or totally any premium flowing from perceived or actual externality effects. Further, externality effects themselves may be far from uniform. For instance, although certified as a district (or neighborhood) for historic purposes, an area may currently contain several distinct neighborhoods or generally lack the cohesiveness necessary to benefit substantially from districting. Because the exact impact of these factors is unknown, we conducted an empirical analysis of residential sales to estimate the impacts of historic district location.
HISTORIC DESIGNATION IN PHILADELPHIA
As the birthplace of American independence, the City of Philadelphia has been one of the leaders in the preservation of historic resources. In 1816, protesters persuaded the city to spend $70,000 to purchase the old Pennsylvania State House (now known as Independence Hall) and restore it as a historical landmark. In 1955, the city established the Philadelphia Historical Commission, the first such agency in the nation to have city-wide jurisdiction regarding the preservation of historic resources.(3)
The city has more than 50 federally designated historic districts, the largest concentration of federal districts in any U.S. city. Federal tax incentives have resulted in significant development and preservation activity in the city. Since 1976, over 675 projects with development costs exceeding $950 million (excluding acquisition) have qualified for federal tax incentives in Philadelphia. A study conducted by the National Trust for Historic Preservation revealed that from October 1, 1982, through September 30, 1984, Philadelphia ranked first nationally in total rehabilitation expenditures attributable to projects using the federal investment tax credit and fourth in the nation as to the number of projects.(4)
The city's federal certification history makes it an interesting case for a study of the effects on market values of federal historic preservation. We statistically tested the hypothesis that federal certification increases residential values. The theoretical framework and hypotheses that support our analysis of federal historic designation effects are presented next.
THE HYPOTHESES AND THE EMPIRICAL FRAMEWORK
The objective of this analysis was to search for the impacts of federal historic designation on home values. To detect the potential price effects of federal historic designation, a standard hedonic pricing model represented by equation 1 was employed. Control variables were assigned to all other factors that may affect market value, including measurements for neighborhood, location, macroeconomic conditions, and time of sale. These variables and the historic designation variable are combined as shown.
(Equation 1 omitted)
LogDSP sub i = Actual sale price of the home, deflated by the Consumer Price Index (CPI) for all urban consumers, in log form.
FED = Dummy variable assigning 1 for a home located in a federal historic district and 0 elsewhere.
X = Physical characteristics of the property. Specifically, the following variables are included: AGE; total number of bedrooms and baths (ROOMS); GARAGE; MASONRY construction versus stone or brick; DETACHED, if home is single family detached; and square feet of lot in logs (SQFTLOT).
N'HOOD = Control variables for neighborhood conditions. The following 1990 census tract figures are utilized: median value of improvements (M'VALUE); median household income (HH'INC); and proportion of houses boarded-up (BOARDED).(5)
LOC = Control variables for location. Dummy variables are used to represent various sectors or locations within the city of Philadelphia. Specifically, (0,1) dummy variables are used to represent the northwest (NWEST) and the northeast (NEASY); and location in a single-family zoned area (SFDZ).(6)
MAC = Macroeconomic conditions or time of sale. We use the national unemployment rate (EMPLY) at time of sale and a sequential month-of-sale variable (TIME) that begins with the month and year of the first sale to account for changes in market conditions.(7)
E sub i = Error term.
Again, because the federal designation of homes is not associated with severe regulations or restrictions, it was expected that the positive externality effects associated with federal designation would dominate any negative effects of constraining rights. In other words, it was expected that the coefficient on ED would be significantly positive.
The functional form chosen for this model, the well-known semilogarithmic type (with deflated sale price in logs), was found to provide the best fit for these data.(8) That is, a nonlinear model more easily handled nonlinear partial effects and interactions among variables.
THE DATA AND THE ESTIMATION RESULTS
The Philadelphia data consist of a random sample of 120 owner-occupied sales occurring from December 1986 to May 1990 from a local multiple listing service (MLS).(9) As usual, the MLS contained limited information on the sale transactions, including street addresses; physical characteristics; sale times; and sale prices. Information on historic designation was collected by going through files of the Philadelphia Historic Commission (observation by observation) using street addresses and maps of federal historic districts.(10) Twenty percent of the sample are located in federally designated historic districts. We combined sale prices, sale dates, physical characteristics, and historic location variables with the location, neighborhood, and macroeconomic variables mentioned earlier. Descriptive statistics on all relevant variables for the study are shown in Table 1. (Table 1 omitted)
The results of the regression analyses are shown in Table 2. (Table 2 omitted) The dummy variable for federal designation produces significant results. The estimated coefficient on FED is 0.228401 and is significantly different from zero at the 95% level of confidence. Transforming the coefficient to its anti-log equivalent (necessary because the variable is a [1,0] variable and price is in log form) yields a 26% (or e sup 0 2 sup 3 - 1) price premium. The positive results are consistent with the results found by Schaeffer and Ahern (1988) and Asabere and Huffman (1992).(11)
As for the nonhistoric influences, variables controlling for physical characteristics of the property ROOMS, SaFTLOT, and DETACHED behave as expected. All are significantly positive at conventional levels. GARAGE is positive but insignificant while MASONRY is negative but insignificant. The AGE variable is positive and significant, which runs counter to the theory that increasing age results in physical obsolescence. The average age of the study sample in Philadelphia, however, is about 47 years (Table 1). Thus AGE may not be an "effective age" variable reflecting the present condition of the building. As buildings get older they undergo extensive renovations. Thus we theorize that in the case of Philadelphia, with its large inventory of older housing, the AGE variable may well be proxying the substantial renovations that have occurred over time.
The neighborhood measure BOARDED is significantly negative as expected. M'VALUE and HH'INC, although positive, are statistically insignificant. The city location variables (NWEST and NEAST) are appropriately negative and significant compared with center-city locations. The single-family zoning (SFDZ) variable is insignificant. Our macroeconomic variables, EMPLY and TIME, are also insignificant. Overall, this model explains nearly 80% (R sup 2 = .78) of the variation in our dependent variable, a very acceptable proportion for studies of this type.
CONCLUSION
A residential property located in a federally certified district carries roughly a 26% price premium relative to property located outside of a federally certified district. Because investment tax credits, generally assumed to be the benefit associated with federal historic districting, are unavailable to owners who rehabilitate owner-occupied property, we attribute the premium to the positive externality effects that flow from the historic designation of a district or zone.(12)
While it is clear that additional studies of the impact of historic certification are needed, these results should alert appraisers to the necessity of developing suitable guidelines for the appraisal of historic properties.
1. The impact of local designation may differ from federal designation because local designation may be more stringent than federal requirements. Of course, local requirements will vary in strength from place to place. For example, a recent study by Asabere and Huffman, "The Adverse Impacts of Local Historic Designation: The Case of Small Apartment Buildings in Philadelphia," The Journal of Real Estate Finance and Economics (forthcoming), found that apartment buildings designated historic by local authorities in the city of Philadelphia sold at a 20% discount. In this study, we originally included a variable to measure the impact of local designation. Only about 6% of our sales, however, were locally designated. Therefore at the suggestion of an anonymous reviewer we dropped the local historic variable from our final report.
2. The few remaining studies on the Impact of historic certification have generated mixed, contradictory, or inconclusive results (see Paul Asabere, George Hachey, and Steven Grubaugh, "Architecture, Historic Zoning, and the Value of Homes," Journal of Real Estate Finance and Economics, v. 2 (1989): 181-195; Virginia Benson and Richard Klein, "The Impact of Historic Districting on Property Values," The Appraisal Journal (April 1988): 223-232; Deborah Ann Ford, "The Effect of Historic District Designation on Single-Family Home Prices," AREUEA Journal (Fall 1989): 353-362; William Lockard and Dudley Hinds, "Historic Zoning Considerations in Neighborhood and District Analysis," The Appraisal Journal (October 1983): 485-497; and Peter Schaeffer and Cecily Ahern, "The Impact of Historic District Designation on Property Values: An Empirical Study," unpublished paper, University of Colorado at Denver (December 1988).
3. Philadelphia City Planning Commission, Analysis of Historic Preservation Techniques for City of Philadelphia (October 30, 1986).
4. National Trust for Historic Preservation, FACTS (Washington, D.C.: National Trust for Historic Preservation, July 12, 1985).
5. The sale observations occur in 13% (46 of 365) of the census tracts in the city. The neighborhood quality variables were based on these census tracts.
6. Philadelphia, located on a peninsula at the confluence of the Delaware and the Schuylkill Rivers, is roughly divided into the downtown area (center city, about three square miles in area), and outlying areas to the south, west, northwest, and northeast. The original city was surveyed and settled in the late 1600s. Areas to the northwest (e.g., Germantown) settled shortly thereafter. Areas in the extreme far northeastern section of the city were not developed until the mid-1900s.
The geographical dispersion of the 120 observations is such that about 20% of the sales occurred within the center city, 50% within the larger northeast, 20% within the northwest, and the remaining 10% within the other sectors of the city. These sectors are used to create the dummy variables to control for specific locations. It is possible that if we do not control for location-related differences sufficiently our historic variable would be mis-specified. We therefore experimented with a number of location variables that would account for potential differences in sale prices as a result of geographical differences, including the distance to center city (as measured to City Hall) and various methods of grouping properties geographically. Our basic findings in this regard, that properties in the northwest and northeast sold for less compared with properties in the immediate center city area, were consistent in the various trials. Our zoning variable choice is also the result of numerous trials across various zoning codes. We are comfortable that we have sufficiently captured or "controlled for" the potential impacts on sale prices of geographical and zoning differences such that our historic variable is not significantly biased by such differences.
7. Our sales occurred over approximately four years. Our macroeconomic variable and our time-of-sale variable should control for the potential impacts associated with changes in market conditions. In addition we use deflated sale price as a dependent variable to control for any inflationary impacts on price over the time period.
8. See G. Box and P. W. Tidwell, "Transformation of the Independent Variables," Technometrics (1962): 531-550; and G. Box and D. Cox, "An Analysis of Transformations, Journal of the American Statistical Association 26 (Ser. B, 1964): 211-243 for discussions of these tests.
9. Not all sales occurred via MLS listings and omission could affect our historic variable. Such services, however, provide the most comprehensive method of retrieving past sales. We have no reason to suspect that eliminating sales not occurring via the MLS significantly biases our sample with respect to historic and nonhistoric property sales.
10. This explains our choice for a rather small sample of 120 observations. The usual caveat associated with small sample studies applies here.
11. Paul Asabere and Forrest Huffman, "Historic Districts and Land Values," The Journal of Real Estate Research (Spring 1992); and Schaeffer and Ahern.
12. An interesting aspect of this study is the size of the premium found for houses (26%) versus the much larger premium (131%) discovered by Asabere and Huffman (1992) for vacant lots zoned residential in the same city. The reduction in premium might be partially attributable to the presence of the functional obsolescence mentioned earlier. Unfortunately, our data do not allow us to separate these items.
Paul K. Asabere, PhD, is professor of real estate at the School of Business and Management at Temple University. He received a PhD from the 
University of Illinois.
Forrest E. Huffman, PhD, is associate professor of real estate at the School of Business and Management at Temple University. Mr. Huffman received a PhD from the University of South Carolina.
My Research
