2003:abstract
This study reports a finding of a significant effect on the selling price of a single-family house given its proximity to a registered sex offender's residence. The effect is an increasing function of proximity that varies with offender classification and with the community notification system employed. For more dangerous offenders, the effect is significant for houses located up to 0.3 mile from an offender. Houses located within 0.1 mile of an offender sold for 17.4 % less, on average, than similar houses located farther away. For less dangerous offenders, the significant effect extends to 0.2 mile from the offender's residence, and the effect is smaller.
**********
Real estate literature is rich with papers reporting the results of studies conducted to determine the effect of externalities on housing value. (1) While the list of previously examined externalities is extensive, a notable exception is the proximity of the residence of a registered sex offender to the subject property. La Fond (2) has shown that the direct and indirect costs of enacting and implementing sexual predator laws are expensive for the public at large. Whether an additional monetary burden must be borne by property owners in close proximity to an offender's residence, however, has not been investigated. Relatively recent additions to the law facilitate an examination of this issue and enable us to begin filling this gap in the literature. In 1996, Congress passed "Megan's Law," (3) which required states to enact laws governing sex offender registration and community notification. Today, every state has complied with the federal requirement and has laws that require offenders to register with police (or a government agency) and specify how registration information is released to the public.
In this study, single-family house transactions that occurred during 2000 in Montgomery County, Ohio, are examined to determine the effect on selling price given the house's proximity to the residence of a registered sex offender. A significant effect was discovered. The effect is an increasing function of proximity that varies with the community notification system employed. Where limited disclosure was employed for more dangerous offenders, the negative effect extends to 0.3 mile from an offender's residence. Compared to comparable houses located farther away, houses located within 0.1 mile of an offender's residence sold, on average, for 17.4% less. Where passive notification was employed for relatively less dangerous offenders, the negative effect was significant for houses located up to 0.2 mile from the offender. Compared to comparable houses located farther away, the houses located within 0.1 mile of an offender's residence sold, on average, for 7.5% less. The results suggest that when appraising a single-family house, appraisers may want to place more reliance upon the cost approach and/or include an appropriate adjustment to a comparable's sale price when using the sales comparison approach.
This article first describes the systems for sex offender classification and public notification employed in Ohio during the study period. Next, the pricing implications of both the categories of sex offenders and the notification systems are described. Then the data and methodology are presented along with a brief explanation of the geocoding process used to determine the distance between each sales transaction and the nearest offender's residence. Finally, the results and their implications for appraisers are discussed.
Sex Offender Classification and Community Notification in Ohio
Courts in Ohio classify sex offenders into one of three categories: "sexual predator" "habitual sex offender," or "sexually oriented offender." The classification depends on the history of the offender and the court's opinion of the likelihood that the offender will commit another offense. All convicted sex offenders in Ohio are required to register with the sheriff's office in the county in which they reside. While members of any category may cause public concern, sexual predators are deemed to present the greatest risk to the community. In Ohio, a sexual predator is defined as a person who has been convicted of, or pleaded guilty to, a sexually oriented offense and who is considered likely in the future to commit additional sexually oriented offenses (4) A habitual sex offender is defined as a person who has been convicted of, or pleaded guilty to, committing a sexually oriented offense, and who has previously been convicted of or pleaded guilty to one or more sexually oriented offenses. (5) Finally, a sexually oriented offender is defined as a person who has been convicted of, or pleaded guilty to, a sexually oriented offense. (6)
Ohio law specifies that each county sheriff's office must follow a practice sometimes referred to as "limited disclosure," i.e., proactive notification that applies only to sexual predators and some habitual sex offenders. Under this system, the sheriff's office must notify a variety of parties within 72 hours after the sexual predator or habitual sex offender moves into a residence. Parties that are notified include owners of houses adjacent to the offender's residence and school officials (who in turn sometimes notify the parents of students). Interested parties may also learn about sexual predators and habitual sex offenders through two "passive notification" mechanisms. In some counties (including the sample county here), information about predators may be viewed over the Internet on the sheriff's office web site. In addition, interested parties may personally request information about sex offenders (in any category) from the sheriff's office. Passive notification is the disclosure system used for sexually oriented offenders and some habitual sex offenders. Under this system, interested parties must initiate contact with the sheriff's office to discover the location of an offender's residence.
Effectively, there are four categories of offenders in Ohio: (1) sexual predators where limited disclosure applies, (2) habitual sex offenders where limited disclosure applies, (3) habitual sex offenders where passive notification applies, and (4) sexually oriented offenders where passive notification applies. For the purpose of this study, however, sex offenders were divided into two groups based on the disclosure procedure: (1) sexual predators and habitual sex offenders where limited disclosure applies, and (2) sexually oriented offenders and habitual sex offenders where passive notification applies. Meaningful analysis based on offender classification is problematic in the present study because there are only six habitual sex offenders in the database.
Price Implications of Offender Classification and Notification Systems
Intuitively, it would appear that if a house is located in close proximity to a sex offender, selling price effects should be negative. Few people would elect to live next door to a felon of any type. Although convicted arsonists and murderers also may pose a risk to a community, current law makes it easier for market participants to identify the location of sex offenders and build that information into transaction prices. (7) Recidivism by sex offenders is well documented in the literature. (8) Analysis of data collected by the United States Department of Justice (DOJ) suggests that proximity to a sex offender's residence increases one's risk of becoming a victim. Approximately every five years, DOJ administers a comprehensive questionnaire to a nationally representative sample of prison inmates. Sex offenders accounted for 8.5% of state prisoners included in the most recent survey, the 1997 Survey of Inmates in State Correctional Facilities (9) The results indicate that while sex offenders commit their crimes over substantial geographic areas, they tend to perpetrate their crimes close to home. Sex offenders reported that 85.1% of their crimes were committed in the same city in which they resided at the time of arrest (compared to 80.2% for other offenders), and 64.9% of sex offenders reported committing their offense in their own neighborhoods (compared to 44.6% for other offenders), (10)
It is plausible that larger price discounts would be associated with the proximity of a house to a more dangerous offender compared to proximity to a less dangerous offender. It also is reasonable that larger discounts would be associated with a notification system where government authorities take an active role in the process compared to a system where they do not. No published studies, however, have documented that the public recognizes or fails to recognize the difference between offender classifications. Therefore, it is uncertain whether the public distinguishes between offender classifications, and whether any differences in selling price discovered here are due to the relative risk posed by the offender or due to the notification system employed. If the public does not distinguish between offender classifications, any selling price difference attributed to difference in offender classification should disappear when the same notification system is used for all offenders.
A selling price results from the negotiations between the seller and buyer. The presence of an offender may motivate owners to accept a low offer to consummate a sale, and the model employed here will capture that effect. However, in order to estimate the actual selling price effect of proximity to an offender, both price and marketing time should be investigated because, from the seller's perspective, extra time on the market lowers the present value of the selling price. Unfortunately, the transaction data set used in this study does not include reliable time on the market information. (11) Because marketing time is not included in the model used in this study, the selling price effect discovered may understate the effective selling price effect. In fact, if a sufficient number of owners wait for an undiscounted offer, failure to include time on the market would make it impossible to detect any selling price effect. While knowledgeable buyers may refuse to bid on a house located in close proximity to an offender's residence, or may lower their bid, the price offered by uninformed buyers will be unaffected. Therefore, without time on the market in the model, if almost all owners are willing to search until an uninformed buyer is located, no selling price effect should be detected.
The existence of knowledgeable buyers is a critical element in order for selling price discounts to occur. This study could not determine the percentage of knowledgeable buyers in the data sample, but information obtained from the sheriff's office indicates that they are present. Although the sheriff's office does not track web site hits, they reported receiving approximately 3,000 personal requests for information during the study year (about 10% of which came from real
Data
Two data sets are used in this study. The first contains information about registered sex offenders, including their addresses and classifications. This information was obtained from the Montgomery County Sheriff's Office. Their database contained information on 22 sexual predators where limited notification applied, 4 habitual sex offenders where limited disclosure applied, 2 habitual offenders where passive notification applied, and 221 sexually oriented sex offenders where passive notification applied. (12) Therefore, the offender data used in this study contains 26 offenders where limited disclosure applied and 223 offenders where passive notification applied.
The second data set consists of 3,208 single-family houses located in Montgomery County, Ohio, that sold during 2000. (13) Montgomery County contains 461.7 square miles and has a population of 558,427. There are about 226,182 occupied household units in the county and 142,371 of these are owner occupied. Dayton (population 167,475) is the county seat. Another 216,100 people live in the next nine largest cities in the county. Approximately 5% of the population reside in rural areas. Transaction data and property characteristics for sold houses were obtained from public records offices in Montgomery County and the Dayton Area Board of REALTORS[R]. Descriptive statistics of the transactions in this study are presented in Table 1.
Geocoding
An important variable in the study is the distance between each sold house and the residence of the nearest offender. To determine this distance, both the sold house and the offender data sets were geocoded over the Internet using software provided by Tele Atlas. (14) This produced the latitudes and longitudes for observations in both data sets. In addition, Tele Atlas rated each geocoded location for accuracy based upon how well the property description submitted matched the information in their database, and indicated the level of accuracy for each observation by attaching a quality rating ranging from 1 to 6. All geocodes that did not meet the highest quality rating of 1 were eliminated from the study. (15)
Next, using the geocoded latitudes and longitudes, ArcView[R] software was used to calculate the distance from each address in the sold house data set to the nearest address in the offender data set. (16) This distance was calculated twice: once from each sold house to the nearest offender who was subject to limited disclosure, and again from each sold house to the nearest offender subject to passive notification. In determining the closest offender, it was required that the offender be in residence at least one week before the purchase contract was signed.
In addition, the geocoded points were plotted on maps using the ArcView[R] software. A map of the location of sold houses and the residences of offenders subject to limited disclosure is presented in Figure 1, and a map of the location of sold houses and the residences of offenders subject to passive notification is shown in Figure 2. These maps show that the clustering of sold houses and offenders' residences are similar, The source of the shape file used to display Montgomery County in both exhibits was the 1990 Topologically Integrated Geographic Encoding and Referencing System (TIGER[R]) files from the U. S. Census. (17)
[FIGURES 1-2 OMITTED]
Methodology
There are three basic steps in the methodology. The first two steps are used to test the following two null hypotheses:
[H.sub.O]: Ceteris paribus, the selling price of a single-family house is unaffected by its proximity to the residence of a registered sex offender where limited disclosure applies.
[H.sub.O]: Ceteris paribus, the selling price of a single-family house is unaffected by its proximity to the residence of a registered sex offender where passive notification applies.
Stepwise Regression
The objective of the first step of the methodology is to identify variables to be included in a hedonic regression. To accomplish this task, a battery of property characteristic variables is subjected to stepwise regression. (18) Each variable must maintain a p-value of.05 or less in order to enter and remain in the model.
In the second step of the methodology, a binary variable, PROX, is added to the model, and a two-level analysis of covariance (ANCOVA) procedure is performed on the expanded model. (19) PROX is assigned a value of 1 if the subject house was located within a specified area relative to the nearest offender, or 0 if the subject house was located outside the specified area. A series of non-overlapping concentric rings (rings) is used to define the area where PROX equals 1. In the first iteration, PROX equals 1 if the subject house is located 0.1 mile or less from the nearest offender. In subsequent iterations, the maximum radius is expanded by increments of 0.1 mile and observations from previously tested rings are eliminated from the sample. For example, when the maximum radius is set at 0.2 mile (the second ring), houses located no more than 0.2 mile but more than 0.1 mile from an offender are compared to houses located farther away than 0.2 mile from the offender. (20) The iterative process is conducted twice, first for offenders subject to limited disclosure and again for offenders subject to passive notification, to determine the point at which PROX becomes and remains insignificant at the 95% confidence level. When estimating the model for offenders subject to passive notification, any observations located within the distance found to be significant for an offender subject to limited disclosure are eliminated from the sample in order to eliminate the effect of the more dangerous offender. (21) A significant negative estimated coefficient for PROX indicates that the average selling price of houses located within the specified ring is less than the average selling price of houses located farther away.
In the final step of the methodology, the percentage selling price effect for each ring is calculated by dividing the dollar price effect by the average selling price of houses located within the ring. The stepwise regression indicates that the basic hedonic model contains thirty-four variables, as shown in Equation (1). (22)
(1) SP = [alpha] + [[beta].sub.1]SQFT + [[beta].sub.2]AGE + [[beta].sub.3]LOT + [[beta].sub.4]FIRE + [[beta].sub.5]BATH3 + [[beta].sub.6]OWN + [[beta].sub.7]FULL + [[beta].sub.8]WINTER + [SIGMA][[beta].sup.34.sub.n=9]LOC + [xi]
where:
SP = the selling price of the subject property
[alpha] = the intercept
[[beta].sub.n] = the estimated coefficients
SQFT= the amount of living space in square feet
AGE = the age of the house in years
LOT = the size of the lot in acres
FIRE = the number of fireplaces in the house
BATH3 = a binary variable equal to 1 if the house has three or more bathrooms, equal to 0 otherwise (23)
OWN = a binary variable equal to 1 if the house is owner occupied, equal to 0 otherwise (24)
FULL = a binary variable equal to 1 if the house has a full basement, equal to otherwise
WINTER = a binary variable equal to 1 if the house sold during December, January, or February, equal to 0 Otherwise (25)
LOC = a vector of binary variables equal to 1 if the house is located in a particular tax district of 48 tax districts, equal to 0 otherwise (26)
[xi] = the error term
A joint test for functional 10nrta and homoskedasticity of the error term that follows the approach of White (27) was conducted on Equation (1) using the SPEC option in PROC REG available in SAS. (28) Because of the large number of observations in the sample, the estimated variance-covariance matrix degenerated into singularity rendering the test results suspect. Therefore, the joint test was conducted again using only the nonbinary independent variables. Although the singularity problem was eliminated by this adjustment, the null hypothesis of homoskedasticity was rejected for LOT, SQFT, and AGE. A variety of transforms on both the dependent and independent variables were tested; in each case the null hypothesis of homoskedasticity of the error term was rejected. It was apparent from the large number of degrees of freedom in the test results that the large number of observations in the data set caused rejection of the null hypothesis even for slight deviation from homoskedasticity. Therefore, nonquantitative methods for detection of homoskedasticity (i.e., residual analysis) were utilized. Examination of residual plots indicated that eight (high price) outliers were present in the data, but very little heteroskedasticity. The outliers were eliminated from the data, and the functional form that resulted in the highest [R.sup.2], linear, was selected. The final residual analysis indicated that heteroskedasticiry and nonlinearity were not problems.
A collinearity diagnostics program that follows the approach of Belsley, Kuh, and Welch, (29) available on SAS, was conducted. The results indicate a moderate degree of multicollinearity is present in the data, but not enough to be harmful in the sense that the estimates of the regression are highly imprecise or unstable. The highest condition number was 13.79 and the highest proportion of variation for any variable was 44 (the second highest proportion of variance for any variable was .22).
A critical assumption of ANCOVA over and above the assumptions made in regression analysis is that of homogeneity of regression. Specifically, that the slopes of all the regression lines in simple regression (or the slope of the hyperplanes in multiple regression) are equal with respect to the qualitative variable being tested (i.e., PROX). In other words, there should be no interaction between PROX and covariates. The interaction was tested and found to be insignificant for all variables except SQFT. The interaction for SQFT and PROX was investigated and found to be of magnitude and not of direction. Using the method outlined by Tabachnick and Fidell, (30) SQFT was transformed into a blocking variable and the ANCOVA model was reestimated. (31) No significant change occurred in either the estimated coefficient or p-value for PROX. Therefore, the robustness of ANCOVA indicated the model was appropriate.
Results of the ANCOVA Procedure
The results of the ANCOVA procedure, where PROX is set at the maximum significant radii, are shown in Table 2. In Table 2, the explanatory variables are listed in the first column; the respective estimated coefficients for proximity to offenders subject to limited disclosure and passive notification are shown in the second and fourth columns respectively. The p-value for each variable is shown in the third and fifth columns. Examination of Table 2 reveals that the model fits the data well. The adjusted [R.sup.2] indicates that the model explains over 72% of the variation in selling price. Previous hedonic studies have found that selling price tends to be negatively related to AGE and WINTER, and positively related to SQFT, LOT, FIRE, FULL, and BATH3. (32) The sign of each property characteristic variable in the model is consistent with previous research. Because over 37% of all houses in the sample are not owner occupied, OWN was included to control for any price difference that may be attributable to the occupancy intentions of the purchaser. The positive sign on OWN is subject to multiple interpretations. It indicates that buyers who intend to live in the property pay more than buyers who plan to rent it to others while living elsewhere themselves. This could mean that nonoccupant owners are systematically more aware of the presence of nearby offenders and factor that information into their purchase offers. Another possible explanation is that absentee owners may be purchasing houses in poor condition. The study did not prove this because property condition was not a variable in the model, but OWN may be serving as a proxy for property condition.
Focusing on the variable of interest, PROX, the results of the ANCOVA procedure enable the rejection of both null hypotheses. Note that the estimated coefficient for PROX is negative for offenders subject to both notification systems. The negative sign means that there was a significant negative effect on the selling price of single-family houses in the sample due to their proximity to the residence of a sex offender. Specifically, it means that the average selling price for houses located within the specified rings is significantly less than the average selling price for comparable houses located farther away from the offender. The ANCOVA procedure indicated that a significant selling price effect occurs for houses located up to 0.3 mile from the residence of an offender subject to limited disclosure. The ANCOVA procedure also showed a significant selling price effect occurs for houses located up to 0.2 mile from the residence of an offender subject to passive notification. If the maximum radii are extended beyond these distances, no significant difference is observable in an average selling price for houses located within the specified ring and those located farther away.
To show the effect on selling price as the distance from the offender's residence increases, partial ANCOVA procedure results are summarized in Table 3. The results for proximity to offenders subject to limited disclosure are shown in the upper portion of the tablet and the results for proximity to offenders subject to passive notification are shown in the lower portion of the table. PROX (in miles) is shown in the first column. The number of sold houses within each ring (n) is shown in the second column. The dollar price effect due to proximity to an offender is shown in the third column. The p-value for the significance of the difference between selling prices for houses located inside the ring compared to those located farther away is shown in the fourth column. Finally, the percentage price effect, which is the dollar price effect for each ring divided by the average selling price of houses sold within the ring, is shown in the fifth column.
Focusing on the upper portion of Table 3, it is shown that the price effect is significant for houses located up to 0.3 mile from the residence of an offender subject to limited disclosure. Compared to comparable houses located farther away, houses located within 0.1 mile of an offender's residence sold, on average, for 17.4% less. The effect drops as distance from the offender's residence increases. On average, houses located between 0.1 and 0.2 mile from an offender's residence sold for 10.2% less compared to houses located farther from the offender. Also, houses located between 0.2 and 0.3 mile from an offender's residence sold, on average, for 9.3% less. Approximately 7.7% (247) of all the houses in the sample were located within 0.3 mile of an offender subject to limited disclosure. Note that the number of observations in each ring increases as the minimum ring radius is increased (by a constant 0.1 mile). This phenomenon occurs because the area within the expanded ring is larger than the areas within the rings located closer to the offender.
Focusing on the lower portion of Table 3, it is shown that the price effect is significant for houses located up to 0.2 mile from the residence of an offender subject to passive notification. Compared to comparable houses located farther away, houses located within 0.1 mile of an offender's residence sold, on average, for 7.5% less. Again, the effect drops as distance from the offender's residence increases. On average, houses located between 0.1 and 0.2 mile from an offender's residence sold for 5% less compared to houses located farther away from an offender. Approximately 25% (802) of all houses in the sample were located within 0.2 mile of an offender subject to passive notification. (33) Because the sample market included almost ten times the number of offenders subject to passive notification as offenders subject to limited disclosure, it is not surprising that more houses in the sample were located within the significant price effect area for the former classification.
Summary and Conclusions
This study shows that a monetary burden must be borne by house sellers in close proximity to a registered sex offender's residence. Examining single-family house transactions that occurred in Montgomery County, Ohio, during 2000, a significant negative effect upon selling price is discovered due to a house's proximity to the residence of a registered sex offender. The effect is an increasing function of proximity that varies with the community notification system employed, which in turn depends on the risk the offender poses to the community. Limited disclosure is the notification system used for offenders deemed to present a relatively greater risk to the community. Under this system, the sheriff's office notifies owners of houses adjacent to the offender's residence and school officials (who sometimes notify the parents of students). Passive notification is the disclosure system used for offenders deemed to present relatively less risk. Under this system, interested parties must contact the sheriff's office to discover the location of an offender's residence.
It is intuitive that larger discounts would be associated with the proximity of a house to a more dangerous offender compared to proximity to a less dangerous offender. It also is a reasonable assumption that larger discounts would be associated with a notification system where authorities take an active role in the process compared to a system where they do not. Because we are unsure if the public distinguishes between offender classifications, we cannot be certain whether the difference in selling price effect discovered here is due to the relative risk posed by the offender, or if it is due to the notification system employed. If the public does not distinguish between offender classifications, differences in selling price effects should not be present when the same notification system is used for all offenders. Perhaps this could be tested in a state that employs the same notification system for all offenders.
The study results are consistent with both of the above possibilities. Where limited disclosure applies, significant selling price effects are greater and extend farther from an offender's residence than when passive notification applies. Where limited disclosure is employed, the significant negative effect extends to 0.3 mile from an offender's residence. Compared to comparable houses located farther away from an offender, on average, houses located within 0.1 mile of an offender sold for 17.4% less. Houses located between 0.1 and 0.2 mile from an offender sold for 10.2% less, and houses located between 0.2 and 0.3 mile from an offender sold for 9.3% less.
Where passive notification is employed, the significant negative effect extends to 0.2 mile. Compared to comparable houses located farther away, on average, houses located within 0.1 mile of an offender's residence sold for 7.5% less, and houses located between 0.1 and 0.2 mile from an offender sold for 5% less. Despite the relatively compact areas where significant price effects occur, a substantial number of the houses in the sample were located close enough to an offender that they may have been affected. Approximately 7.7% of the houses were located within 0.3 mile of an offender subject to limited disclosure, and approximately 25% of the houses were located within 0.2 mile of an offender subject to passive disclosure.
The study results may actually understate the true financial effect of proximity to an offender's residence because the model did not include a variable for time on the market. From the seller's perspective, extra time on the market lowers the present value of the selling price. The presence of an offender may motivate some owners to accept a low offer to consummate a sale, and the model employed here captures that effect. However, if the owners want an undiscounted price for their house, they may have to extend their search time because knowledgeable buyers will either refuse to make an offer or lower their offer to account for the presence of the offender. To the degree that owners wait for an undiscounted offer from an uninformed buyer, failure to include time on the market will mask the true effect of proximity to an offender on the effective selling price. An examination of additional markets with reliable time on the market data seems a logical extension to this research effort.
To keep the problem tractable, two important assumptions were made concerning the impact of offender proximity on the selling price of nearby houses: that the presence of a more dangerous offender dominates, and that the presence of the nearest offender dominates. This does not imply that the presence of additional offenders located farther away has no effect. Future research efforts could examine the effect of proximity to multiple offenders in the same classification as well as interaction effects between offender classifications.
Implications for Residential Appraisers
What are the implications of this study for residential appraisers? First, this problem is likely to become more widespread because the number of registered offenders is growing. As a result, appraisers may want to modify the appraisal process. As a prerequisite, it is suggested that the appraiser ascertain the local price effect, if any, attributable to proximity to a sex offender. If no effect is present, maintain the status quo. We suspect, however, that the findings presented in this article are not unique. If a price effect is discovered, it is suggested that when estimating value with the sales comparison approach, an adjustment to comparable selling price may be warranted to account for offender proximity. In certain cases, it also may be prudent to place more reliance on the cost approach.
In valuing a single-family house, many appraisers place heavy reliance on the sales comparison approach. In fact, it is not unusual for the final estimate of value to equal the indicated value derived from this approach (with the cost approach used primarily as a device to ensure the reasonableness of the sales comparison's indicated value). This practice can be maintained if the price effect due to offender proximity is identical for the subject property and each comparable property. If this is not the case, appraisers must modify their methodology to accurately estimate value using the sales comparison approach. The potential effect of proximity to an offender must be calculated for the subject property, as well as the effect included in the transaction price for each comparable. Then, each comparable's sale price should be adjusted to account for the difference in offender price effects between the subject and the comparable.
The actions of an appraiser in response to this study also depend, in part, on the purpose of the appraisal. For example, the adjustment described in the preceding paragraph is warranted if the purpose is to support a mortgage loan, or if the appraisal is being prepared for an individual contemplating the acquisition of a house for investment purposes. However, if the appraisal is to establish value for the origination of an insurance policy or for supporting an insurance claim, it is suggested that the cost approach be assigned more importance in arriving at the final value estimate. After all, if the above-recommended adjustment to the sales comparison approach results in a lower value estimate, this does not reduce the replacement cost of the property. Also, to the extent the offender proximity effect is reflected in the comparable sale prices, a lower value estimate could result whether or not the adjustment is made.
Finally, it should be noted that most of the offenders in this study did not change their residence during the study year. However, because offenders are free to move (and report to authorities their new location), the financial burden associated with an offender's presence may be transitory for a particular house owner. A determination of exactly how long it takes for the negative price effect to disappear after the offender leaves remains a topic for further research.
The authors thank the Ohio Link and the Paul Lawrence Dunbar Library at Wright State University for their generous support by providing the ESRI software through a site licensing arrangement. We also thank the Ohio GIS-Net for providing GIS advice, the Department of Urban Affairs and Geography at Wright State University for their assistance in this study, and the reviewers who commented on this paper.
(1.) In general, previous studies find that if an externality is perceived as favorable, it has a positive effect on the value of the subject property; if the externality is perceived as unfavorable, it has a negative effect. Negative price effects have been demonstrated for houses in close proximity to other negative externalities including a variety of environmental hazards. For a review of the environmental hazard literature, see Melissa A. Boyle and Katherine A. Kiel, "A Survey of House Price Hedonic Studies of the Impact of Environmental Externalities," Journal of Real Estate Literature 9, no. 2 (2001): 117-144.
(2.) John Q. La Fond, "The Costs of Enacting a Sexual Predator Law, Psychology, Public Policy and Law 4 (1998): 468-504.
(3.) P.L. 104-145, [section]1, 110 Stat. 1345. One of the stimuli for this law was the case of Megan Kanka, who in 1994 was raped and killed by a repeat sex offender who, unknown to Megan's parents, lived across the street from her home.
(4.) Ohio Revised Code [section] 2950.01; offenses included in this statute are rape, sexual battery, gross sexual imposition, kidnapping, abduction, unlawful restraint, criminal child enticement, corruption of a minor, compelling prostitution, endangering children (under age 18), pandering obscenity, pandering sexually oriented material involving a minor, and illegal use of a minor in nudity-oriented material. Sexual predators must report to the sheriff's office every 90 days for life.
(5.) Habitual sex offenders must report to the sheriff's office once annually for 20 years.
(6.) Sexually oriented offenders must report to the sheriff's office once annually for 10 years.
(7.) Unlike in some other states (e.g., Alaska), house sellers in Ohio are not required to report the presence of sex offenders on the mandatory seller disclosure form.
(8.) See for example, Dennis M. Doren, "Recidivism Base Rates, Predictions of Sex Offender Recidivism, and the 'Sexual Predator' Commitment Laws," Behavioral Sciences & the Law 16 (1998): 97-114; D. M. Greenberg, "Sexual Recidivism in Sex Offenders," Canadian Journal of Psychiatry 43 (1998): 459-465; Michael P. Hagan and Karyn L. Gust-Brey, "A Ten-Year Longitudinal Study of Adolescent Rapists Upon Return to the Community," International Journal of Offender Therapy and Comparative Criminology 43, no. 4 (1999): 448-458; R. K. Hanson, R. A. Steffy, and R. Gauthier, "Long-Term Recidivism of Child Molesters," Journal of Consulting and Clinical Psychology 61 (1993): 646-652; R. A. Prentky, et al., "Recidivism Rates Among Child Molesters and Rapists: A Methodological Analysis," Law and Human Behavior 21 (1997): 635-659; V. L. Quinsey, M. E. Rice, and G. T. Harris, "Actuarial Prediction of Sexual Recidivism," Journal of Interpersonal Violence 10 (1995): 85-105; and M. C. Seto and H. E. Barbaree, "Psychopathy, Treatment Behavior, and Sex Offender Recidivism," Journal of Interpersonal Violence 14 (1999): 1235-1248.
(9.) U.S. Department of Justice, Office of Justice Programs, Bureau of Justice Statistics.
(10.) These figures were obtained by personal communication between the authors and employees of the Bureau of Justice Statistics, United States Department of Justice.
(11.) The local real estate board reports days on the market only for the most recent listing contract; time for expired listings is not included in the figure they report.
(12.) Not all of the offenders lived in Montgomery County for the entire year. Twenty-five of the 26 offenders where limited disclosure applied lived in the county at year-end. Ten of this group did not live in the county at the beginning of the year. All 223 offenders where passive notification applied lived in the county at the end of the year, but 61 of this group did not reside in the county (or had not yet become registered sex offenders) at the beginning of the year.
(13.) The Dayton Area Board of REALTORS[R] reported 5,614 single-family home sales during the study period. Ambiguous geocoding resulted in 115 observations being discarded from the sample. The remainder were eliminated because of incomplete data.
(14.) Tele Atlas can be found at www.geocode.com. Tele Atlas provides the "Block Face Match" (BFM), which represents the best match rather than parcel level accuracy. In essence, rather than specifying the latitude and longitude at a particular point on each property (e.g., front center) the geocode derived from a BFM is actually a geometric estimation. Tele Atlas stores the beginning and ending address range for a block, and knows the number parity (odd or even). For example, the geocode assigned to 150 Eagle Street, would be roughly halfway between the presumed beginning and ending address range of 100 and 198. The interactive web site for Tele Atlas was used in this study for the geocoding because it provides a high level of location confidence. It can accurately position every point in the data set to six decimal points of a degree, and in the Montgomery County, Ohio area, this accuracy translates to less than 20 inches. It is also fast, repeatable, commonly used in geographic information systems (GIS) work, and accessible to anyone conducting a study. There is a charge for the service, but the cost per address is low.
(15.) If points with quality ratings of 2 to 6 are included, the accuracy of the study suffers because the system places poorly geocoded points at the centroid of the zip codes
. This would result in a cluster of sold houses (and/or offenders) located in one place, which is obviously not the case. No sexual predators or habitual offenders were eliminated from the study due to bad geocodes.However, 4 sexually oriented offenders and 11S sales transactions were eliminated for this reason.
(16.) ArcView[R] GIS (Redlands, Calif.: Environmental Systems Research Institute (ESRI), Inc.). In determining the proximity measures used in Equation (1), it was required that the offender had been in residence for at least one week before the purchase contract was signed. At some point in the year, an offender may have lived closer to a sold property than the one used to calculate the proximity measure. If the offender was not in residence prior to the sale, there is no way to attribute any price effect to them.
(17.) TIGER[R] Line Files (Washington, D. C.: United States Department of Commerce, Bureau of the Census, 1992).
(18.) Stepwise regression is the appropriate process to use in this case because the major objective is not to predict the value of the dependent variable; the major objective is the analysis of the independent variables, in particular PROX. In a predictive model, multicollinearity inhibits the analysis of independent variable effects due to the instability of the regression coefficient of each independent variable. In stepwise regression, an independent variable enters the model only if it explains variation in the dependent variable that is not already explained by variables in the model. Therefore, the predictability of the dependent variable is maximized while multicollinearity is minimized. Of course, predictability could be higher by including all possible independent variables in the model because the addition of any variable cannot lower the [r.sup.2]. But the possible presence of multicollinearity in this situation makes the interpretation of an independent variable problematic.
(19.) Analyzing explanatory variables in a regression model where all explanatory variables are qualitative (e.g., 0, 1) is basically equivalent to performing an ANOVA (analysis of variance). In this study, ANOVA is inappropriate because the study model contains both qualitative and quantitative variables (e.g., lot size). When a regression model has both qualitative and quantitative explanatory variables, this is basically equivalent to performing an ANCOVA (analysis of covariance). ANCOVA requires the analyst to test an important additional assumption beyond those in ANOVA. This test is described later in the article.
(20.) By eliminating observations from inner ring(s), the price effect from the inner ring(s) is also eliminated. Hence, any price effect discovered would be due solely to the difference between the houses in the subject ring and those located farther away.
(21.) The study attempted to do the same for offenders subject to limited disclosure by estimating the model after eliminating observations located within the significant range for passive notification offenders. However, because there are so many passive notification offenders, the model lost full rank for all rings. Because the effect of limited disclosure offenders should dominate the effect of passive notification offenders, we do not consider this a major problem, but there may be an interaction effect that the model is not capturing.
(22.) Variables that did not enter and remain in the model include SPRING (transactions that occurred in March, April, and May), FALL (transactions that occurred in September, October, and November), BATH2 (houses with more than one, but less than three, full bathrooms), and 22 location variables.
(23.) The holdout category to control for number of bathrooms was BATHI (where the house had one full bathroom), which was the most prevalent value in the sample. Three observations were recorded in the data as having less than one full bath. Some older houses in the sample area have less than full amenities. These three observations were included in the holdout category.
(24.) Values for this variable were obtained from the Montgomery County Treasurer's Office. If the mailing address for property tax purposes was the same as the property address, it was assumed that the buyer intended to live in the property and the variable was coded 1. If the mailing address differed from the property address, it was assumed that the buyer did not intend to live in the property and the variable was coded 0.
(25.) The holdout category to control for seasonal effects was SUMMER, which was the most prevalent value in the sample.
(26.) The tax district with the most observations in the sample was used as the holdout category for LOC. Montgomery is an extremely heterogeneous environment with neighborhoods that include extremely wealthy satellite cities, inner-city neighborhoods, new upscale suburban neighborhoods, and rural communities. The tax districts, although not perfect, divide the heterogeneous county into somewhat homogeneous subgroups. Inclusion of a variable to describe individual house condition would be preferable. This information was unavailable for the present study.
(27.) H. White, "A Heteroskedasticity-Consistent Covariance Matrix Estimator and a Direct Test for Heteroskedasticity," Econometrica 48 (1980): 817-838.
(28.) SAS/STAT[R] User's Guide: Version 6, 4th ed., vol. 2 (Cary, NC: SAS Institute, Inc., 1989). Whenever dealing with microeconomic data there is the possibility that the error terms do not follow the classical assumption of homoskedasticity (i.e., that the variance of the error term is constant for all observations). If the error terms of the equation are heteroskedastic (i.e., the variance of the error term is related to the size of a particular independent variable), ordinary least squares estimators will be inefficient (i.e., they will not have the minimum variance). Use of an incorrect functional form can result in incorrect estimators. Because the tests conducted indicated that no specific functional form was significantly better than the others, we present the linear model results for expository expedience.
(29.) David A. Belsley, Edward Kuh, and Roy E. Welch, Regression Diagnostics: Identifying Influential Data and Sources of Collinearity (New York: John Wiley and Sons, 1980).
(30.) Barbara G. Tabachnick and Linda S. Fidell, Using Multivariate Statistics (New York: Harper & Row, 1983).
(31.) Blocking variables is a recommended method to correct for a violation of constant covariates. In this process, the range of the culprit quantitative variable (e.g., AGE, which, for example, ranges from 0 to 90) is divided into subgroups and these are treated as a series of qualitative variables (e.g., AGE1 if AGE [greater than or equal to] 0 or [less than or equal to] 10, AGE2 if AGE >10 or [less than or equal to] 20, and so forth).
(32.) See for example, James E. Larsen, "Money Illusion and Residential Real Estate Transfers," Journal of Real Estate Research 4, no. 1 (1989): 13-19, and James E. Larsen, "Leading Residential Real Estate Sales Agents and Market Performance," Journal of Real Estate Research 6, no. 2 (1991): 241-249, where the study areas examined included the county analyzed in the present study.
(33.) The number of observations mentioned here (802) does not match the total number of observations shown for the first two rings in the lower portion of Table 3 because (as explained previously) observations located within 0.3 mile of an offender subject to limited disclosure were eliminated from the sample before the test reported in the lower portion of Table 3 was conducted.
James E. Larsen, PhD, is a professor of finance in the Raj Soin College of Business at Wright State University in Dayton, Ohio. Larsen has published numerous articles in journals such as The Appraisal Journal, Real Estate Appraiser and Analyst, Journal of Real Estate Research, Journal of Real Estate Practice and Education, Real Estate Appraiser, and Journal of the American Real Estate and Urban Economics Association. He is a member of the Ohio Real Estate Commission's Education and Research Fund Advisory Committee, and is author of Real Estate Principles and Practices, published by John Wiley & Sons, Inc. Contact: james.larsen@wright.edu
Kenneth J. Lowrey is a lecturer in the Department of Urban Affairs and Geography at Wright State University (WSU) in Dayton, Ohio. He teaches geography and geographic information systems (GIS). His latest research is on the topic of gated communities and has been published in Urban Geography. Lowrey has produced numerous maps for use on the WSU campus and in southwestern Ohio. He is also the WSU representative to the Ohio GIS-Net, a statewide GIS organization representing the state universities in Ohio. Contact: kenneth.lowrey@wright.edu
Joseph W. Coleman, PhD, is an associate professor of management science and information systems in the Raj Soin College of Business at Wright State University in Dayton, Ohio. Coleman has published numerous articles in journals such as The Appraisal Journal, IEEE Transactions on Reliability, and Communications in Statistics: Simulation and Computation. Contact: joseph.coleman@wright.edu ..Source.. by James E. Larsen & Kenneth J. Lowrey & Joseph W. Coleman
No comments:
Post a Comment