Home » 2014 Categories » 2014 Durham Paper » Northgate Mall’s Effect on Surrounding Property Values

Northgate Mall’s Effect on Surrounding Property Values

By James Seago   Northgate Mall’s Effect on Surrounding Property Values

I. Introduction & Motivation

Over the course of the last few decades economists and scholars have produced a significant amount of research on the various factors influencing the value of residential properties. Major determinants of property values include the physical characteristics of a property, the environmental and amenity attributes, the financial conditions of the sale and, most importantly, the location of a property. Homebuyers will consider many different facets of a property when determining the price they’d be willing to pay for their new home. The number of bathrooms, number of bedrooms, size of the lot, square footage of the property and additional amenities are all essential components that factor into their decision. However, as the saying goes, only three things matter in the real estate industry: location, location, location. This is the homebuyer’s most important decision.

The effects of certain locational determinants, such as proximity to public transport, highways, schools and churches, on the value of property are well documented. However, despite the acknowledgement of the role that nearby commercial land uses have on pricing homes, current studies and literature on the topic suggest that there is a clear lack of consensus on whether the externalities created by commercial land developments negatively or positively influence surrounding property values. Shopping centers, in particular, represent a very unique influence in the impact they have on housing prices. The benefits associated with close proximity to a variety of retail stores and restaurants, are arguably offset by increased levels of traffic, noise pollution and crime. These attributes have led to many economists and real estate professionals debating the true bearing that shopping centers can have on local neighborhood housing prices.

In this paper I will introduce and examine my findings on how the prices of residential properties are affected by their distance from Northgate Mall. Through my research I hope to provide a broader understanding of the direct influence a neighborhood shopping center can have on property values in the surrounding area. The conclusion of this paper will then aim to compare these results to those of similar studies that have previously been conducted.

II. Literature Review

Despite the large quantity of studies and literature that cover the effect of commercial land use on the value of neighboring properties, there is a distinct division of opinion on the conclusions that have been made. Stull (1995), for example, found a quadratic relationship between home values and the amount of commercial development in an overall residential area. The studies’ results suggested that small quantities of commercial development in a local area would have a positive impact on home values. However, once commercial development exceeded 5% of the total neighborhood land, property prices would begin to experience a substantial decline. A more recent study by Song and Knaap (2004) drew similar conclusions, showing that commercial development had no negative effect on the property values that they had assessed. The findings showed that housing prices increased as their distance from neighborhood commercial land uses shortened. Furthermore, homeowners that lived within walking distance from the commercial development were likely to pay an additional premium due to improved accessibility. Despite these results, the paper does conclude that the size of a particular commercial development can have powerful effects on neighboring home values and that larger commercial developments are more likely to create a negative impact.

On the other hand, a study by Grether and Mieszkowski (1978) indicated that there is a small positive correlation between housing prices and distance from industrial activity and public housing zones. The findings are based on a hedonic pricing model that is used to evaluate the effect that proximity to industrial land uses has upon home values.

The aforementioned empirical research exhibits differing results on the impact of commercial developments on property prices because in these cases, and in many others, the research has failed to recognize the tremendously localized character of the effect. Colwell, Gujral and Coley (1985) showed, in their comparison of how house prices in a neighborhood changed from before and after the introduction of the local shopping center, that property values within 1500 feet of the development decreased as proximity increased. However, after this critical distance was reached houses displayed an increase in value the closer they were situated to shopping and other development amenities. Li and Brown (1980) also explored the idea that commercial developments can have both positive and negative consequences on the surrounding region. The study assessed both the potential negative effects generated on housing prices, as a result of the aesthetics and noise pollution created by the commercial development, and also the positive influence of “accessibility” to the shopping center. Closer proximity to the industrial land improves access to shopping, various developmental amenities and to work places for homeowners. The empirical research that they conducted suggested that homes within 1760 feet (one third of a mile) of the commercial development diminish in value the closer they are to the development site. Similar to the studies produced by Colwell, Gujral and Coley (1985) once this 1760 feet threshold is passed, residential homes begin to command higher prices the closer they are to the development. The conclusions of the paper add that the positive impact created by the “accessibility effect” outweighs the effects that are realized by the negative externalities.

Aydin, Crawford and Smith (2011) sought to expand on these findings and applied them to a large commercial development in the Town Center Improvement District in Montgomery County. The primary goal of their research was to assess the negative externalities incurred by residential properties that would be generated through “Commercial Development Spillover”. Once again the results from this study, despite a much larger commercial development and a distinctly different location, shared many commonalities. The research demonstrated that any negative impacts that were generated from commercial developments were limited to areas of very close proximity. The increased size of the commercial property, compared to those from the aforementioned literature, resulted in an increased radius of negative spillover effects (approximately one mile past the TCID’s boundary). However, this increased size also led to a far greater rise in positive externalities past this point and thus these positive externalities far outweighed the negative.

The objective of this paper will be to assess whether similar patterns arise in areas surrounding Northgate Mall through the use of a hedonic regression model that shares many characteristics to the model implemented by Aydin, Crawford and Smith (2011).

III. Data  & Methodology

This paper utilizes a hedonic regression model to interpret the effect of Northgate Mall on the value of nearby housing. Hedonic regression provides the most apt method to statistically estimate the relationship between the market value of a property and the property’s characteristics, including distance from the shopping center. The model was comprised of the following features that play the most significant role in determining the value of a property: Lot area (Loti), Square Footage of property (SFi), number of bedrooms (Bedi), number of bathrooms (Bathi) and the distance from Northgate Mall is split up into five segments (see appendix) that are represented by five dummy variables in the model (D1i, D2i… D5i).

The paper categorizes 250 different houses that have been sold since 2012 into 5 different distance segments.1 These segments represent a portion of land that is situated within certain radius (0.5 miles, 1.0 miles, 2.0 miles, 3.0 miles and 4.0 miles) from the shopping center. The data for these different property characteristics were accumulated from Zillow.com2 and individually collected based on whether homes were placed within the required segment. A sample size of 50 different houses was chosen from each geographic segment to represent the overall population of houses from within that area. Collection of data ensured that a minimal of 12 different properties were selected from each the North, South, East and West to account for any anomalous results or other confounding factors that could influence housing prices. For example segment (D3) contained properties in the Southeast that are in close vicinity to Brightleaf Square and 9th Street. Once the data was successfully collected OLS regressions were run with the log of housing price as the response variable. The three models that were created are shown below. The first model is the most standard model, the second seeks to improve the fit of the model with the incorporation of square footage squared and the third assesses the size of the lot squared:

Screen Shot 2014-04-09 at 12.34.35 AM

IV. Empirical Review

The results of the regression are shown in Table 1 (see appendix). The coefficients of the number of bedrooms, numbers of bathrooms, housing size and lot size are predictably all positive. Irrespective of a property’s distance from Northgate Mall one would anticipate that an increase in any of the aforementioned factors would lead to a positive rise in the value of home. Houses that were situated within a 0.5-mile radius of the shopping center appeared to suffer from negative externalities as predicted in our hypothesis. The coefficient displayed by D1 under the first model was -.05, which indicates that houses suffer a small drop in prices when within a 0.5-mile radius from Northgate Mall. One limitation of the model is the small sample size that has been selected to represent a much wider set of homes and this is a potential explanation for the low t-statistic demonstrated by D1.

The most interesting findings from running these models is that this data has shown almost identical patterns to the results shown by both Aydin and Crowel. D2 (segment of 1.0-mile radius from Northgate) has the strongest positive coefficient of all variables from the model. A coefficient of 0.25 and strong t-statistic (3.04) suggests that properties within this area experience a strong increase in price as a consequence of their distance from the commercial development. The coefficients of D3 and D4, 0.155 and 0.002 respectively, further give credence to the results produced by Aydin and Crowel. The coefficients demonstrate that house prices in both areas are still positively impacted by their relative proximity to Northgate Mall, however once the critical point is reached (in our case 0.5-mile radius) property prices begin to fall the further one moves away from the commercial development.

V. Conclusions

There are several limitations that constrict the true power of the results provided by the research done in this paper. With more time the empirical research would have ideally included a much wider array of datasets and would have attempted to differentiate certain house prices based on other confounding factors such as additional areas of influence.

Despite the limitations of the model the results of this research provide further backing for the proposed theory that negative externalities of commercial developments are realized in a very localized surrounding area. Once this area is passed properties experience positive changes to their prices and their values increase the closer they are to, in this case, the shopping center. The fact that these results are similar across a diverse range of commercial developments and different cities suggests that this theory has very powerful implications.

VI. Works Cited

Aydin, Recai, Evert Crawford, and Barton A. Smith. “Commercial Development Spillover Effects Upon Residential Values.” Southwestern Economic Review 37 (2011): 47–62.

Colwell, Peter, Surinder Gurjral and Christopher A. Coley. 1985. “The Impact of a Shopping Center on the Value of Surrounding Properties.” Real Estate Issues 10 (1): 35-39

Grether, David M. and Peter Mieszkowski. 1980. “The Effects of Non-residential Land Uses on the Prices of Adjacent Housing: Some Estimates of Proximity Effects.” Journal of Urban Economics 8 (1): 1-15.

Li, Mingche M. and James H. Brown. 1980. “Micro-Neighborhood Externalities and Hedonic Housing Prices.” Land Economics 56 (2): 125-141.

Song, Yan and Gerrit-Jan Knaap. 2004. “Measuring the Effects of Mixed Land Uses on Housing Values.” Regional Science and Urban Economics 34 (6): 663-680.

Stull, William. 1975. “Community Environment, Zoning, and the Market Value of Single- Family Homes.” The Journal of Law and Economics 18 (2): 535-557.

Appendix I 

Table 1

Screen Shot 2014-04-09 at 12.36.22 AM


Appendix II 

Screen Shot 2014-04-09 at 12.37.23 AM


  1. James,

    The empirical section of your paper was very strong. I can see you put a lot of effort and thought into it. Manually collecting 250 data points is certainly no easy task. I have listed below some suggestions as to how you could improve this paper:

    1. In the literature review, there is extensive discussion about the critical threshold distance that separates a positive or negative impact on property values. If you could insert a theoretical model that explains this phenomenon, it could be helpful for the reader.

    2. Perhaps you could replicate the data analysis to the area surrounding South Point Mall? From my experiences, Northgate Mall primarily attracts a low-income clientele. The facilities and store options at Northgate reflect this. I guess what I’m trying to say is that Northgate is a unique type of commercial development property, and not representative of the whole spectrum. By incorporating and extending your study to South Point Mall (a more luxurious and high-end mall) and the surrounding area, you can address this issue.

  2. It is clear that you put a lot of work into analysing this data and reaching the conclusions you did. Your literature review was very interesting and clearly demonstrated why this topic was of interest to you. There is clearly not an easy answer to this question and therefore it is necessary to evaluate on a case by case basis.

    I agree with Bryan’s comment above, that a comparison using the same techniques with Southpoint Mall would be interesting. Of course, you can always add more variables to your regression to further your analysis too.

  3. James,

    I found the focus of your paper to be very interesting. I wrote my term paper on the effect of professional sports stadiums on housing prices. We often assume these complexes to provide a positive economic benefit, but it’s clear that that’s not always the case. The effect that commercial development has on the price of residential property is clearly one that homebuyers ought to consider more closely.

    The contrast in impacts between the accessibility effect and negative externalities from the presence of Northgate Mall is interesting. I’m intrigued by the threshold point that is reached at which property values begin to actually rise as homes locate closer to the Shopping Mall. Much like in the Aydin and Crowel paper did you investigate any further into the exact point at which this trend begins to take place? Since this point is likely, based on your results, to lie within the 0.5 mile radius I would recommend expanding your analysis by reducing the distance between the radii of the distance segments in order to provide a more accurate representation of the impact of property valuation closer to the site. The current distance between the segments appears slightly far apart and may have led to the analysis including homes that are closely linked to other factors of influence in the area.

    Also, given the types of customers that regularly visit Northgate Mall (mid – low income) I think a further analysis of the socio economic status and type of demographics of those in the area may play a strong role in the analysis that was conducted. This could then be compared to the effects of a higher end mall, such as Southpoint, can have on the surrounding area. Perhaps the negative externalities or the accessibility effect is more pronounced depending on the size or perceived quality of the commercial development in question.

    -Billy Marsden

  4. James,

    I enjoyed reading your paper, as having lived two blocks away from Northgate Mall this past academic year it was quite relevant on a personal level. Specifically, I found your exploration of both sides of the division in opinion on the effect commercial development has on housing values to be quite refreshing, since papers often only selectively incorporate past literature that buttresses their theses.

    Additionally, I thought your empirical approach was quite robust. However, I have two suggestions on ways to improve the paper. First, I believe there is a more salient approach to account for confounding variable than utilizing twelve houses in each direction. I think a random selection of houses by radius, combined with an inclusion of pertinent explanatory variables such as local school performance or crime rate, would yield better coefficient estimates. Also, I think a reduction of the first few radius distances would help in determining the critical distance where commercial development begins to bolster home prices. Colwell, Gujral and Coley (1985) and Li and Brown (1980), two papers mentioned in the Literature Review, found that the threshold was 1500 ft. and 1760 ft., both significantly less than the smallest radius distance employed in your paper, .5 miles (2640 ft.).

    – Jay Bishen

  5. James,

    It is clear that you put a lot of research and effort into formulating your paper. Your analysis was well structured. The previous research in you conducted prior to writing your paper was clearly thorough as well as helpful to your study. As Jay said in his comment above, I too lived closed to the Northgate Mall this past year, and this analysis was particularly interesting to me.

    If you were to expand on this topic, an interesting comparison you can make is how property surrounding a newly constructed commercial development is effected. My assumption is that the benefits of living very close to a newly constructed commercial development are high. However, I think it would be interesting to find the point at which the benefits of proximity begin to follow the model that you analyzed.

  6. This paper is well articulated and presents a convincing argument. Although your statistical model is useful, I have concerns about the appropriateness of your data. My first criticism is your assertion that 0.5 miles is the critical point at which the rate of the mall benefits home prices declines. Your dataset groups houses based on the distance to the mall, but this distinction is artificial. Although proximity is certainly relevant to the benefit home prices enjoy, there are other factors that could affect the impact of the mall on prices. For example, the travel infrastructure (i.e. how different areas travel differently to the mall) could make the distance a home close to the mall travels greater than a home that is farther from the mall in a linear sense.
    After reading the paper, I am curious what the effects the mall had directly after it was built. This local amenity shock could be revealing. Furthermore, I am unsure whether the mall’s decision to establish itself at that location can be ignored in your model. It is possible there are location factors which induced the mall to locate there and also drove up home prices close to the location.
    -Garrett Kingman

  7. Dear James,

    I applaud you for collecting your own 250 individual data points. I wrote my Durham paper on proximity effects as well and although I started by collecting data points from Zillow, the process proved tedious and was, therefore, discontinued. Based on the data resources that you had, I think the hedonic regression adequately estimated the relationship between property value and the property’s characteristics, including proximity to Northgate so the following are just suggested improvements that I think can be made to your approach.

    Firstly, I think you should have measured your radii differently. Residents would resonably be willing to walk up to 2 miles but would most like start driving for any distance after that, thus removing all positive proximity externalities. I think you should have categorized your max distance as >2.0miles and disincluded 3.0 and 4.0 miles. I also think you should have categorized your distances in smaller increments, especially closer to the mall, so as to find the optimal distance or the threshold between being too close to a mall (and thus, receiving negative externalities) and being too far to receive the benefits of close proximity. This value, as proven by your research, is in between 0.5 miles and 1.0 miles away.

    Secondly, I do not know how much Model 2, which incorporated square footage squared, and Model 3, which incorporated size of the lot squared, actually contribute to your model (however, I do not know the adjusted R squared values and will not be quick to pass judgement). I think it would be beneficial to instead focus your extension on the aspects of the malls itself such as the number or quality of the tenants, if you could control for it. The mall is depreciating and has been losing tenants for the past 5 years. I think it would be very interesting to see how prices have appreciated/depreciated in the last year (since you published your report). Perhaps you can examine how property values have changed over a 5-year period controlling for factors like the quality of tenants and the good-service ratio. It might be worthwhile to create a mall-wide or tenant-specific cash flow analysis, perhaps over a 5-year, 10-year, or 30-year period and integrate this into your study. As a general note, it’s interesting to see that the mall landed new tenants in 2011, however, vacancy rates have continued increasing and additional tenants have been more activity based and targeted towards low-income sectors (Cosmic Glow Golf and Save More, a furniture discount store). It would be interesting to see how these tenants, and different anchor department stores (Sears, Belk, Macy’s, etc) affect property prices. Another thing you may want to control for is the incidence of crime, which has been quite volatile for Northgate, in particular. Central areas like large malls spawn gathering areas and thus these effects on property values must be factored in as well.

    If it was turned into a thesis, I think that the profound growth of Walltown and should be factored in because that area has experienced more exponential growth that was not controlled for. The effects of the closure of South Square Mall in Durham may also be examined.

    Overall, great job and it was a pleasure to read your paper.

Leave a comment

Your email address will not be published. Required fields are marked *