Playing in the Airport Zone: The Amenity Separation Effect and the Impact of Airport Proximity on Resort Property Values

1. Introduction

Airport proximity influences real estate markets by offering benefits such as improved accessibility while also presenting drawbacks like increased noise and congestion. Although the residential sector has extensively examined this trade-off (Cohen & Coughlin, 2008; McMillen, 2004; Nelson, 1980), limited research exists on vacation resort properties despite their growing economic significance. Unlike typical housing, resort properties are primarily valued for tranquility, scenic views, and recreational access rather than for commuting and accessibility convenience (Belej et al., 2023; Stringam, 2010). Recognizing this difference is essential to understanding how airport proximity impacts property values.

Vacation resort properties differ from year-round housing in ownership structure and buyer intent. Owners frequently purchase these units as second homes or short-term rental investments rather than as primary residences. The market values experiential and locational amenities more heavily than commuting convenience. Buyer motivation in this market diverges significantly from the permanent-resident context that underpins most airport-proximity research (Nelson, 1980; McMillen, 2004; Cohen & Coughlin, 2008), influencing how buyers weigh accessibility against disamenity (Bernard & Cook, 2015; Lawton et al., 2016). Therefore, a hedonic model designed for year-round residents cannot be directly applied to the resort market because the attributes valued by buyers and their relative importance differ sufficiently to require independent estimation.

In 2023, Mobile and Baldwin Counties accounted for 41.2% of Alabama’s tourism economic impact, which totaled $23.5 billion (Alabama Tourism Department, 2024). Understanding how proximity to airports influences resort values has direct implications for planning, investment, and tax policy. Regions that depend on tourism are particularly vulnerable to fluctuations in visitor demand, and seasonal variations increase the importance of accurately identifying airport-related effects for strategic planning and investment (Ridderstaat et al., 2014). Sustainable development of resort regions relies on safeguarding the amenities that initially attract visitors, making airport proximity a critical factor in planning rather than merely an academic concern (Dreizis et al., 2023). However, no hedonic study has analyzed vacation resort properties in a coastal tourism market utilizing airport-specific disaggregation and local spatial estimation.

We use a hedonic price model (HPM) and an adaptive-bandwidth geographically weighted regression (GWR) to analyze forty years of resort property sales data. This study makes four important contributions to the existing literature. First, it provides the first reliable spatial estimate of the impact of airport proximity on resort properties within a coastal tourism market. Second, it demonstrates that ordinary least squares results incorrectly indicate no proximity effect, while our spatial corrections reveal strong, multi-faceted effects. Third, it expands on research into residential property airport proximity by developing the Amenity Separation Effect, a new interpretation of the airport accessibility-disamenity trade-off, highlighting that resort properties are directly connected to transportation infrastructure even at considerable distances. Fourth, it offers practical guidance for public officials, policymakers, and planning professionals.

2. Literature Review

The literature on airport proximity consistently documents a negative residential price gradient. Nelson’s (1980) survey established foundational evidence. McMillen (2004) estimated a 9% reduction in property values near Chicago O’Hare International Airport. Jud and Winkler (2006) found a 9% decline following airport expansion announcements. Mense and Kholodilin (2014) documented losses of up to 12% beneath Hamburg flight paths. Studies on high-noise levels report discounts ranging from 20% to 23%. Nelson’s (2004) meta-analysis places the central noise discount between 0.5% and 0.6% per decibel. Additional city-level studies reinforce this pattern. Chalermpong (2010) documents comparable discounts near Bangkok’s Suvarnabhumi Airport. Trojanek et al. (2017) find similar effects in Poznan, Poland. Ozdenerol et al. (2015) link general traffic and aircraft noise to reduced housing values in Memphis. Espey and Lopez (2000) add that the discount intensifies with proximity rather than merely being within a fixed airport-influence radius.

The relationship is not uniformly negative. Cohen and Coughlin (2008), Affuso et al. (2019), and Florido-Benítez (2023) document cases in which accessibility improvements offset disamenity costs. Wassmer (2019) demonstrates that the direction of effects varies across different markets. Lipscomb (2003) finds that proximity to Atlanta’s Hartsfield International Airport increases nearby property values, providing evidence that accessibility benefits can dominate in markets with strong economic ties to air travel. Brueckner (2003) extends this reasoning beyond housing by showing that increased airline traffic stimulates broader urban economic development, which can positively influence local property markets even when noise disamenities are present. Thanos et al. (2015) complicate this understanding by revealing that the noise-price relationship is nonlinear and that stigma effects can persist even after noise levels decline.

Boes and Nüesch (2011) use a natural experiment around a runway closure to confirm that noise effects are causal rather than merely correlational, addressing a common concern in the cross-sectional hedonic literature. Accessibility benefits in nonurban areas may outweigh disamenity costs for permanent residents, resulting in a positive impact on prices, which contrasts with the expected negative proximity effect (Green, 2007; Partridge et al., 2010; Tang et al., 2022). This is logical, though, because residents with limited choices highly value those that exist. Conversely, resort buyers, who visit for shorter, more unpredictable stays, may prioritize tranquility over accessibility and lack demand for the latter. Tranquility, scenic landscapes, and recreational amenities primarily determine vacation resort property values, rather than accessibility preferences (Belej et al., 2023). Major and Lash (2003) demonstrate that proximity to beaches generally dominates resort pricing over other locational factors. The tranquility premium is not restricted to coastal settings. Sulyok et al. (2023) find a similar premium for proximity to national parks in rural vacation markets. Sztubecka et al. (2022) show that perceived environmental quality influences value even within built recreational landscapes such as spa parks. Florido-Benítez (2023) and Hao et al. (2020) caution that airports can disturb the serenity that attracts visitors to amenity destinations. Any externality that disrupts the sense of escape sought by resort buyers, including airport noise, carries a cost that conventional models may fail to fully capture.

2.1. Hedonic Studies of Vacation and Resort Properties

A smaller body of research applies hedonic methods directly to vacation and resort real estate, consistently finding that location-specific amenities dominate the pricing structure. Spahr and Sunderman (1999) demonstrate that proximity to a resort community itself commands a premium for surrounding properties, providing evidence that resort amenities radiate value outward much like an urban center does for its suburbs. Wassmer (2022) confirms that proximity to short-term vacation rentals has a distinct price effect in resort communities, separate from the structural characteristics of the home. Belej et al. (2023) apply hedonic methods specifically to vacation properties, finding that experiential attributes, in addition to structural features, explain a substantial share of price variation.

Adjacent hospitality research reinforces this pattern outside the sales market. Yang et al. (2018) find that guests weigh a hotel’s location attributes similarly to how buyers value a resort property. Caballero Galeote and García Mestanza (2020) show that tourists and residents near Málaga’s airport perceive its proximity differently, highlighting that the value of being near an airport depends on who performs the valuation. None of these studies incorporate airport proximity as an explanatory variable, a deficiency we address directly.

2.2. General Aviation as a Distinct Externality Class

A common belief is that airport externalities increase with airport operational volume (Nelson, 2004). Researchers often consider general aviation airports as minor sources of externalities. Batóg et al. (2019) disagree with this common view, arguing that general aviation airports form a policy-relevant category of their own rather than a scaled-down version of commercial aviation, and demonstrating further that general aviation airport operations affect resort properties in four significantly different ways from commercial properties. This is especially true in areas where local zoning and land-use plans have not kept pace with the growth of general aviation. Batóg et al. (2019) also disagree with the common view by demonstrating that general aviation airport operations affect resort properties in four significantly different ways from commercial properties. First, general aviation activity frequently involves repeated, low-altitude flying in various patterns around the airfield. This differs from the narrow, linear corridors typical of commercial airline flight patterns. Second, although larger jet-engine aircraft may seem noisier, smaller piston-engine propeller aircraft produce noise frequencies that penetrate more effectively through ordinary construction than modern engines do. Third, general aviation airfields typically operate during daytime and fair-weather hours, precisely when resort visitors most value their surroundings, so the timing and frequency of general aviation activity work against the resort environment. Fourth, most general aviation airports are not subject to federal noise regulations. Consequently, they have not been eligible for any of the significant federal support for airport noise mitigation and soundproofing that has benefitted many commercial service airports. Cellmer et al. (2019) and Batóg et al. (2019) further demonstrate that geographically disaggregated hedonic methods, such as Geographically Weighted Regression, are necessary to accurately capture these effects. A single global coefficient cannot reflect the sharp local variations around individual airfields. We apply this disaggregated approach for the first time to vacation resort properties rather than the single-family residential parcels examined in previous studies.

3. Theoretical Framework

3.1. Hedonic Pricing Model and GWR

Our analysis employs the Hedonic Pricing Model to decompose property values into key attributes, including proximity to airports, environmental quality, recreational amenities, and structural characteristics. Developed by Lancaster in 1966 and expanded by Rosen in 1974, the Hedonic Pricing Model is suitable for resort properties because it considers both tangible and intangible value drivers. We utilize geographically weighted regression as a complementary tool to capture local variations that the global model cannot account for, since airport impacts differ by location. Geographically weighted regression also surpasses alternative spatial techniques, such as the spatial expansion method, in accurately capturing detailed local variation without requiring the analyst to prespecify how coefficients change across space (Bitter et al., 2007).

3.2. The Amenity Separation Effect

McMillen (1997) predicts that rents increase with proximity to a transportation hub because the destination, the transportation hub, creates value. This is the amount buyers are willing to pay for greater proximity to the transportation hub. Yet resort markets operate very differently. Resort markets are places where the shoreline, the beach, the harbor, the boardwalk, the course, the vista, even the most remote wilderness are the valued amenities. Airports simply provide accessibility. The amenity and the access are spatially coincident, meaning that a visitor is indifferent between a 30-minute and a 60-minute transfer from the airport to the resort property after landing. Since the visitor is indifferent, the airport proximity accessibility benefit can only be local. We term this the Amenity Separation Effect. Because accessibility and amenity are spatially decoupled in resort markets, only the amenity itself capitalizes into price, generating two testable predictions. First, we show that resort values decline as proximity to the airport increases, though the decline is less pronounced in highly competitive coastal markets. Second, because resort buyers prioritize tranquility over mobility, we demonstrate that the Amenity Separation Effect reverses the positive nonurban general aviation accessibility premium documented in the residential literature. In other words, we confirm that an amenity that raises value for permanent residents confers no such benefit for resort buyers.

3.3. Endogeneity

Hedonic price models can have a specific bias. Property values might influence airport investment decisions as much as airport proximity affects property values (Epple, 1987; Baranzini et al., 2008). For example, cities might build or expand airports near high-value areas that are already growing, rather than the other way around. We control for environmental features, demographics, and local amenities to avoid bias from unobserved development pressure, which can distort estimates (Plantinga et al., 2002). Our geographically weighted regression adds another layer of protection by allowing coefficients to differ across locations. This helps capture unobserved local factors that a single fixed model would overlook (Nakaya, 2015).

This examination addresses two questions. The first question is whether airport externalities are reflected in resort property values. This is an important empirical question as it is a precondition for most policy and valuation contexts. The second question is whether the theoretical prediction observed in the residential literature also applies to resort properties. The significance of air traffic generated by resort buyers could have considerable implications for this question, whether resort buyers prefer airports for their associated accessibility or tranquility. Confirming or refuting these predictions directly tests whether the resort market functions like its residential counterpart or follows the distinct logic proposed by the Amenity Separation Effect.

Hypothesis 1. 

There is no relationship between airport proximity and vacation resort property values.

Rejecting Hypothesis 1 demonstrates that airport externalities are reflected in resort property values, extending the evidence previously documented only for residential (Nelson, 1980; McMillen, 2004; Cohen & Coughlin, 2008) and commercial properties (Ekeroma et al., in press) to the resort sector.

Researchers confirm that the values of general aviation airports are capitalized positively because the accessibility benefit outweighs the disamenity cost. Studies by Green (2007), Partridge et al. (2010), and Tang et al. (2022) support this. Our Amenity Separation Effect predicts that a resort buyer in a nonurban area will prefer coastal tranquility over regional accessibility. The buyer does not derive additional compensating benefits from airport proximity, even though disamenities remain present.

Hypothesis 2. 

In nonurban areas, proximity to general aviation airports has no negative impact on vacation resort property values.

Rejecting Hypothesis 2 shows that general aviation airports have an overall negative impact on resort property values, even in nonurban areas. This reverses the previously documented positive accessibility benefit (Green, 2007; Partridge et al., 2010; Tang et al., 2022) and confirms that resort buyers do not gain a compensating advantage from being close to a general aviation airport.

4. Methodology

4.1. Study Area, Airport Classification, and Sales Transaction Data

Our study area includes Mobile and Baldwin Counties, located along the coast of Alabama. The region contains nine airports (Table 1). Each airport is classified as either Urban or Nonurban by the U.S. Federal Aviation Administration (FAA). Airports that serve the metropolitan Mobile core and have an established transportation network are classified as Urban. Those that serve the lower-density coastal corridor and island communities, where transportation options are limited, are classified as Nonurban. All six Baldwin County zip codes in the resort sampling frame (36542, 36547, 36561, 36532, 36549, and 36564) are situated within the Gulf Shores, Orange Beach, Fairhope, and Foley corridor. These zip codes are served exclusively by Nonurban general aviation airports. The only Mobile County resort zip code, 36528 for Dauphin Island, is served by Jeremiah Denton Airport, which is also a Nonurban general aviation facility. Therefore, the resort sample is entirely nonurban and is served by general aviation airports. The primary Commercial Service airport, Mobile Regional, is located outside the resort sampling frame.

We obtain transaction data from the Mobile and Baldwin County Tax Assessor offices. We concentrate on small residential-style resort properties such as condominiums, vacation homes, townhomes, and villas because their transaction frequency provides a reliable basis for hedonic analysis (Bełej et al., 2020; Soler & Gemar, 2018). Our classification of airports by operational role, rather than treating all airports as functionally equivalent, follows precedent from Polish general aviation studies. As of 2022, Mobile County had over 411,000 residents and 186,000 housing units, while Baldwin County had over 246,000 residents and 132,000 housing units. Baldwin’s smaller and faster-growing population reflects the retiree and second-home in-migration that characterizes much of the Gulf Coast vacation market. Waterfront access is a primary value driver throughout this Gulf of Mexico corridor (Dahal et al., 2019). Isolating the airport effect from the beach effect is essential to a credible estimate. We reduce the initial dataset of 10,425 entries to 9,841 residential units after removing properties exceeding $1 million, 7,500 adjusted square feet, ten bedrooms, and those built before 1925. These thresholds remove luxury outliers and larger commercial properties inconsistent with the residential-style resort focus of this study. We deflate all prices to 2023 dollars using the Bureau of Labor Statistics Housing Consumer Price Index (Hillinger, 2008). We use the Claude artificial intelligence platform, version Sonnet 5, to assist in developing the literature review outline and manuscript framing (Anthropic, 2026).

4.2. Variable Specification

The dependent variable is the inflation-adjusted sale price. Airport proximity appears in the model in two forms. The Distance to Airport variable measures Euclidean distance in kilometers to the nearest airport of any category. The General Aviation and Commercial Service categorical variables identify the nearest airport’s category and quantify the price difference associated with proximity to that airport type. These categorical coefficients represent property-level price effects rather than per-kilometer value changes and capture the total capitalized value of being near an airport of that specific type. Because noise contour data are unavailable, we could not separate the effect into its noise and accessibility components. The General Aviation and Commercial Service variables are coded based on the FAA role classification of the nearest airport at the time of sale. Since no resort property’s nearest airport changed during the sample period, we treat this classification as time-invariant following previous research on general aviation hedonic models (Batóg et al., 2019; Cellmer et al., 2019).

Control variables comprise the distance to the nearest beach, measured in kilometers, together with the airport distance measure. Distance to the nearest park and distance to the nearest school were also tested but were not statistically significant and were dropped from the final specification. Distance to the nearest hospital and distance to the nearest transportation hub were statistically significant, but categorical facility-type indicators are retained in their place, since facility type captures the relevant amenity or disamenity distinction more directly than raw distance for these background infrastructure controls (Koster & Rouwendal, 2012). Structural controls include square footage, year built, counts of rooms and bathrooms, airport type, hospital type, school ranking tier, and transportation hub type.

4.3. Data Preparation

We addressed missing values in beds, baths, year built, and square footage using Multiple Imputation by Chained Equations (Raghunathan et al., 2001; Van Buuren & Groothuis-Oudshoorn, 2011; White & Carlin, 2010). We selected this method over listwise deletion because removing incomplete records could introduce systematic bias if missingness correlates with property vintage or transaction era, which vary significantly across our forty-year sample. We did not impute airport distance or category values. We employed exploratory ordinary least squares to identify statistically insignificant controls to exclude from the final estimation, including distance to parks, schools, Level 2 trauma centers, and general hospitals.

Table 2. Resort Property Characteristics.

Variable	Overall, N = 9,841	Baldwin, N = 8,865	Mobile, N = 976	p-value
	Mean (Standard Deviation)
Sales Price	205,618 (136,777)	212,793 (133,387)	140,451 (149,447)	<0.001
Sales Year	2,010 (7)	2,010 (8)	2,010 (7)	0.500
Distance: Airport	8.28 (6.15)	8.58 (6.25)	5.54 (4.34)	<0.001
Distance: Beach	7.08 (7.47)	5.79 (4.62)	18.87 (14.66)	<0.001
Distance: Hospital	9.86 (8.84)	9.00 (7.14)	17.68 (16.03)	<0.001
Distance: Park	9.22 (5.54)	9.23 (5.37)	9.15 (6.89)	0.062
Distance: School	6.36 (5.27)	6.51 (5.30)	5.02 (4.86)	<0.001
Distance: Transportation	47.13 (18.23)	49.76 (16.12)	23.25 (18.98)	<0.001
Square Feet	2,012 (978)	2,045 (984)	1,711 (870)	<0.001
Year Built	1,992 (16)	1,992 (15)	1,988 (17)	<0.001
Rooms	2.98 (0.90)	2.99 (0.89)	2.86 (0.95)	<0.001
Bathrooms	1.88 (0.74)	1.89 (0.73)	1.84 (0.80)	0.005
Airport Category	# of Resort Properties (% of Sample)			<0.001
General Aviation	9,606 / 9,841 (98%)	8,865 / 8,865 (100%)	741 / 976 (76%)
Commercial Service	235 / 9,841 (2.4%)	0 / 8,865 (0%)	235 / 976 (24%)
Hospital Type	# of Resort Properties (% of Sample)			<0.001
Birthing Center	268 / 9,841 (2.7%)	0 / 8,865 (0%)	268 / 976 (27%)
General Hospital	6,415 / 9,841 (65%)	6,221 / 8,865 (70%)	194 / 976 (20%)
Level 1 Trauma Center	73 / 9,841 (0.7%)	0 / 8,865 (0%)	73 / 976 (7.5%)
Level III Trauma Center	3,007 / 9,841 (31%)	2,633 / 8,865 (30%)	374 / 976 (38%)
Women’s Center	78 / 9,841 (0.8%)	11 / 8,865 (0.1%)	67 / 976 (6.9%)
School Ranking	# of Resort Properties (% of Sample)			<0.001
Top 1%	114 / 9,841 (1.2%)	0 / 8,865 (0%)	114 / 976 (12%)
Top 10%	6,145 / 9,841 (62%)	5,286 / 8,865 (60%)	859 / 976 (88%)
Top 5%	3,582 / 9,841 (36%)	3,579 / 8,865 (40%)	3 / 976 (0.3%)
Transportation Type				<0.001
Greyhound Bus Service	1,240 / 9,841 (13%)	388 / 8,865 (4.4%)	852 / 976 (87%)
Amtrak Train Service	8,601 / 9,841 (87%)	8,477 / 8,865 (96%)	124 / 976 (13%)
Note. Mean and Standard deviation/p-value comparing data from both counties. All distances reported in KM. Wilcoxon rank sum test; Pearson’s Chi-squared test.

Table 2. Resort Property Characteristics.

Variable	Overall, N = 9,841	Baldwin, N = 8,865	Mobile, N = 976	p-value
	Mean (Standard Deviation)
Sales Price	205,618 (136,777)	212,793 (133,387)	140,451 (149,447)	<0.001
Sales Year	2,010 (7)	2,010 (8)	2,010 (7)	0.500
Distance: Airport	8.28 (6.15)	8.58 (6.25)	5.54 (4.34)	<0.001
Distance: Beach	7.08 (7.47)	5.79 (4.62)	18.87 (14.66)	<0.001
Distance: Hospital	9.86 (8.84)	9.00 (7.14)	17.68 (16.03)	<0.001
Distance: Park	9.22 (5.54)	9.23 (5.37)	9.15 (6.89)	0.062
Distance: School	6.36 (5.27)	6.51 (5.30)	5.02 (4.86)	<0.001
Distance: Transportation	47.13 (18.23)	49.76 (16.12)	23.25 (18.98)	<0.001
Square Feet	2,012 (978)	2,045 (984)	1,711 (870)	<0.001
Year Built	1,992 (16)	1,992 (15)	1,988 (17)	<0.001
Rooms	2.98 (0.90)	2.99 (0.89)	2.86 (0.95)	<0.001
Bathrooms	1.88 (0.74)	1.89 (0.73)	1.84 (0.80)	0.005
Airport Category	# of Resort Properties (% of Sample)			<0.001
General Aviation	9,606 / 9,841 (98%)	8,865 / 8,865 (100%)	741 / 976 (76%)
Commercial Service	235 / 9,841 (2.4%)	0 / 8,865 (0%)	235 / 976 (24%)
Hospital Type	# of Resort Properties (% of Sample)			<0.001
Birthing Center	268 / 9,841 (2.7%)	0 / 8,865 (0%)	268 / 976 (27%)
General Hospital	6,415 / 9,841 (65%)	6,221 / 8,865 (70%)	194 / 976 (20%)
Level 1 Trauma Center	73 / 9,841 (0.7%)	0 / 8,865 (0%)	73 / 976 (7.5%)
Level III Trauma Center	3,007 / 9,841 (31%)	2,633 / 8,865 (30%)	374 / 976 (38%)
Women’s Center	78 / 9,841 (0.8%)	11 / 8,865 (0.1%)	67 / 976 (6.9%)
School Ranking	# of Resort Properties (% of Sample)			<0.001
Top 1%	114 / 9,841 (1.2%)	0 / 8,865 (0%)	114 / 976 (12%)
Top 10%	6,145 / 9,841 (62%)	5,286 / 8,865 (60%)	859 / 976 (88%)
Top 5%	3,582 / 9,841 (36%)	3,579 / 8,865 (40%)	3 / 976 (0.3%)
Transportation Type				<0.001
Greyhound Bus Service	1,240 / 9,841 (13%)	388 / 8,865 (4.4%)	852 / 976 (87%)
Amtrak Train Service	8,601 / 9,841 (87%)	8,477 / 8,865 (96%)	124 / 976 (13%)
Note. Mean and Standard deviation/p-value comparing data from both counties. All distances reported in KM. Wilcoxon rank sum test; Pearson’s Chi-squared test.

4.4. Geographically Weighted Regression

GWR extends the global hedonic model by allowing coefficients to vary by location (Brunsdon et al., 1996; Fotheringham et al., 2009). We estimate Equation (1):

yᵢ = β₀(uᵢ,vᵢ) + Σₖ βₖ(uᵢ,vᵢ)xᵢₖ + ϵᵢ

(1)

where coefficients βₖ(uᵢ, vᵢ) are estimated at each location (uᵢ, vᵢ) using a proximity-based kernel. An adaptive bandwidth of 1,365 nearest neighbors is utilized, selected by minimizing the corrected Akaike Information Criterion as described by Fotheringham et al. (2009) and Nakaya (2015). The adaptive specification addresses the uneven geographic distribution of resort transactions across the two counties. Global Moran’s I statistic confirms significant residual spatial autocorrelation in the ordinary least squares residuals, with I equal to 0.0935 and a p-value less than 0.000. This confirms the precondition for applying geographically weighted regression, as noted by Anselin (1995). However, the local coefficient estimates derived from geographically weighted regression raise concerns regarding multiple testing, because thousands of local regressions increase the likelihood of false positives at individual locations, as discussed by da Silva and Fotheringham (2016). Therefore, emphasis is placed on patterns that are consistent across broad areas of the study region, rather than on isolated local estimates.

5. Results

5.1. Global OLS and the Case for Local Estimation

Our OLS estimation shows that our structural variables behave as expected (Table 3). The airport distance variable, Distance Airport, yields a coefficient of -960 with a 95% confidence interval ranging from -2,716 to 796 and a p-value of 0.300. This may lead some to believe that airport proximity is not reflected in market prices. However, the geographically weighted regression results demonstrate why this conclusion is incorrect. The global estimate represents the average of two opposing relationships. Because their standard errors conceal these differences, the estimate suggests an unsupported null hypothesis. Additionally, the General Aviation dummy has a coefficient of -122,337 with a p-value less than 0.001, indicating that both results are consistent. The average linear coefficient on distance is near zero, reflecting offsetting positive and negative local sign directions. The impact of general aviation is large and consistently negative.

5.2. GWR Coefficient Distribution

Our geographically weighted regression analysis produces a coefficient matrix derived from 9,820 local regressions, with the removal of 21 units due to the absence of neighboring observations (Table 4). The geographically weighted regression model demonstrates improved fit, with an Akaike Information Criterion of 262,006.8 compared to the global ordinary least squares value of 263,473.8. The highest local R-squared values, ranging from 0.4 to 0.45, are concentrated along the southern coastal beach. This pattern aligns with the influence of beach proximity in determining coastal resort prices (Dahal et al., 2019). Inland regions exhibit lower local R-squared values, between 0.15 and 0.20.

5.3. Hypothesis Tests and Spatial Disaggregation

We test Hypothesis 1 using the entire sample, and Hypothesis 2 using the Baldwin County subset (Table 5). The Baldwin County subset is served exclusively by nonurban general aviation airports.

We reject Hypothesis 1, which asserts that there is no relationship between airport proximity and resort property values across both airport types and all geographic contexts. The overall average effect for general aviation is -$3,962 dollars, approximately 1.9% of the overall mean price of $205,618. The overall effect for primary commercial service airports is -$3,769, approximately 1.8% of the overall mean price of $205,618. All observed effects are negative.

We reject Hypothesis 2, which claims that nonurban general aviation airports do not negatively affect resort values. The Baldwin County general aviation coefficient is -$4,213, which is approximately 2% of the mean price of $212,793. This finding contradicts the previously established benefit of nonurban area residential accessibility and confirms the criteria for our Amenity Separation Effect, which asserts that resort buyers who prefer tranquility over accessibility incur a cost, rather than gain any advantage, from proximity to a general aviation airport (Green, 2007; Partridge et al., 2010; Tang et al., 2022).

5.4. Four Empirical Patterns

Four distinct patterns emerge in our results. The first pattern shows that the airport effect is consistently negative across airport types and geographic contexts, indicating that there is no apparent accessibility premium. The second pattern reveals that proximity to airports costs nearly three times as much in nonurban Baldwin County as in Mobile County, with discounts of $4,613 per kilometer compared to $1,659. This suggests that the externalities associated with noise and traffic congestion heavily impact Baldwin County, a region that relies on quiet and tranquil surroundings. The third pattern indicates that proximity zones for general aviation yield a larger discount than those for commercial service, which contradicts predictions based on operational scale. The fourth pattern is reflected in the global continuous-distance estimate, which shows a slight negative impact of -$960, suggesting that proximity has little effect. However, the localized market model presents a very different picture. The value increases by $7,992 for each additional kilometer away from the airport.

6. Discussion

6.1. Why the Resort Discount Is Smaller than the Residential Literature Reports

Our proximity zone effects, which range from 1.1% to 2.0% of the transaction price, are significantly smaller than the 9% to 23% discounts documented in the residential literature. Three reasons explain this difference. First, resort property owners do not experience airport-related noise and congestion daily. Therefore, their disutility is much less than that of residents who live there year-round. Second, the beach stands out as a major amenity. Prices decrease by $14,363 dollars for every kilometer farther from the beach, which is three and a half times the impact of airport proximity. Third, our proximity measure has a limitation. It considers all properties within a certain radius of an airport equally, regardless of whether they are directly under the flight path or miles away with no immediate exposure. This limitation tends to weaken the results, so the coefficients reported should not be viewed as upper bounds. The actual impacts are likely to be much larger. A fourth factor also influences the results. Resort transactions tend to cluster in a narrow coastal strip. Consequently, properties several kilometers from an airport are within a short absolute distance compared with the sprawling metropolitan samples that generate the 9% to 23% estimates. The smaller discount observed does not indicate a weaker underlying preference for quiet. Instead, it reflects a market where nearly every property is, in absolute terms, fairly close to an airport.

6.2. The Residential Reversal That Does Not Appear

The residential literature demonstrates that permanent nonurban residents consider airport proximity an advantage due to limited accessibility options. They find the benefits of this proximity outweigh the associated disamenity costs (Green, 2007; Partridge et al., 2010; Tang et al., 2022). However, our findings reveal the opposite for resort property owners. The Amenity Separation Effect confirms a preference for tranquility over mobility. This feature, which increases value for year-round residents, does not contribute to the value of resort properties and may even diminish it by disturbing their peaceful environment. The divergence between residential property expectations for nonurban general aviation airports and the results for resort properties represents our primary theoretical contribution. Our Amenity Separation Effect finding suggests that resort buyer motivation may have greater influence on resort property values than geographic setting alone.

6.3. Why General Aviation Exerts the Larger Discount

General aviation airports are associated with larger discounts than commercial service airports in both Mobile and Baldwin counties. This finding contradicts logical expectations. It would seem that larger, busier commercial service airports with higher traffic volume and larger aircraft would exert the greatest negative influence on property values. However, this is not observed in either urban Mobile or nonurban Baldwin. The underlying reasons for this must be examined carefully. One possible explanation involves aviation traffic patterns. Commercial airports typically operate within narrow, well-defined flight corridors. In contrast, general aviation fields often feature repetitive, circular, low-altitude flight patterns. Noise impacts also vary. Many smaller propeller planes at general aviation airports operate without federal noise standards and produce noise at frequencies that penetrate more deeply than those of larger jet aircraft. Typically, general aviation flights occur during daylight hours under favorable weather conditions. These times often coincide with periods when resort owners enjoy peaceful outdoor activities, free from airport noise disturbances. Regulatory factors may also contribute to the observed differences. Commercial airports have long been required to comply with federal noise statutes. Although these regulations increase operating costs, they also facilitate funding for land acquisition, noise easements, and sound insulation projects. In contrast, general aviation airports are not obligated to adhere to these statutes and have not received comparable funding for noise mitigation. Some of the larger discounts associated with properties near general aviation airports may result from the fact that commercial airports have benefited from years of noise mitigation efforts, whereas general aviation airports have not.

6.4. The Global Null and the Local Signal

If we only conducted a standard regression that applies a uniform approach, we would conclude that the insignificant airport proximity coefficient of -960 indicates that airport proximity does not influence prices in this resort market. However, this approach would overlook the localized spatial analysis provided by the geographically weighted regression, which demonstrates variation by location. Our robust analysis shows largely increasing price impacts ranging from $2,683 to $19,757 per kilometer farther from the airport. In some cases, the impact is opposite, reaching as much as negative $33,782 per kilometer, possibly because being farther from the airport reduces accessibility. Calculating the average impact produces a figure closer to zero, but this does not accurately reflect the true relationship. The absence of a proximity effect in the average does not imply it does not exist. It simply means that a single number cannot capture the differences that vary from one location to another.

7. Policy and Practical Implications

7.1. Ad Valorem Assessment

Counties generally assess property through mass appraisal employing common valuation modeling techniques. When assessors apply a uniform valuation, they may overlook the actual impacts related to general aviation facilities. The results demonstrate significant variation in the airport proximity effect across different locations. This issue is not about accuracy but rather about the uniformity that the profession currently expects, according to the International Association of Assessing Officers (2013). It is recommended to use Global Moran’s I to test for spatial autocorrelation in assessment-to-sale residuals prior to conducting jurisdiction-wide assessments of resort property valuation indications near general aviation airports. This procedure is important for maintaining a fair tax base and ensuring fairness because airports generate local economic activity, as noted by Green (2007). Assessors who ignore spatial variation risk overtaxing quiet, amenity-sensitive parcels and undervaluing parcels closer to economic activity.

7.2. Appraisal Practice

Because the airport’s proximity impact on resort property values varies by location, using a comparable sale from farther away can cause a hidden pricing error. Appraisers selecting comparables for resort properties should focus on nearby properties similarly affected by the airport. They should also use the location-specific GWR coefficient to justify airport-related adjustments. This practice is not limited to routine appraisals. It is essential for easement valuations, partial takings, and condemnation cases. In these situations, broad market averages lack the precision that parcel-level estimates can provide.

7.3. Noise Compatibility Planning

FAA noise statutes apply only to commercial airports (Federal Aviation Administration, 2023). Our research indicates that general aviation airports in tourism-dependent areas need similar attention. Three cost-effective actions can help reduce noise impacts. First, air traffic patterns could be adjusted to direct planes away from resort properties. Second, operators could reduce operating hours during peak tourist seasons. Third, operators could work with local flight schools to adopt less impactful altitude patterns. None of these measures require additional costs or regulatory approvals. Other airport authorities have already incorporated similar noise-routing strategies into formal environmental reviews (Josimović et al., 2016), demonstrating that these are practical, low-effort solutions rather than untested ideas.

7.4. Disclosure Policy

Pope (2008) demonstrates that mandatory noise disclosure alters the way noise is incorporated into home prices. This results in current prices reflecting incomplete information available to buyers. Alabama does not require disclosure of aviation noise, so buyers assume responsibility for assessing the information. This situation is significant because resort buyers often originate from outside the area and may visit only once, potentially during a time when flight training traffic is light. Implementing a disclosure requirement for properties within a mapped airport-influence zone, similar to existing requirements for flood zones and coastal construction, would incur minimal administrative costs. Such a requirement would address an information gap that disproportionately disadvantages the least-informed buyers. Since the estimates are based on a market without disclosure requirements, they likely underestimate the price discount that would result if comprehensive air-traffic information were available.

8. Limitations

Our study has certain limitations. Without noise contour or flight track data for the general aviation facilities in our sample, it is not possible to decompose the airport effect into distinct noise and accessibility components. Additionally, we cannot determine which parcels are located directly under active flight paths and which are merely nearby. The resort sample is primarily nonurban, so the contrast between urban and nonurban areas resulting from county disaggregation reflects differences in price levels, density, and proximity to metropolitan regions. Unobserved features of resort properties, such as floor level, water views, and on-site amenities, may account for most of the price variation not explained by our model. These features could also be correlated with proximity to airports in ways that are not measurable. From a methodological perspective, our design is cross-sectional; therefore, the results indicate associations rather than causal relationships. This distinction is important because cross-sectional hedonic coefficients address different questions than treatment effects identified through a quasi-experimental design (Angrist & Pischke, 2008). Our data sample covers multiple distinct market periods, including the 2006 peak, post-hurricane recoveries, and short-term rental restructuring after 2012. These cycles are well documented in the broader housing market literature (Burnside et al., 2016). Disaggregating these impacts remains a task for future researchers. Finally, Geographically Weighted Regression coefficients can be locally collinear even when global variance inflation factors are within acceptable ranges (Wheeler & Tiefelsdorf, 2005). This limitation is inherent to local regression, and while our adaptive bandwidth selection helps mitigate it, it cannot entirely eliminate the issue.

9. Conclusions

This study investigates how the proximity of airports influences the values of vacation resort properties along the coast of Alabama. The analysis employs hedonic pricing and geographically weighted regression techniques on 9,841 property sales spanning from 1983 to 2023. Our results support both of our proposed hypotheses. First, proximity to airports is reflected in resort property values as properties situated near airports tend to sell for less. Specifically, properties near general aviation airports experience an average discount of $3,962, while those near primary commercial airports see an average reduction of $3,769. Second, proximity to general aviation airports in the nonurban Baldwin corridor results in a decrease in prices by $4,213, accounting for approximately 2% of the average sale price. This observation contrasts with prior residential studies, which generally find that small airports in nonurban areas increase nearby home values (Green, 2007; Partridge et al., 2010; Tang et al., 2022). The difference becomes understandable when considering buyer motivation. Resort buyers seek peace and quiet rather than convenient travel. A small airport that helps year-round residents travel within nonurban areas becomes an undesirable feature for vacationers seeking a tranquil retreat. In essence, the impact of an airport on resort property value depends not only on the property’s location but also on the buyers’ motivations. We term this divergence in buyer motivation the Amenity Separation Effect, our study’s central theoretical contribution and the mechanism underlying both hypothesis tests.

Three of our findings have implications beyond the study area. First, general aviation airports provide larger price discounts compared to commercial airports, contrary to what airport size alone would suggest. This likely results from small-plane flight patterns, distinct engine noise, and a regulatory gap, where federal noise regulations do not address general aviation. Second, a standard regression across the entire market shows no impact attributable to an airport. However, our geographically weighted approach reveals that this is misleading. Local effects differ in direction across various locations and cancel out when averaged. The substantially better fit of our spatial model confirms that localized, place-by-place estimation is necessary to observe the true pattern. Third, our estimates offer researchers elsewhere a pre-change baseline for studying how resort-adjacent markets respond to airport relocations and noise-disclosure policy shifts, since our data were collected before Mobile relocated its airport and prior to the implementation of noise-disclosure requirements. Notably, commercial passenger service is moving from Mobile Regional to Aeroplex at Brookley in 2027 (The Traveler, 2026), which will reverse the current classifications of these two airports. Since both are situated within the urban core of Mobile outside our resort sampling frame, this shift does not affect the present findings. However, it will offer a natural context for future research on how residential, commercial, and resort property markets respond to changes in airport type.

Our results have direct practical value for multiple audiences. Assessors evaluate whether resort properties are fairly and uniformly valued. Appraisers select comparable sales near airports. Airport sponsors consider low-cost operational changes. Policymakers think about noise-disclosure requirements. Local planners direct land uses such as workforce housing to less amenity-sensitive locations. As tourism-dependent coastal regions work to balance airport connectivity with the natural quiet that attracts visitors and sustains their economies, the location-specific analysis demonstrated here provides a necessary tool for making those tradeoffs with clarity.

This study demonstrates that property markets driven by tourism require their own valuation logic rather than relying on borrowed assumptions from residential markets. Research on airport proximity has traditionally focused on the year-round resident, but coastal economies such as Mobile and Baldwin Counties depend on many resort buyers who play in the airport zone but choose to live and work elsewhere. As tourism demand research increasingly acknowledges that accessibility is only one of many factors influencing destination competitiveness (Dogru et al., 2021; Song et al., 2023), hedonic studies of resort real estate should treat buyer motivation as a primary variable rather than a secondary consideration.