Do Greenways Impact Proximate Residential Property Prices?

This paper is an abridged version of a thesis submitted for the award of an MSc in Quantitative Economics at University College Dublin in 2022.

Abstract

As public, non-market goods, recreational amenities such as greenways lack a comprehensive monetary estimate of value which often means that the case in support of their development is incomplete. Though the published evidence confirms the existence of a range of monetary and non-monetary benefits of these amenities, studies have been historically subjective in nature, focusing on public perception rather than hard empirical data. Using the hedonic pricing method and GIS tools, and applying semi-logarithmic regression model techniques, this paper confirms the existence of a ‘greenway effect’ for both the Great Western Greenway in Co. Mayo and Waterford Greenway in Co. Waterford, with each kilometre further from these greenways reducing the mean sale price by magnitudes of 1.22% (or €2,264 based on the mean sale price of the sample properties), and 1.35% (€2,325) respectively. These findings are consistent with published international empirical studies.

Introduction

Ireland’s network of greenways—dedicated corridors for walking and cycling—has expanded substantially over the past decade, galvanised by national strategy and strong local demand. These routes are designed to improve quality of life, safely connect places for active travel, and enhance environmental amenities. Two flagships illustrate the trend: the Great Western Greenway in Co. Mayo, opened in 2011 along a former railway line from Westport to Achill, and the Waterford Greenway, launched in 2017 between Waterford City and Dungarvan. Policy attention and research output have both risen sharply over this period, reflecting the practical importance of understanding the benefits and costs of greenways. The potential benefits of greenways are well cited e.g. health, connectivity, tourism, ecological value. However, public investment and political decision-making generally often hinge on monetary evidence. Cost–benefit analysis struggles when amenities are non-market goods without explicit prices. One credible way to monetise value is to study how nearby housing markets capitalise the presence of the amenity into property prices. If purchasers are willing to pay more to live near a greenway, a “greenway effect” should appear, controlling for property characteristics and other location factors. To identify this effect, this paper therefore asks two related questions. First, does proximity to a greenway measurably affect residential sale prices? Second, what is the magnitude of that effect, and how does it compare with estimates reported internationally?

Literature Review

Literature on the value of open space and trail amenities has evolved notably over the past four decades. Early contributions relied heavily on perception studies i.e. surveys of homeowners, real-estate agents, and community groups near trails and parks. These studies catalogue concerns about privacy, noise, litter, and anti-social behaviour, while also registering perceived benefits such as recreational access, aesthetic improvements, and a general uplift in neighbourhood prestige. Perception evidence, while instructive for planning, was inevitably subjective and sometimes contradictory across locations. As data availability improved, researchers increasingly turned to behaviour-based approaches, especially hedonic pricing. The hedonic method treats the property as a bundle of attributes—structural, locational, environmental—and regresses observed transaction prices on those attributes to infer implicit prices for features that do not have stand-alone markets. In this framework, the marginal willingness to pay for a greenway is revealed by how prices vary with distance to the amenity, once other determinants are controlled for.

As shown in Table 1, a large body of empirical work now reports that trails and greenways tend to increase nearby residential values or, at minimum, do not depress them. Depending on context, estimated premiums typically fall in the low single digits, with credible studies reporting 2–5% uplifts for single-family homes in urban or suburban settings.

Table 1: A Selection of Empirical Studies in the Literature Exploring the Effect of Greenway Proximity on Residential Property Values. Source: Author

Nevertheless, estimation is not trivial. Amenities often co-locate with transport corridors, waterways, or scenic areas; if greenways parallel a primary road, distances to each may be highly correlated. Some environmental or neighbourhood attributes e.g. school performance, small-area income, or crime are hard to measure at appropriate granularities, risking omitted variable bias. Researchers therefore place emphasis on parsimony, diagnostics for heteroscedasticity and multicollinearity, functional-form checks, and clear reporting of limitations.

Empirical Strategy, Data and Methods

This study examines how different features of a housing unit affect its price. It expresses the final sale price of a home as a bundle of individually priced structural, locational, environmental and time characteristics e.g. size, number of bedrooms, age, location, and nearby amenities such as parks or waterfronts. The model estimates how much each of these features contributes to the final sale price. To ensure that the results are reliable, the study assumes the housing market works in a reasonably competitive and informed way but there are always risks since elements that are not easily quantifiable e.g. school quality or crime level are excluded. The best model is one that fits the data best.

The analysis draws on 2019 residential sales from the Residential Property Price Register (RPPR). To ensure comparability across the year, each transaction is adjusted to a common December 2019 price level using the Central Statistics Office’s Residential Property Price Index (National excluding Dublin), assuming no material property alterations between sale and the reference date. Property addresses are manually geocoded to obtain coordinates and Eircodes, enabling precise calculation of straight-line distances to the greenway alignment and the coastline, and driving distances to primary roads and the largest adjacent urban centres.

A key design choice concerns the definition of proximity. A 15-kilometre catchment is used for each greenway. While some studies adopt narrow bands (e.g., 0.5 miles), shrinking the radius here would sharply reduce sample sizes and jeopardise model stability—particularly in Mayo, where settlement is more dispersed. The chosen radius captures adjacent and accessible neighbourhoods that plausibly use the amenity for recreation or commuting.

The dependent variable is the log of the adjusted sale price. Structural regressors include floor area, lot size, bedrooms, age, a new-build dummy, and a house (versus apartment) dummy. Locational regressors comprise straight-line distance to greenway, driving distance to the largest nearby town (Westport in Mayo and Waterford City or Dungarvan in Waterford), and driving distance to the nearest national / primary route. The environmental control is straight-line distance to the sea or waterfront. A time dummy indicates whether the sale executed in the first half of 2019. Estimation proceeds from a reduced model—with one representative from each attribute group—to progressively richer specifications, with diagnostics and AIC guiding the final model choice.

Case Study 1: Great Western Greenway (Mayo)

Sales data records 1,316 residential property sales in County Mayo in 2019. Excluding out-of-catchment addresses, unlocatable records, and duplicates, this leaves 151 geocoded properties spanning Westport, Newport, Mulranny, Achill, and nearby rural areas.

Figure 1: Geolocated Residential Property Sales Proximate to GWG. Source; Author, applying geocoded data to QGIS and Googlemaps

In the proximate greenway area, the mean characteristics include a price of €185,551 (December 2019 basis), floor area of 136m², lot size of 2,080m², 3.2 bedrooms, average age of 36 years, 2.97 km from the greenway, 14.1 km driving distance to Westport, 5.3 km to a primary road, and 1.45 km to the waterfront. Most properties are second-hand (96%) and houses (81%) and 53% were sold in the first half of the year.

Greenway alignment closely follows the N59/N5 for extended stretches, making distances to the greenway and to the primary road highly correlated. Distance to Westport also moves with these as a large share of sales cluster near the town. Structural variables behave as expected with bedrooms and floor area strongly correlated to price. Once statistical adjustments are made to correctly specify the functional form of the model, the following regression and robustness results are as follows.

Table 2: Final Regression and Robustness Testing Results – Case Study 1 (Mayo) *** p<0.01, ** p<0.05, * p<0.1

The preferred specification (model 5)[i] is read as follows: 

• SIZE: Each additional 1m² in property size increases the property sale price by 0.13%. This result is statistically significant at the 5% level.
• LOTSIZE: The positive coefficient confirms that that bigger lot sizes are associated with higher sale prices. However, the result is not statistically significant. 
• BEDS: Each additional bedroom increases the property sale price by 15.6%. This result is significant at the 1% level. This variable is a strong predictor of sale price, an intuitive and common result in the literature.
• UNIT: The property being a house increases the sale price by 38.8%, which is statistically significant at the 1% level.
• GRNWAY: For each 1km increase in distance from the Greenway, the sale price reduces by 1.22%. This result is statistically significant at the 5% level.
• DISTTOWN: For each 1km increase in driving distance to Westport (a major town), the sale price reduces by 0.9%. This result is statistically significant at the 5% level.
• AGE and TIME: While an increase in the age of a property has an expected negative coefficient and therefore negative impact on sale price, the result is not statistically significant at conventional levels in any models. However, the proximate area did not include many properties of advanced age since most properties sold in 2019 were under 50 years old. The result is similarly the case for TIME. 

The results are intuitive. Property prices are higher where a property is bigger, has more bedrooms, or is a house rather than an apartment. Homes closer to Westport and newer homes also tend to sell for higher prices, although age isn’t a significant factor in Mayo. For every extra kilometre a property is from the greenway, its price falls by about 1.22%. That works out to roughly €2,264 less per kilometre for an otherwise similar home. Distance from the waterfront also has a positive effect on price, but once other factors are included, this effect is not statistically significant. This is reasonable because the area includes a mix of coastal and inland homes, and many coastal features overlap with other desirable amenities.

Case Study 2: Waterford Greenway (Waterford)

The Waterford catchment produces a much larger sample. From 1,580 county sales in 2019, applying the same exclusions to case study 1 yields 757 geocoded observations within the catchment area covering Waterford City, Kilmacthomas, Dungarvan and surrounding rural localities. Average traits include a price of €172,229, floor area of 131m², lot size of approximately 535m², 3.26 bedrooms, mean age of 31 years, 2.39 km from the greenway, 2.4 km driving distance to the nearest major town (Waterford or Dungarvan), 5.3 km to a primary route, and 1.42 km to the waterfront. Second-hand properties comprise 80% of sales, houses 85% and 53% of sales executed in the first half of the year.

Figure 2: Geolocated Residential Property Sales Adjacent to the Waterford Greenway. Source: Author, applying geocoded data to QGIS and Googlemaps.

Correlations show familiar structural relationships and a weaker collinearity profile among locational distances than Mayo. Because the Waterford Greenway does not shadow the primary route as closely along its entire length, the correlation between distances to the greenway and to a primary road is materially lower; the correlation between town distance and road distance is also muted given the twin-centre geography.

Table 3: Final Regression and Robustness Testing Results – Case Study 2 (Waterford) *** p<0.01, ** p<0.05, * p<0.1

The preferred specification (model 5)[ii] is read as follows: 

• SIZE: Each additional 1m² in property size, increases the property sale price by 0.36%. This result is statistically significant at the 1% level.
• LOTSIZE: The positive coefficient confirms that that bigger lot sizes are associated with higher sale prices. This result is statistically significant at the 10% level.
• BEDS: Each additional bedroom increases the property sale price by 17.1%. This result is significant at the 1% level. This regressor is a strong predictor, a common result in the literature. This result is statistically significant at the 1% level.
• AGE: Each additional year in age reduces the sale price by 0.274%. This result is significant at the 1% level.
• UNIT: The property being a house increases the sale price by 10.6%. This result is statistically significant at the 5% level.
• DISTTOWN: A 1-kilometre increase in driving distance to either Waterford or Dungarvan (both major towns/a major city) increases the sale price by 0.6%.
• DISTWATER: A 1-kilometre increase in distance to the sea or waterfront reduces the sale price by 2.95%. This result is statistically significant at the 1% level.
• GRNWAY: A 1-kilometre increase in distance from the greenway reduces the sale price by 1.35%. This result is statistically significant at the 5% level.
• TIME:  The timing of the sale matters, unlike in case study 1. A sale taking place in the other half of the year increases the sale price by 5.2%.

As with the GWG, bigger homes and those with more bedrooms in Waterford sell for higher prices. Houses also sell for more than apartments. Older properties tend to be cheaper, and this effect is clearly visible in Waterford City. Homes farther from the waterfront are worth less. Prices drop by about 2.95% for every extra kilometre from the sea. The greenway also matters. Each additional kilometre from the Waterford Greenway is linked to a 1.35% lower sale price or about €2,325 on average.

Discussion and Limitations

The two independently estimated semi-elasticities −1.22% (Mayo) and −1.35% (Waterford) per kilometre—are remarkably close in magnitude despite different geographies, market contexts, and network maturities. Both sit comfortably within the range reported by comparable empirical studies of trail and open-space amenities. As these coefficients are marginal per-kilometre effects, the total proximity premium depends on the local distribution of distances. All else being equal, a property one kilometre from a greenway will transact at a meaningful premium to an otherwise similar property ten kilometres away.

Across both areas, structural characteristics are powerful predictors of price. In Waterford, age discounts and waterfront proximity further shape price variation, reflecting the coastal urban–rural mix. In Mayo, larger rural lots and more dispersed settlement mean lot size is less precisely estimated and waterfront proximity is less informative after other controls. A central identification challenge is the co-location of features. In Mayo, the greenway parallels the N59/N5 for long segments, making distances to the greenway and to the primary road highly collinear. If primary roads have their own influence (positive via accessibility, negative via noise and pollution), including both distances can inflate standard errors and obscure attribution. The modelling strategy therefore drops the primary-road distance in the preferred specifications. Interpreting the greenway coefficient, the negative semi-elasticity on distance is best viewed as a capitalised accessibility premium. Closer properties offer lower travel costs (time and convenience) to greenway recreation and active travel routes, improved perceived health and wellbeing opportunities, safer cycling corridors, and, in some settings, aesthetic value from green open space. Households pay for these attributes in the housing market. As with any local public good, value likely varies with perceived quality (surface, width, signage, ancillary facilities), maintenance, and integration into the wider network.

Several limitations are salient to this study including a lack of reliable geocoded and property-specific data and omitted attributes such as school performance and crime. Further studies would benefit from richer geocoded administrative data and systematic reporting of property attributes to improve area-specific evaluations and to facilitate more nuanced spatial modelling.

Policy Implications and Conclusions

This paper confirms that a positive ‘greenway effect’ exists that can (cautiously) be used as an input to broader cost–benefit appraisals alongside tourism, active travel, health, and environmental co-benefits. This finding may also alleviate community concerns during planning and route alignment discussions, especially in rural areas where scepticism is often most acute. Finer-grained analysis by housing type or tenure would help identify distributional effects and inform targeted mitigation for any localised disamenities e.g., parking pressure at trailheads or peak-period congestion. Beyond this headline effect, the analysis demonstrates the feasibility of area-specific hedonic appraisal using publicly available data, while also spotlighting significant data gaps. Improving the availability and quality of geocoded, property-level attributes would enable more granular and accurate local assessments, help isolate amenity impacts from co-located infrastructures, and strengthen the evidence base for investment decisions.

References

Asabere, P., and Huffman, F. (2009) ‘The Relative Impacts of Trails and Greenbelts on     Home Price’. The Journal of Real Estate Finance and Economics, Vol. 38, 408-419.

Campbell, H., and Munroe, D. (2007) ‘Greenways and Greenbacks: The Impact of the Catawba Regional Trail on Property Values in Charlotte, North Carolina’. South-eastern Geographer, Vol. 47, No. 1.

Correll, M., Lillydahl, J. H and Singell, L. (1978) ‘The Effects of Greenbelts on Residential Property Values: Some Findings on the Political Economy of Open Space’. Land Economics. Vol. 54, No. 2, pp. 207-217.

Crompton, JL. and Nicholls, S. (2005) ‘The Impact of Greenways on Property Values: Evidence from Austin, Texas’. Journal of Leisure Research. Vol. 37, Issue 3, pp. 321-341.

Crompton, JL. and Nicholls, S. (2019) ‘The Impact of Greenways and Trails on Proximate Property Values: An Updated Review’. Journal of Park and Recreation Administration. Vol. 37, Issue 3.

Hunter, R., Dallat, M., Tully, M., Heron, L., O’Neill, C., and Kee, F. (2020) ‘Social return on investment analysis of an urban greenway’. Cities & Health, DOI.

Lindsey, G., Man, G., Payton, S., and Dickson, K. (2004) ‘Property values, recreation values, and urban greenways’. Journal of Park and Recreation Administration, 22(3), 69–90.

Racca, D., and Dhanju, A. (2006) ‘Project Report for Property Value/Desirability Effects of Bike Paths Adjacent to Residential Areas’. Center for Applied Demography & Survey Research.

Youngre, N. (2019) ‘Does converting abandoned railways to Greenways impact neighbouring housing prices?’ Landscape and Urban Planning. Vol. 183, 157-166.

[i] 

[ii]

Cover Image: Along the Great Western Greenway, County Mayo | Landscape and Travel | Flickr. The paper’s author adapted the image slightly. License information here.