1# RECORD c8acd4a142f8c89ba5c53a61a81bfd91957160a62# TITLE: Do people value farmers markets: A spatial hedonic pricing model approach3# YEAR: 2024.0 VENUE: Agricultural and Resource Economics Review DOI: 10.1017/age.2024.14# AUTHORS: Yanan Zheng; Meng Yang; Henry An; Feng Qiu5# PAGES: 21 CANDIDATE TABLE PAGES (1-based): [9, 11, 12, 13, 14, 15, 17]6# ABSTRACT: 7 Farmers markets can generate positive externalities by improving food access and negative externalities through pollution. The presence of both may influence people’s willingness to pay (WTP) for living nearby. This study employs spatial hedonic pricing models to estimate the WTP for living near farmers markets in Edmonton, Canada. We find an inverted U-shaped relationship between proximity to a farmers market and property values. Our results suggest that local governments might consider the economic impact of building new or relocating existing farmers markets on residential housing values, alongside the benefits of improved access to high-quality food sources.8910===== PAGE 1 (score 5, numeric-paren 0, table-words 0, var-words 5) =====1112RESEARCH ARTICLE13Do people value farmers markets: A spatial14hedonic pricing model approach15Yanan Zheng1, Meng Yang216, Henry An3 and Feng Qiu3171Department of Agricultural and Resource Economics, University of California, Davis, Davis, CA, USA,182School of Public Administration, Zhongnan University of Economics and Law, Wuhan, Hubei, China and193Department of Resource Economics and Environmental Sociology, University of Alberta, Edmonton, AB,20Canada21Corresponding author: Meng Yang; Email: yangmeng@zuel.edu.cn22(Received 24 May 2023; revised 25 October 2023; accepted 29 December 2023; first published online 14 February232024)24Abstract25Farmers markets can generate positive externalities by improving food access and negative26externalities through pollution. The presence of both may influence people’s willingness27to pay (WTP) for living nearby. This study employs spatial hedonic pricing models to28estimate the WTP for living near farmers markets in Edmonton, Canada. We find an29inverted U-shaped relationship between proximity to a farmers market and property30values. Our results suggest that local governments might consider the economic impact of31building new or relocating existing farmers markets on residential housing values,32alongside the benefits of improved access to high-quality food sources.33Keywords: Externality; farmers market; spatial hedonic pricing model; willingness to pay34JEL Classifications: D12; R3035Introduction36The extensive growth of farmers markets in Canada and worldwide has been advocated by37various agents including governments, private entrepreneurs, and consumers over the past38decades with the goal of improving healthfulness, access to fresh produce, and39environmental sustainability (Hunt 2007; Feagan and Morris 2009; Farmer et al. 2019;40Garner 2019). Many regional governments expand local food systems through the41promotion of farmers markets as a way to facilitate public health and promote local42economic development (Gillespie et al. 2008; Wittman et al. 2012; Kahin et al. 2017;43Malagon-Zaldua et al. 2018). Private entrepreneurs also argue for the creation of new44farmers markets by emphasizing the reduction in food miles from supporting local45producers (Martinez et al. 2010).46Consumers also have various motivations to support farmers markets. Some consumers47value farmers markets for their perceived higher quality and freshness of food products48(Ruelas et al. 2012; Dodds et al. 2014; Yu et al. 2017). Others view farmers markets as places49© The Author(s), 2024. Published by Cambridge University Press on behalf of Northeastern Agricultural and Resource50Economics Association This is an Open Access article, distributed under the terms of the Creative Commons Attribution51licence (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted re-use, distribution and reproduction,52provided the original article is properly cited.53Agricultural and Resource Economics Review (2024), 53, 277–29754doi:10.1017/age.2024.155https://doi.org/10.1017/age.2024.1 Published online by Cambridge University Press565758===== PAGE 9 (score 27, numeric-paren 0, table-words 2, var-words 21) =====5960Table 1. Variable definition and summary statistics61Variables62Description63Mean64Std. dev.65Dependent variable66Price67Log of transacted price (in 2016 CAD$)6812.98690.3470Key variable of interest71Farmers market72Distance to the nearest Farmers market732.86741.4975Structural variables76Living area77Square feet of living space781,559.6479620.8080Lot size81Square feet of lot size825,873.90834,338.7184Bedroom85Number of bedrooms862.92870.6588Bathroom89Number of bathrooms901.64910.6692House condition934 if the house condition is excellent, 3 if the house condition is good, 2 if the house94condition is average, and 1 if the house condition is poor952.99960.8397Basement condition984 if the house condition is excellent, 3 if the house condition is99good, 2 if the house condition is average, and 1 if the house condition is poor1002.441010.85102Garage103Number of garages1041.831050.47106House age107Age of the house10829.2010923.21110Locational variables111Shopping center112Number of shopping centers within each house’s 1,000-m buffer1130.161140.43115Grocery store116Number of grocery stores within each house’s 1,000-m buffer1170.301180.72119Bar120Number of bars within each house’s 1,000-m buffer1210.531221.25123Downtown124Distance to the downtown area12510.571264.31127(Continued)128Agricultural and Resource Economics Review129285130https://doi.org/10.1017/age.2024.1 Published online by Cambridge University Press131132133===== PAGE 10 (score 16, numeric-paren 0, table-words 1, var-words 12) =====134135Table 1. (Continued )136Variables137Description138Mean139Std. dev.140University141Distance to the University of Alberta14211.441433.96144Hospital145Distance to a hospital1465.051472.35148River buffer1491 if the house is located within a 2 km buffer area of the North150Saskatchewan River, 0 otherwise1510.311520.46153Golf154Area (in m2) of a golf course within a 600-m buffer1557,155.2615640,950.36157Green space158Area (in m2) of park and natural preserve within a 600-m buffer15955,037.0716087,327.32161Neighborhood variables162Population density163Neighborhood-level population density1643,071.331651,036.51166Low education167The share of residents whose highest level of education is high school or lower1680.371690.12170Low income171The share of households who have a relative low income (less than C$30,000)1720.121730.10174Unemployment175The share of residents who are unemployed1760.091770.04178Senior179The share of the senior population aged over 651800.141810.08182Public school183The share of residents who are located in a public school zone1840.751850.06186Married187The share of married residents1880.491890.06190Walk191The share of residents who rely upon walking as the primary mode of transportation1920.031930.04194Public195transportation196The share of residents who rely upon public transportation as the primary197mode of transportation1980.131990.05200286201Yanan Zheng et al.202https://doi.org/10.1017/age.2024.1 Published online by Cambridge University Press203204205===== PAGE 11 (score 30, numeric-paren 15, table-words 4, var-words 5) =====206207The parameter estimates of the OLS and SAR models are shown in Table 3. The208coefficients on the spatially lagged dependent variable from the four SAR models (i.e., ρ)209are positive and significant, which confirm the presence of spatial dependence and the210necessity of employing spatial HPMs. Therefore, the rest of the section only discusses the211estimates from the spatial models and omits the discussion of the biased OLS estimates.212Our key variables of interest are Farmers Market and Farmers Market2. The positive213and significant coefficients on Farmers Market and the negative and significant coefficients214on Farmers Market2 from four out of the five SAR models suggest an inverted U-shaped215relationship between distance to farmers markets and property values. The only exception216is the specification using the nearest 10 weight, which is less suitable than the nearest 20217and nearest 30 specifications based on Z-scores from a spatial autocorrelation test.218Specifically, it indicates that property values were valued the highest when the distance to219the nearest farmers market was around 4.60 km to 4.90 km depending on the weighting220matrix.3 When the distance to the nearest farmers market was below this benchmark,221property values fell as the distance dropped. This might be attributable to the negative222externalities such as congestion and noise brought by the farmers markets. Even though223farmers markets also generate positive externalities for nearby residents, negative224externalities may offset the positive ones. Moreover, when the distance to the nearest225farmers market was above the benchmark, property values fell as the distance increased.226This might be due to the fact that residents were not able to enjoy the benefits brought by227farmers markets when living too far away. Though Edmonton citizens were car-228dependent, a 4.5-km drive typically took approximately 10 min. Overall, the results show229property values initially increased and then decreased with the distance to farmers230markets, which is consistent with our conceptual hypothesis. Our results regarding the231dominance of negative externalities seem to be in contrast with previous studies (e.g.,232Collins 2020) that found the opposite, that is, that positive externalities dominate, and233housing values decrease with proximity to a farmers market. Our finding is not wholly234unexpected because the effect of the farmers market on housing values is likely to be235context- and/or region-specific. Edmonton is a large urban city, and its two most236prominent farmers markets are very large and attract a large clientele. In fact, Collins237(2020) noted that “nuisances are more likely to be associated with larger, well-established238markets,” which is an apt description of Edmonton’s main farmers markets.239Table 2. Results of spatial dependence testing240Tests241Moran’s I242LM spatial lag243Robust LM spatial lag244800-m weights2450.155***2461,245.300***247207.430***2481000-m weights2490.133***2501,028.200***251170.950***252Nearest 10 weights2530.214***2541,398.800***255264.840***256Nearest 20 weights2570.188***2581,515.400***259288.030***260Nearest 30 weights2610.173***2621,415.300***263248.750***264Note: Asterisks denote levels of significance (***for 1%).2653Based on the coefficient estimates from the SAR model with the nearest 20 weight matrix, the turning266point is calculated to be about 4.60 km to the nearest farmers market. Similarly, with the 1,000-m weight267matrix, the turning point is calculated to be about 4.90 km.268Agricultural and Resource Economics Review269287270https://doi.org/10.1017/age.2024.1 Published online by Cambridge University Press271272273===== PAGE 12 (score 95, numeric-paren 76, table-words 1, var-words 8) =====274275Table 3. Regression estimation results276Variables277OLS278Spatial lag models279Nearest 30 weights280800-m weights2811000-m weights282Nearest 10 weights283Nearest 20 weights284Farmers market2850.0111**2860.0106***2870.0114***2880.0092**2890.0096**2900.0105**291(0.0044)292(0.0041)293(0.0041)294(0.0040)295(0.0040)296(0.0041)297Farmers market2298−0.0009299−0.0011*300−0.0012*301−0.0010302−0.0010*303−0.0011*304(0.0007)305(0.0006)306(0.0006)307(0.0006)308(0.0006)309(0.0006)310Structural variables311Living area3120.0003***3130.0002***3140.0002***3150.0002***3160.0002***3170.0002***318(4.07E-06)319(3.95E-06)320(3.96E-06)321(3.99E-06)322(3.97E-06)323(3.96E-06)324Lot size3257.04E-06***3266.67E-06***3276.81E-06***3286.29E-06***3296.58E-06***3306.73E-06***331(3.80E-07)332(3.58E-07)333(3.61E-07)334(3.53E-07)335(3.54E-07)336(3.56E-07)337Bedroom338−0.0188***339−0.0120***340−0.0135***341−0.0101***342−0.0106***343−0.0117***344(0.0029)345(0.0027)346(0.0028)347(0.0027)348(0.0027)349(0.0027)350Bathroom3510.0319***3520.0275***3530.0287***3540.0270***3550.0263***3560.0268***357(0.0031)358(0.0029)359(0.0029)360(0.0029)361(0.0029)362(0.0029)363House condition3640.0180***3650.0164***3660.0168***3670.0161***3680.0158***3690.0159***370(0.0020)371(0.0019)372(0.0019)373(0.0018)374(0.0018)375(0.0019)376Basement condition3770.0449***3780.0445***3790.0446***3800.0431***3810.0440***3820.0443***383(0.0020)384(0.0019)385(0.0019)386(0.0019)387(0.0019)388(0.0019)389(Continued)390288391Yanan Zheng et al.392https://doi.org/10.1017/age.2024.1 Published online by Cambridge University Press393394395===== PAGE 13 (score 113, numeric-paren 91, table-words 1, var-words 10) =====396397Table 3. (Continued )398Variables399OLS400Spatial lag models401Nearest 30 weights402800-m weights4031000-m weights404Nearest 10 weights405Nearest 20 weights406Garage4070.1169***4080.1105***4090.1122***4100.1041***4110.1057***4120.1069***413(0.0039)414(0.0037)415(0.0037)416(0.0037)417(0.0037)418(0.0037)419House age420−0.0024***421−0.0022***422−0.0022***423−0.0023***424−0.0022***425−0.0022***426(0.0001)427(0.0001)428(0.0001)429(0.0001)430(0.0001)431(0.0001)432Locational variables433Shopping center434−0.0035435−0.0039436−0.0042437−0.0046438−0.0042439−0.0039440(0.0039)441(0.0037)442(0.0037)443(0.0036)444(0.0036)445(0.0036)446Grocery store447−0.001448−0.0015449−0.0016450−0.0008451−0.0008452−0.0009453(0.0024)454(0.0022)455(0.0022)456(0.0022)457(0.0022)458(0.0022)459Bar460−0.0023*4610.00134620.00124630.00024640.0014650.0013466(0.0014)467(0.0013)468(0.0013)469(0.0013)470(0.0013)471(0.0013)472Downtown473−0.0054***474−0.0127***475−0.0143***476−0.0081***477−0.0096***478−0.0108***479(0.0014)480(0.0013)481(0.0013)482(0.0013)483(0.0013)484(0.0013)485University of Alberta486−0.0114***487−0.0028***488−0.0012***489−0.0070***490−0.0056***491−0.0043***492(0.0015)493(0.0015)494(0.0015)495(0.0014)496(0.0014)497(0.0014)498Hospital499−0.0037***500−0.0050***501−0.0054***502−0.0035***503−0.0043***504−0.0049***505(0.0011)506(0.0010)507(0.0010)508(0.0010)509(0.0010)510(0.0010)511River buffer5120.0302***5130.0106***5140.0088***5150.0189***5160.0164***5170.0158***518(0.0044)519(0.0042)520(0.0042)521(0.0041)522(0.0041)523(0.0041)524(Continued)525Agricultural and Resource Economics Review526289527https://doi.org/10.1017/age.2024.1 Published online by Cambridge University Press528529530===== PAGE 14 (score 86, numeric-paren 67, table-words 1, var-words 7) =====531532Table 3. (Continued )533Variables534OLS535Spatial lag models536Nearest 30 weights537800-m weights5381000-m weights539Nearest 10 weights540Nearest 20 weights541Golf5421.29E-07***5438.55E-08***5441.11E-07***5459.61E-08***5461.00E-07***5471.21E-07***548(3.99E-08)549(3.76E-08)550(3.80E-08)551(3.71E-08)552(3.72E-08)553(3.74E-08)554Green space5554.12E-07***5562.22E-07***5572.38E-07***5582.11E-07***5592.05E-07***5602.11E-07***561(2.07E-08)562(2.05E-08)563(2.06E-08)564(2.02E-08)565(2.03E-08)566(2.04E-08)567Neighborhood variables568Population density569−1.21E-05***570−8.11E-06***571−9.33E-06***572−6.01E-06***573−6.22E-06***574−7.99E-06***575(1.94E-06)576(1.83E-06)577(1.85E-06)578(1.81E-06)579(1.81E-06)580(1.82E-06)581Low education582−0.6492***583−0.3169***584−0.3494***585−0.3163***586−0.2940***587−0.2971***588(0.0291)589(0.0293)590(0.0295)591(0.0285)592(0.0288)593(0.0292)594Low income595−0.1937***596−0.1819***597−0.1761***598−0.1678***599−0.1733***600−0.1710***601(0.0383)602(0.0361)603(0.0364)604(0.0356)605(0.0357)606(0.0359)607Unemployment608−0.3507***609−0.0319610−0.0162611−0.1392*612−0.1048613−0.0769614(0.0819)615(0.0776)616(0.0784)617(0.0763)618(0.0766)619(0.0771)620Senior6210.1006***6220.1018***6230.1040***6240.0975***6250.0959***6260.0847***627(0.0316)628(0.0298)629(0.0300)630(0.0293)631(0.0294)632(0.0296)633Public school6340.0536350.1528***6360.1471***6370.1364***6380.1327***6390.1346***640(0.0349)641(0.0330)642(0.0334)643(0.0324)644(0.0325)645(0.0328)646Married647−0.1510***648−0.3225***649−0.3283***650−0.2247***651−0.2480***652−0.2591***653(0.0545)654(0.0518)655(0.0525)656(0.0506)657(0.0508)658(0.0512)659(Continued)660290661Yanan Zheng et al.662https://doi.org/10.1017/age.2024.1 Published online by Cambridge University Press663664665===== PAGE 15 (score 37, numeric-paren 24, table-words 2, var-words 5) =====666667Table 3. (Continued )668Variables669OLS670Spatial lag models671Nearest 30 weights672800-m weights6731000-m weights674Nearest 10 weights675Nearest 20 weights676Walk6771.0992***6780.6467***6790.6997***6800.7167***6810.6742***6820.7033***683(0.0703)684(0.0675)685(0.0679)686(0.0662)687(0.0664)688(0.0667)689Public transportation690−0.4968***691−0.4335***692−0.4605***693−0.3662***694−0.3746***695−0.4024***696(0.0559)697(0.0526)698(0.0531)699(0.0520)700(0.0521)701(0.0523)702Constant703Control704Control705Control706Control707Control708Control709Monthly dummies710Control711Control712Control713Control714Control715Control716Adjusted R27170.823718Rho7190.4077200.4037210.3537220.3867230.395724Log likelihood7254,419.847264,869.087274,801.747284,952.957294,950.077304,911.76731AIC732−8,721.68733−9,618.20734−9,483.50735−9,785.90736−9,780.10737−9,703.50738Note: Asterisks denote levels of significance (*for 10%, **for 5%, ***for 1%).739Agricultural and Resource Economics Review740291741https://doi.org/10.1017/age.2024.1 Published online by Cambridge University Press742743744===== PAGE 16 (score 39, numeric-paren 0, table-words 3, var-words 33) =====745746Nearly every coefficient on the structural variables has the expected sign and is747statistically significant. The results suggest that larger houses with better conditions tended748to be more highly priced than others. In particular, property values benefited from an749increase in basement and house conditions, the number of bathrooms and size of garages,750as well as the square feet of living area and lot size. The lot size had a very small effect on751housing values, which is consistent with other studies such as Feng and Humphreys752(2018). Property values were penalized for older age. One unexpected finding is the753negative and statistically significant coefficient on the number of bedrooms. Other studies754that found a similar result (e.g., Li et al. 2015) proposed energy cost as the possible755explanation.756Almost all locational attributes are found to have capitalization effects on property757values. For the five variables controlling for the effects of commercial activities (shopping758center, grocery stores, bars, the downtown area, and the University of Alberta), we find759property prices increased as the household got closer to the downtown area and the760University of Alberta, which is in accordance with existing studies (e.g., Morancho 2003;761Jim and Chen 2006). Similarly, living close to hospitals in general positively affected762property prices. Good accessibility to green space and golf courses, as well as better scenic763river views, were also positively related to housing values. Our results consistently support764the positive contribution of water bodies, golf courses, parks, and public reserves to the765increasing house values, as suggested by numerous empirical studies (Larson and Perrings7662013; Sander and Polasky 2009; Samad et al. 2020).767In terms of the neighborhood characteristics, the results show that property values768tended to be higher if they were located in neighborhoods with a lower population density769and a higher portion of senior residents. Property values also benefited from an increase in770the percentage of residents who were located in public school zones and relied on walking771as the primary mode of transportation. On the other hand, property values were lower if772the neighborhoods had a higher proportion of residents with low education, low income,773and who were married. Property values were also penalized for being in an area with a large774percentage of residents who relied upon public transportation as the primary mode of775transportation. One potential reason is that properties in such neighborhoods tended to776have better access to transportation infrastructure, and while accessibility to public777transportation could have a positive effect due to convenience, the negative effect caused778by congestion, noise, and pollution may dominate the positive effect. Moreover, about 83%779of the residents in the sampled neighborhoods had cars, and this may explain why people780do not value public transportation in Edmonton as highly as in other cities.781Valuing marginal effects and quantifying marginal WTP/WTA782We calculate the marginal effects and marginal WTP/WTA for distance to farmers783market using seven distance points: mean distance (2:68 km), lowest distance (0.63 km =784mean distance −1.5 S.D), lower distance (1.37 km = mean distance −1 S.D), low distance785(2.12 km = mean distance −05 S.D), high distance (3.60 km = mean distance 0.5 S.D),786higher distance (4.35 km = mean distance 1 S.D), and highest distance (5.10 km = mean787distance 1.5 S.D). Total marginal effects and marginal WTP/WTA are reported in788Table 4. Although the magnitude of marginal effects and marginal WTP/WTA varies by789weight matrix, the sign and statistical significance remain the same. Because the SAR790model with the nearest 20 weights has the lowest AIC value and the highest log-likelihood791value, the following discussion focuses on the estimation results from this model.792The results indicate that at the mean distance level, property values increased by 0.59%793of the average property price if their distance to the nearest farmers market increased by794292795Yanan Zheng et al.796https://doi.org/10.1017/age.2024.1 Published online by Cambridge University Press797798799===== PAGE 17 (score 70, numeric-paren 56, table-words 1, var-words 4) =====8008011 km. Its estimated range is from 1.35% for houses located at the lowest distance (0.63 km)802away from a farmers market to −0.17% for houses located at the highest distance (5.09 km)803away from a farmers market.804For houses located 2.68 km (the mean distance) away from a farmers market, our805estimates suggest that households were willing to pay C$2,724.72 to live 1 km further away.806For houses located at the lowest distance (0.63 km) away from a farmers market, the807households were willing to pay C$6,225.82. At the other end, for houses located at the808highest distance (5.1 km) away from a farmers market, the households were willing to809accept C$776.38 to live 1 km further away. The estimation results are consistent with our810conceptual framework assumptions: people’s marginal WTP for living further away from811farmers markets declines with the average distance to farmers markets. Using the812transaction values of individual houses and their distances to the nearest farmers market,813Table 4. Marginal Effects and WTP/WTA for Access to Farmers Markets814800-m815weights8161000-m817weights818Nearest 10819weights820Nearest 20821weights822Nearest 30823weights824Marginal825effects826Lowest827distance8280.0155***8290.0166***8300.01228310.0135**8320.0149**833Lower834distance8350.0126***8360.0137***8370.00998380.0110**8390.0121**840Low841distance8420.0097***8430.0108***8440.00768450.0084**8460.0093**847Mean848distance8490.0069***8500.0079***8510.00538520.0059**8530.0065**854High855distance8560.0040***8570.0050***8580.00308590.0034**8600.0037**861Higher862distance8630.0011***8640.0022***8650.00068660.0008**8670.0009**868Highest869distance870−0.0017***871−0.0007***872−0.0017873−0.0017**874−0.0019**875Marginal876WTP877Lowest878distance8797,128.77***8807,638.13***8815,629.118826,225.82**8836,870.90**884Lower885distance8865,808.79***8876,309.01***8884,562.538895,058.78**8905,583.56**891Low892distance8934,488.81***8944,979.89***8953,495.958963,891.75**8974,296.22**898Mean899distance9003,168.83***9013,650.76***9022,429.379032,724.72**9043,008.87**905High906distance9071,848.84***9082,321.64***9091,362.799101,557.68**9111,721.53**912Higher913distance914528.86***915992.51***916296.21917390.65**918434.19**919Highest920distance921−791.12***922−336.61***923−770.37924−776.38**925−853.16**926Note: Asterisks denote levels of significance (**for 5%, ***for 1%).927Agricultural and Resource Economics Review928293929https://doi.org/10.1017/age.2024.1 Published online by Cambridge University Press930931932===== PAGE 18 (score 8, numeric-paren 0, table-words 0, var-words 8) =====933934we further derive the total economic effect of farmers markets on the prices of all houses935sold during the study period. At the aggregate level, we find that the negative externalities936generated by farmers markets dominate the positive ones. The reduction in transacted937housing values is estimated to be C$23,190,452. Using the 2019 property tax rate (0.9%),938the annual reduction in tax revenue from these transacted houses was about C$208,714939(City of Edmonton 2023a).940Conclusion941Farmers markets may generate both positive and negative externalities for nearby942residents, which may influence how they value the presence of these markets when943evaluating housing options. However, little attention is paid to quantify people’s WTP/944WTA for living in close proximity to farmers markets. By employing SAR models, this945study examines the impact of proximity to farmers markets on property values and946estimates people’s marginal WTP/WTA for living near farmers markets. The results947suggest that the relationship between property values and proximity to farmers markets948was nonlinear, with an ideal distance (i.e., highest value) of approximately 4.60– 4.90 km.949In terms of the corresponding WTP values, our estimates suggest that at the mean distance950(2.86 km) to a farmers market, households were willing to pay C$2,724.72 for every 1-km951increase in the distance to the farmers market. Moreover, households’ marginal WTP for952living further away from a farmers market decreased with their distance to the farmers953market. Overall, we find the negative externalities associated with farmers markets954dominated the positive ones, resulting in a C$23,190,452 reduction in transacted housing955values. Assuming a 0.9% property tax rate, the annual reduction in tax revenues from those956houses is estimated to be about C$208,714.957This study does have several limitations. First, like Warsaw and Phaneuf (2019) and958Collins (2020), there is a potential for endogeneity bias in the location of farmers markets.959That is, it is possible that housing values in a particular area are influencing the location of960farmers markets rather than vice versa. We have attempted to address this potential bias as961best we can with our commercial controls and spatial approach but it remains a potential962issue. Therefore, we recommend that the reader interprets the effect as correlational963instead of causal.964Our findings have several implications. First, introducing new farmers markets to965improve a neighborhood’s food environment may lead to some unintended consequences.966When the distance to the nearest farmers market is below a certain threshold, the value of967the property falls as its distance to a farmers market drops. Because property taxes are the968most substantial source of a municipality’s revenue, decreased property values might result969in substantial revenue loss, which could lead to reduced supplies of public goods and970services. Second, for neighborhoods with limited access to farmers markets, the local971government can introduce new farmers markets or relocate the existing ones because our972results show that when the distance to the nearest farmers market was above certain a973threshold, people required compensation to live further away from the farmers market.974When making the locating or relocating decisions, governments should take proximity to975farmers markets into consideration. Last, to reduce the potential negative externalities976associated with larger farmers markets, greater consideration should be given to improve977the operation of farmers market and limit the nuisance effects during market hours. This978may include enhanced traffic control, expanded street and sidewalk cleaning services, and979a greater police or security presence during market hours in the surrounding area.980294981Yanan Zheng et al.982https://doi.org/10.1017/age.2024.1 Published online by Cambridge University Press983