1# RECORD 9848e68efcee710bc05a55f8576e298c00f7e04b2# TITLE: The effect of remediating PCB-contaminated sediments on home prices in Milwaukee, Wisconsin, USA3# YEAR: 2022.0 VENUE: Agricultural and Resource Economics Review DOI: 10.1017/age.2022.264# AUTHORS: R. Wilkinson; Richard T. Melstrom5# PAGES: 18 CANDIDATE TABLE PAGES (1-based): [9, 10, 11, 17, 18]6# ABSTRACT: Abstract This paper estimates the effect of environmental remediation on housing prices in the Milwaukee Estuary Area of Concern (AOC) using a hedonic analysis of individual arms-length sales before and after a major remediation action between 2011 and 2015. Our design leverages this before-after comparison as well as the proximity of homes to remediation and the AOC boundary. Measuring the effect of AOCs in the housing market has always been a difficult task, given that water in AOCs can provide a mix of amenities and disamenities. Indeed, we find little evidence of a negative proximity effect when applying hedonic analysis to cross-section data. However, when we apply the analysis to a repeated cross section in a quasi-experimental framework, we find statistically significant evidence that living near the affected part of the AOC became more desirable after cleanup.789===== PAGE 1 (score 9, numeric-paren 0, table-words 0, var-words 9) =====1011RESEARCH ARTICLE12The effect of remediating PCB-contaminated13sediments on home prices in Milwaukee,14Wisconsin, USA15Robin Wilkinson and Richard T. Melstrom*16Loyola University Chicago, Chicago, IL 60660, USA17*Corresponding author: Email: rmelstrom@luc.edu18(Received 27 May 2022; revised 9 November 2022; accepted 11 November 2022; first published online 2219December 2022)20Abstract21This paper estimates the effect of environmental remediation on housing prices in the22Milwaukee Estuary Area of Concern (AOC) using a hedonic analysis of individual arms-23length sales before and after a major remediation action between 2011 and 2015. Our design24leverages this before-after comparison as well as the proximity of homes to remediation and25the AOC boundary. Measuring the effect of AOCs in the housing market has always been a26difficult task, given that water in AOCs can provide a mix of amenities and disamenities.27Indeed, we find little evidence of a negative proximity effect when applying hedonic analysis28to cross-section data. However, when we apply the analysis to a repeated cross section in a29quasi-experimental framework, we find statistically significant evidence that living near the30affected part of the AOC became more desirable after cleanup.31Keywords: areas of concern; Great Lakes; hedonic method; sediment remediation32JEL Codes: Q25; Q53; Q5833Introduction34Areas of Concern (AOCs) were established in 1987 as part of the Great Lakes Water35Quality Agreement (GLWQA) between the United States and Canada. AOCs are coastal36locations around the Great Lakes region that have been highly degraded from human37activity to a point where water-related beneficial uses are significantly impaired38(EPAa). The goal of the GLWQA’s AOC program is to restore AOCs by removing benefi-39cial use impairments (BUIs), which can range from restrictions on fish and wildlife40consumption, to beach closings, to degradation of esthetics; BUIs are directly experienced41by those living in the vicinity of AOCs. An AOC can be delisted once all BUIs associated42with it have been removed. To accomplish this, AOCs have been the targets of extensive43and expensive remediation efforts. Between 1985 and 2019, the United States and Canada44spent $22.78 billion on AOC remediation efforts (Hartig et al. 2020). With currently 6 of45the original 43 AOCs delisted, substantial restoration work remains to be done.46© The Author(s), 2022. Published by Cambridge University Press on behalf of the Northeastern Agricultural and Resource47Economics Association. This is an Open Access article, distributed under the terms of the Creative Commons Attribution48licence (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted re-use, distribution and reproduction,49provided the original article is properly cited.50Agricultural and Resource Economics Review (2023), 52, 71–8851doi:10.1017/age.2022.2652https://doi.org/10.1017/age.2022.26 Published online by Cambridge University Press535455===== PAGE 9 (score 31, numeric-paren 0, table-words 3, var-words 22) =====5657and the census tract shapefile for the City of Milwaukee to determine the tract associated58with each property. Homes sales occurred in 213 tracts.59Results and discussion60Table 2 presents the main coefficients of interest in the three models. For a complete set of61coefficient estimates, see the Appendix. These coefficients measure the percent change in62sales price associated with a 1% increase in distance (a decrease in proximity) to the AOC,63depending on whether a home is closest to the Milwaukee, Menominee, or Kinnickinnic64Rivers, or the treated area. For the sake of brevity, we will not describe the other coeffi-65cients in the model (i.e. on the structural attributes and lake and downtown distance vari-66ables), except to note that the estimates are in-line with expectations and consistent with67prior research. Model (1) provides no evidence that proximity affects price. In fact, the68Table 1. Summary statistics of Milwaukee home sales in the pre- and postproject phases69Variable70Mean71Std.72dev.73saleprice74Sales price75167051 12162976finsqft77Total finished square footage in thousands781.584790.83080acres81Acreage of lot820.097830.06784bdrms85Number of bedrooms862.963871.33088hbath89Number of half bathrooms900.250910.46592fbath93Number of full bathrooms941.480950.64296age97Age of property (2018 – year built)9877.6669932.510100distance101Straight-line distance to nearest part of AOC (100 s of meters)10220.89410314.784104distance to LP105Straight-line distance to Lincoln Park (100 s of meters)10693.02910739.304108downtown109Straight-line distance to downtown (100 s of meters)11055.65211128.414112lake113Straight-line distance to Lake Michigan (100 s of meters)11443.09111530.717116Dummy Variables117Count118Mean119ranch120Ranch-style house1214441220.113123colonial124Colonial-style house1251991260.051127capecod128Cape Cod-style house1296501300.166131condo132Condominium1338511340.217135Milwaukee136Nearest the Milwaukee River in the AOC13715391380.392139Menomonee140Nearest the Menomonee River in the AOC14113281420.338143Kinnickinnic144Nearest the Kinnickinnic River in the AOC14510581460.267147Treated148Nearest the Milwaukee River segment with water quality149improvements, downstream of the remediation at Lincoln Park1508331510.212152Total sales1533,925154Agricultural and Resource Economics Review15579156https://doi.org/10.1017/age.2022.26 Published online by Cambridge University Press157158159===== PAGE 10 (score 40, numeric-paren 22, table-words 6, var-words 3) =====160161coefficients on ln distance162163 × Menominee and ln distance164165 × Kinnickinnic are negative,166which runs contrary to expectations, though neither is statistically significant. The167coefficient on ln distance168169 × Milwaukee is also not significantly different from zero.170Now consider the coefficients in model (2), which are estimated using both the171pre-172and173postproject174data.175The176coefficients177on178ln distance179180 × Menominee181and182ln distne183184 × Kinnickinnic are significantly negative, which implies that homebuyers place185a premium on living near the Menominee and Kinnickinnic Rivers. The coefficients on186ln distance187188 × Milwaukee and ln distance to LP189190 distances are both statistically insignifi-191cant. However, the coefficient on ln distance to LP192193 × Post is significantly negative.194This indicates that a proximity premium for homes near Lincoln Park developed after195remediation.196Finally, consider the coefficients of model (3), which are also estimated using the197pre-198and199postproject200data.201The202coefficients203on204ln distance205206 × Menominee207and208ln distance209210 × Kinnickinnic are significantly negative, similar to model (2). The coefficient211on ln distance212213 × Milwaukee is not significantly different from zero. However, the214Table 2. Hedonic models – coefficients of interest215Model216(1)217(2)218(3)219ln distancei220221×Menomineei222−0.0275223−0.0705*224−0.0789**225(0.0707)226(0.0361)227(0.0366)228ln distancei229230×Kinnickinnici231−0.0551232−0.0923**233−0.1006**234(0.0701)235(0.0364)236(0.0368)237ln distancei238239×Milwaukeei2400.00232410.00692420.0448243(0.0654)244(0.0355)245(0.0374)246Postt247−0.0141248−0.0141249(0.0122)250(0.0130)251ln distance to LPi2522532540.0143255(0.0177)256ln distance to LPi257258×Postt259−0.0231*260(0.0129)261ln distancei262263×Treatedi264−0.0532*265(0.0295)266ln distnci267268×Treatedi×Postt269−0.0231*270(0.0139)271Housing structure characteristics272Yes273Yes274Yes275Month fixed effects276Yes277Yes278Yes279Tract fixed effects280Yes281Yes282Yes283R-squared2840.82962850.79662860.7971287Observations2881,0332893,9252903,925291Standard errors in parentheses below coefficients. * and ** indicate significance at 10% and 5% levels, respectively.29280293Robin Wilkinson and Richard T. Melstrom294https://doi.org/10.1017/age.2022.26 Published online by Cambridge University Press295296297===== PAGE 11 (score 45, numeric-paren 25, table-words 6, var-words 5) =====298299coefficient on ln distance300301 × Treated is significantly negative, which indicates a price302premium for homes in the treated area relative to those in proximity to other parts of303the Milwaukee River. Turning to the effect of remediation in the treated area, the coeffi-304cient on ln distance305306 × Treated × Post is significantly negative. To be specific, the coef-307ficients on ln distance308309 × Treated and ln distance310311 × Treated × Post imply that before312remediation a 100% increase in distance (e.g. 0.5 km to 1 km) from the AOC (in the treated313area) was associated with a 5.3% reduction in price, which is $8,880 at the average price,314and that after remediation the same change was associated with a 7.6% reduction, which is315$12,740 using the average sale price for properties in our analysis (Table 1).316Robustness checks317We performed three robustness checks to test the sensitivity of the estimates to the defi-318nition of the treated area, controlling for time-varying unobservables, and the timing of319capitalization effects. These results are shown in Table 3. The first robustness check revises320the definition of the treated area in model (3) by reducing the area from 8 to 5 km321Table 3. Robustness checks applied to model (3)322Treated area 5 km323downstream324Add tract-by-post fixed325effects326Mature-project327ln distancei328329×Menomineei330−0.0797**331−0.0795**332−0.0257333(0.0393)334(0.0379)335(0.0311)336ln distncei337338×Kinnickinnici339−0.1008**340−0.0966**341−0.0573*342(0.0391)343(0.0380)344(0.0318)345distancei346347×Milwaukeei3480.02823490.0730*3500.0771**351(0.0367)352(0.0382)353(0.0318)354Postt3550.0282356–357–358(0.0367)359–360–361ln distancei362363×Treatedi364−0.0300365−0.0239366−0.0048367(0.0529)368(0.0432)369(0.0407)370ln distancei371372×Treatedi×Postt373−0.0303**374−0.0718375−0.1078**376(0.0133)377(0.0471)378(0.0414)379Housing structure380characteristics381Yes382Yes383Yes384Month fixed effects385Yes386Yes387Yes388Tract fixed effects389Yes390Yes391Yes392Tract-by-post fixed effects393No394Yes395Yes396R-squared3970.79683980.80793990.8124400Observations4013,9254023,9254034,437404Note that the effect of Postt is not identified when the regression includes tract-by-post effects.405Standard errors in parentheses below coefficients. * and ** indicate significance at 10% and 5% levels, respectively.406Agricultural and Resource Economics Review40781408https://doi.org/10.1017/age.2022.26 Published online by Cambridge University Press409410411===== PAGE 12 (score 20, numeric-paren 0, table-words 6, var-words 8) =====412413downstream of Lincoln Park. These results are shown in the first column of Table 3. This414results in a more precisely estimated, larger (in absolute value) treatment effect than in415Table 2. This should increase confidence that remediation affected sales prices, particularly416for homes just downstream of Lincoln Park. The second column returns to the417original definition of the treated area and adds tract-by-post fixed effects to correct for418the influence of time-varying neighborhood characteristics that could potentially be corre-419lated with the timing and location of remediation (e.g. major capital improvements or420changes in land use). The addition of these fixed effects increases the coefficient on421ln distance422423 × Treated × Post; however, the estimate is less precise and no longer signifi-424cantly different from zero at conventional levels. Finally, the last robustness check uses an425alternative pre-post comparison, by replacing the 2,892 postproject sales with 3,404426“mature-project” sales that occurred between August 2017 and July 2018, in case the effect427of cleanup takes longer to capitalize than a single year after the project ends. With tract-by-428post fixed effects, the coefficient on ln distance429430 × Treated × Post is significantly negative431and substantially larger relative to the earlier estimates. This suggests the effect of reme-432diation could take several years after a project end to fully materialize.4433Additional discussion434The results indicate that hedonic analysis based on distance cannot always estimate the435effect of an AOC on home prices. The coefficients on the proximity variables in model436(1) provide no evidence that homebuyers either discount or place a premium on a home’s437proximity to the Milwaukee Estuary AOC. The variation in sign between different rivers in438the AOC adds further uncertainty as to the actual proximity effect. Prior research has439found that home values tend to rise with distance (Braden et al. 2008a, Braden et al.4402008b), a result which we are unable to replicate. This does not mean that buyers are indif-441ferent to the presence of AOCs. Rather, our results suggest that, at least in some housing442markets, the proximity effect of an AOC can be tempered or entirely offset by amenities443near the water, as appears to be the case in Milwaukee.444The overall effects of proximity notwithstanding, the results imply that buyers care445about the presence of PCBs and remediation because cleaning up part of the446Milwaukee Estuary led to capital gains that increased with proximity. Both the coefficient447on distance to remediation at Lincoln Park in model (2) and the coefficient on distance to448the downriver segment in model (3) were significantly negative, implying that living closer449to these areas became more valuable after remediation. We can use the change in the price450gradient associated with the cleanup effect to estimate the capital gain from remediation,451using the formula 100 × βTPlndistancei lndig452stance453~454, where dig455stance456~457is the hypo-458thetical boundary distance of the effect. Given we trimmed the data to sales within4595 km, and measured distance in hundreds of meters, dig460stance461~462= 50. Using the coefficients463in model (3) (the estimates are similar using model (2)), for a property in the treated area464adjacent to the water, which we define as within 150 m, the implied capital gain is 8%, or465$13,538 using the average property sales price. For properties affected by the improvement466but located 3 km away, the implied capital gain is just 1%, or $1,972. Across all homes in467the treated area, the implied capital gain is 3%, or $5,683. Note that the gain is larger if we468were to use the coefficients in the robustness checks. For example, the implied capital gain4694An important caveat, though, is that a dam in the treated area was removed in 2018, and the mature-470project estimates may pick up the effect of removal. The direction of any bias, though, is unclear, because471removal was expected to improve water quality and restore habitat (a benefit) but reduce water levels for472upstream recreation (a potential loss).47382474Robin Wilkinson and Richard T. Melstrom475https://doi.org/10.1017/age.2022.26 Published online by Cambridge University Press476477478===== PAGE 17 (score 67, numeric-paren 48, table-words 1, var-words 8) =====479480Appendix481Table A1. Hedonic models – all coefficients482Model483(1)484(2)485(3)486Finsqft4870.2830444880.2801724890.280050490(0.024368)491(0.010226)492(0.010258)493Bdrms494−0.039027495−0.037703496−0.037248497(0.016597)498(0.007050)499(0.007058)500Fbath5010.1756025020.1416405030.140987504(0.026654)505(0.011502)506(0.011494)507Hbath5080.1266075090.1246495100.125128511(0.026258)512(0.011963)513(0.011927)514Age515−0.012805516−0.011156517−0.011028518(0.002421)519(0.001245)520(0.001255)521Age25220.0000765230.0000675240.000066525(0.000016)526(0.000008)527(0.000008)528Waterfront5294.4988265300.9871845310.980320532(1.325469)533(0.383445)534(0.383740)535Acres536−11.702370537−1.001644538−1.014829539(3.848793)540(0.705101)541(0.698689)542Acres25430.1597655440.1426915450.142736546(0.038555)547(0.024794)548(0.017641)549Cape Cod5500.1530715510.1860785520.188487553(0.051420)554(0.017676)555(0.024864)556Colonial5570.145395558−0.003397559−0.007865560(0.1270)561(0.052669)562(0.052892)563Condo5640.1208795650.0814675660.081110567(0.048037)568(0.021277)569(0.021237)570Ranch571−0.0709645720.1423485730.162981574(0.084647)575(0.049215)576(0.049872)577DistancetoDowntown578−0.256703579−0.3224345800.241554581(0.070046)582(0.038786)583(0.092125)584DistancetoLake5850.2604775860.2111595870.241554588(0.130816)589(0.088278)590(0.092125)591ln distance592593×Menomineei594−0.027456595−0.070494596−0.078889597(0.070704)598(0.036149)599(0.036620)600(Continued)601Agricultural and Resource Economics Review60287603https://doi.org/10.1017/age.2022.26 Published online by Cambridge University Press604605606===== PAGE 18 (score 20, numeric-paren 12, table-words 1, var-words 4) =====607608Table A1. (Continued )609Model610(1)611(2)612(3)613ln distancei614615×Kinnickinnici616−0.055137617−0.092289618−0.100555619(0.070136)620(0.036414)621(0.036838)622ln distncei623624×Milwaukee6250.0022916260.0069066270.044785628(0.065402)629(0.035543)630(0.037371)631Postt632−0.014117633−0.014066634(0.012180)635(0.013048)636ln distance to LPi6376386390.014304640(0.017707)641ln distance to LPi642643×Postt644−0.023126645(0.012902)646ln distancei647648×Treatedi649−0.053156650(0.029504)651ln distancei652653×Treatedi×Postt654−0.023111655(0.013905)656Month fixed effects657Yes658Yes659Yes660Tract fixed effects661Yes662Yes663Yes664Cite this article: Wilkinson, R. and R. T. Melstrom (2023). “The effect of remediating PCB-contaminated665sediments on home prices in Milwaukee, Wisconsin, USA.” Agricultural and Resource Economics Review 52,66671–88. https://doi.org/10.1017/age.2022.2666788668Robin Wilkinson and Richard T. Melstrom669https://doi.org/10.1017/age.2022.26 Published online by Cambridge University Press670