Disentangling the Impacts of Environmental Contamination from ...

Disentangling the Impacts of Environmental Contamination from ...

Disentangling the Impacts of Environmental Contamination from Locally Undesirable Land-uses (LULUs) on Residential Property Values Xiangping Liu, Laura Taylor, and Daniel Phaneuf June 25, 2010 Literature on Environmentally Contaminated Sites Use distance to a site listed on National Priority List (NPL) to proxy its impact Study a single or several NPL sites in one area Summarized in Kiel and Williams (2007) Literature on Environmentally

Contaminated Sites, contd. Heterogeneity of the impact by site Kiel and Williams (2007; all NPL sites) Stigma Messer et al. (2006) McCluskey and Rausser (2003) Dale et al. (1999) Our Contribution Particular focus is to separate two impacts: Contamination/cleanup (listing/delisting) LULU (commercial or industrial properties) Control for sites that are on both federal and

state priority lists Our Contribution, contd. Use accurate site boundary rather than the centroid Estimate heterogeneous impacts of site types Commercial vs. industrial sites Landfill, solvent, and military sites vs. other sites Use parcel-level transaction data, with careful attention to local unobservable community characteristics Identification strategy--geographic matching Key issue: identify treated group & construct

unobserved counterfactual for the treated observations. Distance matters, but distance can be correlated with local unobservable characteristics. Local community characteristics, if not controlled for, could bias the estimation. Identification strategy, contd. Treated observations: Residential properties within 0.3-mile buffer to the boundary of a listed site Identify land-use of each listed site Control observations: Residential properties within 0.3-mile buffer to the

boundary of a clean commercial/industrial (COM/IND) site Identify land-use of each clean site Match clean COM or IND sites to contaminated sites based on their spatial locations Identification strategy--geographic matching Match treated & control observations based on their spatial locations: 1. Treated residential properties are within 0.3 mile buffer of listed site Illustration of geographic matching

Identification strategy--geographic matching Match treated & control observations based on their spatial locations: 1. Treated residential properties are within 0.3 mile buffer of listed site 2. Create buffer ring of 0.7-1.0 mile from boundary of listed site Illustration of geographic matching Identification strategy--geographic matching Match treated & control observations based on their spatial locations: 1. Treated residential properties are within 0.3 mile buffer of listed site

2. Create buffer ring of 0.7-1.0 mile from boundary of listed site 3. Any clean commercial/industrial site lying within buffer ring serves as a control com/ind site Illustration of geographic matching Identification strategy--geographic matching Match treated & control observations based on their spatial locations: 1. Treated residential properties are within 0.3-mile buffer of listed site 2. Create buffer ring of 0.7-1.0 mile from centroid of listed site

3. Any clean commercial/industrial site lying within buffer ring serves as a control com/ind site 4. Control residential properties are within 0.3-mile buffer of control com/ind site Illustration of geographic matching Identification strategy--geographic matching Match treated & control observations based on their spatial locations: 1. Treated residential properties are within 0.3-mile buffer of listed site 2. Create buffer ring of 0.7-1.0 mile from centroid of listed site 3. Any clean commercial/industrial site lying within

buffer ring serves as a control com/ind site 4. Control residential properties are within 0.3-mile buffer of control com/ind site Illustration of geographic matching Heterogeneous effects and local community level unobservable Conduct analysis for COM and IND sites separately Examine the impact by site types, eg. landfill, solvent, military sites, and other sites separately Control for site fixed effect Check the robustness of the result by controlling both site fixed effect and local community characteristics

Data Study area: Minneapolis/St. Paul metropolitan statistical area, Minnesota Contaminated sites: All sites listed on state or federal registers of contaminated sites. 108 sites (51 listed after 1990 / 59 delisted after 1990) Site boundaries (manual identification) Land-use Contaminate type

Residential property transactions 1990-2007 housing attributes: acres, # rooms, bedrooms, baths, age, school district. Data Community characteristics: Census demographics at block group level are compiled from Geolytics Local land-use characteristics (within 0.5 miles or 1 mile) % land in: commercial use; industrial use; residential use; apartments, open space, water, highway

Example of treated & control residential properties Legend Control residential Listed sites Treated residential Clean com/ind Empirical specification Mathematical representation of estimating equations --Listing Ln(sales price)=*treat+*tl+*treat_tl*treat+*tl+*treat_tltreat+*tl+*treat_tl*treat+*tl+*treat_tltl+*tl+*treat_tl*treat+*tl+*treat_tltreat_tl +*tl+*treat_tlconstant+*tl+*treat_tl*treat+*tl+*treat_tlX+*tl+*treat_tl site fixed effect+*tl+*treat_tl time effect --De-listing

Ln(sales price)=*treat+*tl+*treat_tl*treat+*tl+*treat_tltreat+*tl+*treat_tl*treat+*tl+*treat_tltdl+*tl+*treat_tl*treat+*tl+*treat_tltreat_tdl +*tl+*treat_tlconstant+*tl+*treat_tl*treat+*tl+*treat_tlX+*tl+*treat_tl site fixed effect+*tl+*treat_tl time effect Preliminary sample results: listing All listed sites, no impact heterogeneity in land-use of site Housing Most recent characteristics transaction Sites listed in any year Listed since Jan. 1991(and all below are same) Y Y Y Y

Only sales within 7 years of listing Y Y Estimated effect 0.00352 -0.0296*treat+*tl+*treat_tl -0.0463*treat+*tl+*treat_tl*treat+*tl+*treat_tl -0.0360*treat+*tl+*treat_tl*treat+*tl+*treat_tl -0.0648*treat+*tl+*treat_tl*treat+*tl+*treat_tl*treat+*tl+*treat_tl -0.0459*treat+*tl+*treat_tl*treat+*tl+*treat_tl # Obs

85,079 24,491 17,058 24,487 17,054 8,740 Preliminary sample results: listing, continued. Heterogeneity by land-use: Commercial versus Industrial (all covariates & most recent transaction) COM IND

COM IND Excluding landfill, solvent & military sites N N Y Y Estimated effect -0.114*treat+*tl+*treat_tl*treat+*tl+*treat_tl*treat+*tl+*treat_tl 0.026

-0.0639 -0.536*treat+*tl+*treat_tl*treat+*tl+*treat_tl*treat+*tl+*treat_tl # Obs 5,943 5,034 5,848 3,525 Preliminary sample results: de-listing All de-listed sites, no impact heterogeneity in land-use of site Housing Most recent Estimated characteristics transactions effect

# Obs Site delisted any year -0.120*treat+*tl+*treat_tl*treat+*tl+*treat_tl*treat+*tl+*treat_tl 58,126 Site delisted during Jan. 1991 and Jan. 2007(and all below are same) -0.107*treat+*tl+*treat_tl*treat+*tl+*treat_tl*treat+*tl+*treat_tl 45,830 Y -0.105*treat+*tl+*treat_tl*treat+*tl+*treat_tl*treat+*tl+*treat_tl 31,243 Y -0.102*treat+*tl+*treat_tl*treat+*tl+*treat_tl*treat+*tl+*treat_tl 45,820 Y Y -0.095*treat+*tl+*treat_tl*treat+*tl+*treat_tl*treat+*tl+*treat_tl 31,233 Only sales within 7 years of delisting

Y Y -0.076*treat+*tl+*treat_tl*treat+*tl+*treat_tl*treat+*tl+*treat_tl 17,885 Preliminary sample results: de-listing, continued Heterogeneity by land-use: Commercial versus Industrial (all covariates & most recent transaction) COM IND COM IND Excluding landfill,

solvent & military sites N N Y Y Estimated effect -0.210*treat+*tl+*treat_tl*treat+*tl+*treat_tl*treat+*tl+*treat_tl 0.0451*treat+*tl+*treat_tl*treat+*tl+*treat_tl*treat+*tl+*treat_tl 0.0424*treat+*tl+*treat_tl 0.0460*treat+*tl+*treat_tl*treat+*tl+*treat_tl Obs

9,958 15,058 7,812 13,596 Preliminary sample results: landfill, solvent, military sites separately Heterogeneity by site types: Landfills, solvent sites and military sites (all covariates & most recent transaction) Landfills Solvent sites Listing/Delisting Listing

Delisting Listing Delisting Estimated effect Obs -0.133 95 -0.178*treat+*tl+*treat_tl*treat+*tl+*treat_tl*treat+*tl+*treat_tl 4,382 -0.0866*treat+*tl+*treat_tl*treat+*tl+*treat_tl*treat+*tl+*treat_tl 4,403 -0.00710 1,793 Conclusion

The effect of contamination--listing a site reduces nearby residential property value 3-7 % The listing has larger impact on residential property surrounding a industrial site than a commercial site about 7-12% for a COM site 47% an IND site Delisting some NPL/PLP sites improves nearby residential property by 5%. However, there exist stigma effect of cleanup on the nearby residential property for the landfill sites. Work-to-do Matching NPL/PLP sites based on local community characteristics rather than geographical location (a propensity score matching or matching on local

community characteristics directly). This method allows us to further check how strong local community characteristics or unobservables affect estimation results. Separate regressions for detached residential properties, townhouse, and condo The End Thank you! Preliminary sample results: de-listing, continued Heterogeneity by land-use: Commercial versus Industrial (all covariates & most recent transaction) Excluding landfill, solvent Estimated effect

& military sites COM IND COM IND COM (0-7) IND (0-7) COM (0-7) IND (0-7) N N Y Y

N N Y Y -0.210*treat+*tl+*treat_tl*treat+*tl+*treat_tl*treat+*tl+*treat_tl 0.0451*treat+*tl+*treat_tl*treat+*tl+*treat_tl*treat+*tl+*treat_tl 0.0424*treat+*tl+*treat_tl 0.0460*treat+*tl+*treat_tl*treat+*tl+*treat_tl -0.132*treat+*tl+*treat_tl*treat+*tl+*treat_tl*treat+*tl+*treat_tl 0.0204 -0.0534*treat+*tl+*treat_tl*treat+*tl+*treat_tl 0.0136 Obs

9,958 15,058 7,812 13,596 5,807 7,371 4,455 6,239 Preliminary sample results: de-listing, continued De-listed sites, exluding landfill, solvent & military sites Housing &

Neighborhood Most recent Estimated Demographics transactions effect # Obs Site delisted any year Site delisted during Jan. 1991 and Jan. 2007(and all below are same) 49029 36769 25184 36760 25175 13413 Y

Y Y 0.0298*treat+*tl+*treat_tl*treat+*tl+*treat_tl 0.0329*treat+*tl+*treat_tl*treat+*tl+*treat_tl 0.0183 0.0218 Y Y -0.0301*treat+*tl+*treat_tl*treat+*tl+*treat_tl

Y Only sales within 7 years of delisting -0.0292*treat+*tl+*treat_tl*treat+*tl+*treat_tl Preliminary sample results: de-listing, continued De-listed sites, only landfill, solvent & military sites Housing & Neighborhood Most recent Estimated Demographics transactions effect # Obs Site delisted any year Site delisted during Jan. 1991 and Jan.

2007(and all below are same) 9,039 9,039 6,047 9,038 6,046 4,479 Y Y Y -0.201*treat+*tl+*treat_tl*treat+*tl+*treat_tl*treat+*tl+*treat_tl

-0.235*treat+*tl+*treat_tl*treat+*tl+*treat_tl*treat+*tl+*treat_tl -0.192*treat+*tl+*treat_tl*treat+*tl+*treat_tl*treat+*tl+*treat_tl -0.214*treat+*tl+*treat_tl*treat+*tl+*treat_tl*treat+*tl+*treat_tl Y Y -0.153*treat+*tl+*treat_tl*treat+*tl+*treat_tl*treat+*tl+*treat_tl Y Only sales within 7 years of delisting -0.201*treat+*tl+*treat_tl*treat+*tl+*treat_tl*treat+*tl+*treat_tl

Recently Viewed Presentations

  • Internal Controls They Are Everyones Business Valdosta State

    Internal Controls They Are Everyones Business Valdosta State

    It is much less expensive to implement sound internal controls than it is to investigate and prosecute fraud or to recover from the bad publicity surrounding a fraud. From the famous Bernie Madoff and Enron headlines. In June of 2012,...
  • Lysbilde 1 - NAKU

    Lysbilde 1 - NAKU

    Trine Lise Bakken, PPU Ullevål sykehus John Petter Mykletun, Sykehuset Buskerud Olav Ose Evensen, Nordlandssykehuset Asbjørn Strømmen, Trondheim kommune Thor Kristiansen, pårørenderepresentant Nanna Gjersøe, pårørenderepresentant Geir Bjarte Gravdal, NAKU I tillegg har følgende bidragsytere vært hentet inn: Linda Morkemo, Trondheim...
  • Urban Economics, Ninth Edition, Chapter 20

    Urban Economics, Ninth Edition, Chapter 20

    The Tiebout model is a formal model of interjurisdictional mobility. The simple version of the Tiebout model is based on five assumptions. 1.Municipal choice. A household chooses the municipality (or school district or other local jurisdiction) that provides the household's...
  • Swingy - Perceptions

    Swingy - Perceptions

    The next day he saw the same Israelite involved in another fight. Moses went to him and said: "You seem to be a quarrelsome fellow. You have a new quarrel with one person or another each day." Fearing that Moses...
  • Antonia Susan Byatt POSSESSION A Romance

    Antonia Susan Byatt POSSESSION A Romance

    Antonia Susan Byatt POSSESSION. A Romance Structure CHAPTER II R. H. ASH, Ragnarök Nordic epic poem Mortimer Cropper - Trustee of the Stant Collection at Robert Dale Owen University, Harmon City, New Mexico.
  • Brief Review of the Literature - City Reformed Presbyterian ...

    Brief Review of the Literature - City Reformed Presbyterian ...

    Part of the OT Wisdom literature . ... Canadian Psychology, 52(1), 20-28. doi: 10.1037/a0022048. Barriers to an Open Dialogue. ... A great manual for Christian couples. The first few chapters focus on the heart of sex and being a great...
  • Florida's Bathymetry - University of Florida

    Florida's Bathymetry - University of Florida

    Ejection trajectories of grains blasted into the air by rainsplash The Raindrop CTSV Raindrop Size and Effects Rain Splash Experiments Measuring Net Downslope Rainsplash Transport Creep General term for slow, downslope mass movement of material in response to gravity on...
  • Mercantilism - Federal Reserve Bank of Dallas

    Mercantilism - Federal Reserve Bank of Dallas

    Principles of Mercantilism:Ships. Merchant fleet is a necessity to facilitated the flow of goods in and out of a nation. Naval fleet assures protection of transport. Focus changes from the protection of resources through military might to protection/ support through...