The Effect of Emission Information on Housing Prices: Quasi-Experimental Evidence from the European Pollutant Release and Transfer Register


In this paper, we study whether the release of pollutant emission information has an effect on housing prices. The event under study is the publication of the first wave of emission quantity data from the European Pollutant Release and Transfer Register in 2009. Our analysis is based on quarterly housing prices at the German postal code level for the years 2007–2011 and provides the first evidence from Europe on this research question. Estimating a differences-in-differences model and controlling for observable differences in land use, housing type distribution, tax revenues and other postal code area characteristics by means of propensity score matching, we find no significant effect of the release of emission information on the value of houses in affected postal code areas. This result survives a number of robustness checks designed to assess whether our findings are due to data aggregation issues or the actual treatment definition. This leads to the conclusion that on an aggregate level the 2009 publication of E-PRTR data did not have an immediate and noticeable effect on housing prices in Germany.

This is a preview of subscription content, log in to check access.

Fig. 1
Fig. 2
Fig. 3
Fig. 4

Change history

  • 24 November 2016

    An erratum to this article has been published.


  1. 1.

    Housing markets have also been used to evaluate changes in utility due to proximity to sex offenders (Linden and Rockoff 2008), school quality, etc.

  2. 2.

    Pope (2008) for example exploits the introduction of disclosure laws that require sellers to provide exact information about airport noise exposure to potential buyers.

  3. 3.

    A list of scientific articles relying on TRI data can be found in the booklet (Environmental Protection Agency 2013) released by the Toxics Release Inventory Program Division.

  4. 4.

    The EPER made Europe-wide emissions data for the year 2001 available in 2004 and emissions data for the year 2004 available in 2006. However, this register received very little public attention. A LexisNexis search involving German newspapers regarding the keyword “EPER” yielded only 7 hits for the time frame before 2009. Mentions of the term were largely concentrated in special interest journals regarding environmental topics or the waste treatment industry such as Entsorga (2004).

  5. 5.

    Examples of comprehensive newspaper articles on the newly available E-PRTR data include “Database of hazards”—Sueddeutsche (2010), “Pollutant information now online”—Hamburger (2009), and “Interesting information on environmental sinners in the neighborhood”—TAZ (2009).

  6. 6.

    Compared to the EEA website the localized website was fairly basic but also easier to use and has been updated substantially over the following years. The current German front-end application for private inquiries is available under: Accessed 21 Sep 16.

  7. 7.

    It is possible to obtain House Price Index values for a longer time horizon and E-PRTR emission data for the subsequent years (i.e. data from 2008 released in 2010, data from 2009 released in 2011 and so forth). However, the scope of this research project is to focus on the immediate impact of the initial data release. The following waves of E-PRTR information would allow inhabitants to continuously update their beliefs, which could potentially dilute the actual effect of interest.

  8. 8.

    This point is made in a recent paper by Abbott and Klaiber (2013). They use matching to account for observable characteristics, but limit matching to potential controls within a certain radius (spatial proximity) to obtain comparable units in terms of unobservable characteristics. The procedure is intuitively similar to the use of fixed effects to capture neighborhood unobservable characteristics.

  9. 9.

    In essence this corresponds to using matching as a non-parametric pre-processing of the data, see Ho et al. (2007). By limiting the analysis to a matched sample, the estimated models become less sensitive to misspecification as there is less implicit interpolation when the treatment and control group are balanced in terms of observable characteristics.

  10. 10. (F+B Forschung und Beratung für Wohnen, Immobilien und Umwelt GmbH, Hamburg). Accessed on 28-10-2013.

  11. 11.

    Database accessible via:

  12. 12.

    All relevant economic activities are listed in the Annex (p. 8 et seq.) to the regulations published in European Union (2006) in January 2006. Official information accessible via: The register includes information of about 29,000 facilities in 32 countries (EU27, Iceland, Liechtenstein, Norway, Switzerland and Serbia).

  13. 13.

    These thresholds will be used as weights to normalize emissions when calculating a weighted severity measure, see Appendix section A.2 “Construction of Weighted Emission Scores” or Chapter 6.1. All pollutants and thresholds are listed under: Accessed 21 Sep 16.

  14. 14.

    See question 5 on

  15. 15.

    Including these reports does not qualitatively affect the results.

  16. 16.

    Interestingly, using the geographic coordinates revealed that in more than 200 cases in 2007 alone, the postal code in the E-PRTR reflected the location of a firm’s main office rather than the location of the actual emission.

  17. 17.

    This shape file has also been used for the remaining maps displayed in this paper.

  18. 18.

    The data can be downloaded from the European Environment Agency website:

  19. 19.

    INdikatoren und KARten zur Raum- und Stadtentwicklung in Deutschland und in Europa—Indicators and maps on spatial and urban development in Germany and Europe.

  20. 20.

    Over the last years, there have been several municipal reforms merging and dividing municipalities. Since our INKAR data refers to the state ultimo 2008 we had to match municipalities from this time period to present municipal structures and then to the postal code areas. As a result, 18 postal code areas were lost in the first step of this process.

  21. 21.

    Note that due to the inclusion of state fixed effects the results cannot be directly inferred from Table 2.

  22. 22.

    In our case, some 12 % of the postal code areas are affected and general equilibrium effects are conceivable if households respond strongly. If treatment has the effect of increasing house prices in surrounding non-treated areas, this would make an estimated effect larger rather than smaller and thus make it more likely that we would find a significant treatment effect.

  23. 23.

    To avoid multicollinearity we left the share of labor force employed in the primary sector, the share of single family housing, and the share of natural land use out of the logit models used for matching.

  24. 24.

    In addition to dropping treated units with propensity scores above the highest score for the potential control units, we also exclude the 2 % of the treated units with the lowest p-score density.

  25. 25.

    Huber et al. (2013) additionally use 1.5 and three times the maximum distance in an assessment of the performance of matching estimators. We choose the lower end specification to avoid including too many poor matches.

  26. 26.

    Following Dehejia and Wahba (2002) we use the weights generated in the matching procedure as frequency weights and estimate weighted regressions. The same weights are used in the calculation of the ATETs.

  27. 27.

    It is well-known that matching techniques can be sensitive to the specification of the logit/probit model. We tested alternative specifications without qualitatively changing the results.

  28. 28.

    Nevertheless, the reporting thresholds are an imperfect proxy for toxicity. They are not directly intended to capture toxicity but rather to ensure that a large fraction of emissions is covered by the register while at the same time minimizing unnecessary burdens for small emitters. Still, when looking across the table of thresholds and individual substances, there is a clear pattern that lower thresholds are associated with substances generally perceived as being dangerous. The full list of pollutants and their thresholds can be accessed under Accessed 21 Sep 16.

  29. 29.

    Similar to the main specification, the isolated treatment dummies are dropped due to the inclusion of postal code area fixed effects in all robustness checks.

  30. 30.

    We also carried out analyses distinguishing between emissions to air and water respectively. With only 6 emissions to soil the data is too thin to analyze this medium separately. Again, no significant effect could be found in either Western of Eastern Germany after matching was carried out. A table of these results is available from the authors upon request.

  31. 31.

    This result survives the robustness check of grouping 2–3 facility postal codes together in an alternative specification but remains at a fairly low level of significance throughout these tests.

  32. 32.

    All results are summarized in Appendix A.11 “Summary of mean comparisons and regression results” in Table 21 to allow for a quick comparison of estimated values across all methods and robustness checks.


  1. Abbott JK, Klaiber HA (2013) The value of water as an urban club good: a matching approach to community-provided lakes. J Environ Econ Manag 65(2):208–224

    Article  Google Scholar 

  2. Angrist JD, Pischke J-S (2009) Mostly harmless econometrics: an Empiricist’s companion. Princeton University Press, Princeton

    Google Scholar 

  3. Banzhaf HS, Walsh RP (2008) Do people vote with their feet? An empirical test of environmental gentrification. Am Econ Rev 98(3):843–863

    Article  Google Scholar 

  4. Bui LTM, Mayer CJ (2003) Regulation and capitalization of environmental amenities: evidence from the toxic release inventory in massachusetts. Rev Econ Stat 85(3):693–708

    Article  Google Scholar 

  5. Chay Kenneth Y, Greenstone M (2005) Does air quality matter? Evidence from the housing market. J Polit Econ 113(2)

  6. Currie J, Davis L, Greenstone M, Walker R (2015) Environmental health risks and housing values: evidence from 1,600 toxic plant openings and closings. Am Econ Rev 105(2):678–709

    Article  Google Scholar 

  7. Davis L (2004) The effect of health risk on housing values: evidence from a cancer cluster. Am Econ Rev 94(5):1693–1704

    Article  Google Scholar 

  8. Davis L (2011) The effect of power plants on local housing prices and rents. Rev Econ Stat 93(4):1391–1402

    Article  Google Scholar 

  9. Dehejia R, Wahba S (2002) Propensity score-matching methods for nonexperimental causal studies. Rev Econ Stat 84(1):151–161

    Article  Google Scholar 

  10. Entsorga (2004) Blick in die Zukunft. Issue: 06/21/2004

  11. Environmental Protection Agency (2013) The toxics release inventory in action: Media, government, business, community and academic uses of tri data.

  12. European Union (2006) Regulation (ec) no 166/2006 of the European Parliament and of the Council of 18 January 2006 (Off J Eur Union).

  13. F+B. F+B-Wohn-Index Deutschland - Methodiksteckbrief, 2012. F+B GmbH

  14. Gamper-Rabindran S, Timmins C (2013) Does cleanup of hazardous waste sites raise housing values? J Environ Econ Manag 65(3):345–360

    Article  Google Scholar 

  15. Greenstone M, Gallagher J (2008) Does hazardous waste matter? Evidence from the housing market and the superfund program. Q J Econ 123(3):951–1003

    Article  Google Scholar 

  16. Hamburger Abendblatt (2009) EU stellt Schadstoffquellen in Europa online. Issue: 11/10/2009:29, 2009. Axel Springer

  17. Hanna BG (2007) House values, incomes, and industrial pollution. J Environ Econ Manag 54(1):100–112

    Article  Google Scholar 

  18. Ho DE, Imai K, King G, Stuart EA (2007) Matching as nonparametric preprocessing for reducing model dependence in parametric causal inference. Polit Anal 15:199–236

    Article  Google Scholar 

  19. Huber M, Lechner M, Wunsch C (2013) The performance of estimators based on the propensity score. J Econ 175(1):1–21

    Article  Google Scholar 

  20. Kuminoff NV, Pope JC (2014) Do “capitalization effects” for public goods reveal the public’s willingnes to pay? Int Econ Rev 55(4):1227–1250

    Article  Google Scholar 

  21. Leuven E, Sianesi B (2003) PSMATCH2: Stata module to perform full Mahalanobis and propensity score matching, common support graphing, and covariate imbalance testing., current version: 4.0.11 (22.09.2014) edition

  22. Linden L, Rockoff JE (2008) Estimates of the impact of crime risk on property values from Megan’s laws. Am Econ Rev 98(3):1103–1127

    Article  Google Scholar 

  23. Lyon TP, Shimshack JP (2015) Environmental disclosure: evidence from Newsweek’s green companies rankings. Bus Soc 54(5):632–675

    Article  Google Scholar 

  24. Mastromonaco R (2015) Do environmental right-to-know laws affect markets? Capitalization of information in the toxic release inventory. J Environ Econ Manag 71:54–70

    Article  Google Scholar 

  25. Moretti E, Neidell M (2011) Pollution, health, and avoidance behavior: evidence from the ports of Los Angeles. J Hum Resour 46(1):154–175

    Google Scholar 

  26. Muehlenbachs L, Spiller E, Timmins C (2015) The housing market impacts of shale gas development. Am Econ Rev 105(12):3633–3659

    Article  Google Scholar 

  27. Neidell M (2009) Information, avoidance behavior, and health: the effect of ozone on asthma hospitalizations. J Hum Resour 44(2):450–478

    Google Scholar 

  28. Oberholzer-Gee F, Mitsunari M (2006) Information regulation: Do the victims of externalities pay attention? J Regul Econ 30:141–158

    Article  Google Scholar 

  29. Palmquist RB (2006) Handbook of environmental economics, vol 2, chapter 16: Property Value models. Elsevier North Holland

  30. Parmeter CF, Pope JC (2009) Quasi-experiments and hedonic property value methods. In: List JA, Price MK (eds) Written for the handbook on experimental economics and the environment. Edward Elgar Publishers, Cheltenham

    Google Scholar 

  31. Pope JC (2008) Buyer information and the hedonic: the impact of a seller disclosure on the implicit price for airport noise. J Urban Econ 63(2):498–516

    Article  Google Scholar 

  32. Rosenbaum PR, Rubin DB (1985) Constructing a control group using multivariate matched sampling methods that incorporate the propensity score. Am Stat 39(1):33–38

    Google Scholar 

  33. Sanders NJ (2014) The response to public information on environmental amenities: New evidence housing markets care about the toxics release inventory. Unpublished manuscript, 6 2014

  34. Schlenker W, Scorse J (2012) Does being a “top 10” worst polluter affect environmental releases? Evidence from the U.S. toxic release inventory. Unpublished manuscript

  35. Sueddeutsche Z (2010) Datenbank der gefahren. Issue: 05/17/2010.

  36. Taylor LO, Phaneuf DJ, Liu X (2016) Disentangling the impacts of environmental contamination from locally undesirable land uses. J Urban Econ 93:85–98

    Article  Google Scholar 

  37. TAZ (2009) Interessantes über Umweltsünder aus der Nachbarschaft - ab sofort kann im Internet jeder nachgucken, mit welchen Schadstoffen Unternehmen die Umwelt belasten. Issue: 06/06/2009:8, 2009. Contrapress

  38. Tiebout C (1956) A pure theory of local expenditures. J Polit Econ 64(5):416–424

    Article  Google Scholar 

  39. Zivin JG, Neidell M, Schlenker W (2011) Water quality violations and avoidance behavior. Am Econ Rev Papers Proc 101:448–453

    Article  Google Scholar 

Download references


We gratefully acknowledge funding from the German Ministry for Education and Research (BMBF) under research grant number 01UN1003. Any opinions expressed in the paper are those of the authors and do not necessarily reflect those of the BMBF. We would like to thank Linda Bui, Dietrich Earnhart, Timo Goeschl, Sabine Grimm, Stephen Kastoryano, Nicolai Kuminoff, Jaren Pope, Nicholas Sanders, Alexander Schürt, V. Kerry Smith, Andrea Weber, the audiences at the 2014 Atlantic Workshop on Energy and Environmental Economics, the 2013 AERE summer conference and the 2013 EAERE conference, as well as seminar participants at the Universities of Heidelberg and Mannheim for their valuable comments. We also thank the editor and two anonymous reviewers for their very constructive comments and suggestions.

Author information



Corresponding author

Correspondence to Kathrine von Graevenitz.

Additional information

An erratum to this article is available at



A.1 Data on Postal Code Area Sociodemographic Characteristics

Table 15 INKAR 2010 Indicators used for matching

Sources for the INKAR database include statistical records on the German federal and state level as well as individual data collections of Federal Ministries, the Eurostat Regio database, GfK (Gesellschaft für Konsumforschung) records, official tax records and the accessibility model of the BBSR (Bundesinstitut für Bau-, Stadt- und Raumforschung). The variables in the data set are described in Table 15. The INKAR data we use is the 2010 publication, which contains information about the municipalities at the end of 2008.

The data on postal code characteristics is available at the municipal level (“Gemeinde”) and for aggregated municipalities (so-called “Gemeindeverbände”), but the unit of analysis is a postal code area as mentioned in the main text. Unfortunately, these two spatial entities do not perfectly overlap. As a result we used GIS software to match postal code areas to municipalities based on the spatial overlay of the two entities. For this purpose we combined a postal code area shape file with a shape file of 11,329 municipalities, both provided by GfK GeoMarketing GmbH as of 2012. A number of municipal reforms have led to restructuring of a few hundred municipalities between 2008 and 2012. As a result, we were unable to merge sociodemographic data to a small number of postal code areas. In total we were able to match 8194 out of our 8212 postal code areas to an INKAR unit.

The spatial overlay analysis was done based on urban area in the municipalities as the INKAR data mainly refers to sociodemographic data such as unemployment levels, tax revenues, and the prevalence of types of residential buildings. Each urban area was identified by the municipality code within which it lay. The postal code area was then allocated to the municipality with the largest share of urban area within the postal code area. Almost 300 postal code areas contained no urban areas. These were matched based on the simple spatial overlap to the municipality with the largest share of the postal code area. Once this allocation was complete based on the 2012 municipalities, we used information on changing municipalities from the German Statistical authorities (Destatis) to trace the municipalities that changed back to their 2008 identification number. For the 8194 postal code areas thus matched to the INKAR units, 4618 where allocated to a single municipality and 3576 to an aggregate (“Gemeindeverband”).

A.2 Construction of Weighted Emission Scores

In order to create postal code quantiles with respect to the severity of treatment (relative severity of emitted substances) we add up the individual emission reports for each postal code weighted by their emission thresholds. For the E-PRTR data release in 2009 this is done according to the following formula:

$$\begin{aligned} {\textit{Weighted}} \ {\textit{emission}} \ {\textit{score}}_{i}(2009)=\sum _{f_{i}=1}^{F_{i}}\left[ \sum _{p=1}^{P}\frac{{\textit{Quantity}}_{p,f_{i}}(2007)}{{\textit{Threshold}}_{p}}\right] \end{aligned}$$

Here, \(f_{i}=1,\ldots ,F_{i}\) denotes the individual facilities in the respective postal code while \(p=1,\ldots ,P\) denotes all substances listed in the database with \(Quantity_{p,f_{i}}(2007)\) being the reported quantity of an individual substance by the respective facility for the year 2007. This emission quantity is weighted by the reporting threshold defined in the E-PRTR regulations. Overall, this measure is a good proxy for the severity of emissions recorded for a certain postal code. As the reporting thresholds are reasonable proxies for toxicity and are also available to the public as possible guidelines for interpreting the values in the database, this aggregate measure constitutes a sound basis for creating treatment quantiles. A few examples of computed Weighted Emission Scores are presented in Fig. 5.

Fig. 5

Examples for the construction of weighted emission scores

A.3 Logit Estimations for Propensity Score Matching

See Tables 16 and 17.

Table 16 Logit estimates for matching, part I
Table 17 Logit estimates, part II

A.4 Development of Housing Prices in Germany

See Fig. 6.

Fig. 6

Price trends (House Price Index), unmatched sample

A.5 Testing the Common Trend Assumption

The common trend assumption is tested by estimating the following regression setup:

$$\begin{aligned} Y_{it}= & {} \beta ^{0}+\beta ^{1}(1-Post_{t})\cdot t+\beta ^{2}Post_{t}\cdot t+\beta ^{3}(1-Post_{t})\cdot Treated{}_{i}\cdot t\nonumber \\&+\,\beta ^{4}Post_{t}\cdot Treated_{i}\cdot t+\beta ^{5}Post_{t}\cdot Treated_{i}+\beta ^{6}Post_{t}+\beta _{i}^{7}+\varepsilon _{it} \end{aligned}$$

The main coefficient of interest is the one for a possible trend divergence prior to treatment for the treatment group (\(\beta ^{3}\), “Pre-Trend*Treatment” in Table 18). Given the graphical representation of the trends in housing prices, a linear trend model is the preferable choice for our analysis. A sharp turn in the development of German housing prices can be seen to coincide roughly with the publication of the first E-PRTR wave and is not restricted to either of the groups but a universal feature of housing prices in Germany, which can be explained by the notion that there has been a surge in housing prices after the recent financial crisis due to other investment options becoming less attractive. Most importantly, we observe in Table 18 that the common trend assumption holds for both parts of Germany after the matching procedure has been completed. Any pre-treatment trend differences that may have prevailed in the unmatched sample between treatment and control group are eliminated by nearest neighbor matching. It is also worth noting that in the matched sample there is no significant trend differential after 2009Q2, which already indicates that the treatment (consisting of at least one E-PRTR report within a postal-code) may have had little effect on the trend in housing prices.

Table 18 Common trend regressions

A.6 Histograms of Propensity Scores

See Figs. 7, 8, 9, 10.

Fig. 7

Propensity scores, Western Germany, no buffer

Fig. 8

Propensity scores, Western Germany, 500 m buffer

Fig. 9

Propensity scores, Eastern Germany, no buffer

Fig. 10

Propensity scores, Eastern Germany, 500 m buffer

A.7 Histograms of Propensity Scores (urban areas only)

See Fig. 11, 12.

Fig. 11

Histograms of propensity scores, urban areas, Western Germany

Fig. 12

Histograms of propensity scores, urban areas, Eastern Germany

A.8 Geographical Distribution of Treatment and Control Group (matched sample)

See Fig. 13.

Fig. 13

Treatment and control groups with NN matching

A.9 Geographical Distribution Treatment and Control Group (matched sample, urban areas only)

See Fig. 14.

Fig. 14

Map of treated areas and matched controls, urban areas

A.10 Mean Characteristics and Bias Comparison for Treatment and Control Group (urban areas only)

See Tables 19, 20.

Table 19 Urban areas only: Treatment and control group before and after matching (A), Western Germany
Table 20 Urban areas only: Treatment and control group before and after matching (A), Eastern Germany

A.11 Summary of Mean Comparisons and Regression Results

See Table 21.

Table 21 ATET versus regression coefficients (full and matched sample)

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

von Graevenitz, K., Römer, D. & Rohlf, A. The Effect of Emission Information on Housing Prices: Quasi-Experimental Evidence from the European Pollutant Release and Transfer Register. Environ Resource Econ 69, 23–74 (2018).

Download citation


  • Emissions information
  • E-PRTR
  • Housing market
  • Propensity score matching
  • Quasi-experiment