Environmental Science & Policy 63 (2016) 1–6

Unintended consequences of the Clean Air Act: Mortality rates inAppalachian coal mining communities

Michael Hendryxa,*, Benjamin Hollandb

aDepartment of Applied Health Science, School of Public Health, Indiana University, Bloomington, IN 47505, USAbDepartment of Environmental Health, School of Public Health, Indiana University, Bloomington, IN 47505, USA

A R T I C L E I N F O

Article history:Received 9 February 2016Received in revised form 28 April 2016Accepted 29 April 2016Available online xxx

Keywords:Mountaintop removalClean Air ActUnintended consequencesMortalityAppalachia

A B S T R A C T

The 1990 amendments to the US Clean Air Act (CAA) encouraged the growth of mountaintop removal(MTR) coal mining in Central Appalachia. This study tests the hypothesis that the amendments hadunintended impacts on increasing mortality rates for populations living in these mining areas. We used apanel design to examine adjusted mortality rates for three groups (all-cause, respiratory cancer, and non-cancer respiratory disease) between 1968 and 2014 in 404 counties stratified by MTR and Appalachian/non-Appalachian status. The results showed significant interactions between MTR status and post-CAAperiod for all three mortality groups. These differences persisted after control for time, age, smokingrates, poverty, obesity, and physician supply. The MTR region in the post-CAA years experienced an excessof approximately 1200 adjusted deaths per year. Although the CAA has benefits, energy policies have ingeneral focused on the combustion portion of the fossil fuel cycle. Other components of fossil fuelproduction (e.g. extraction, transport, and processing) should be considered in the comprehensivedevelopment of sustainable energy policy.

ã 2016 Elsevier Ltd. All rights reserved.

Contents lists available at ScienceDirect

Environmental Science & Policy

journal homepage: www.elsevier .com/ locat e/e nvsci

1. Introduction

Amendments to the Clean Air Act (CAA) were implemented in1990 in the United States with the intent to reduce acid rain andother pollution consequences of burning coal in power plants. Coalreserves in the Central Appalachian region of the United States arerelatively low in sulfur content and became more financiallyattractive after these amendments took effect (Copeland, 2015;Milici, 2000). The motivation to use low sulfur coal consequentlyincreased mining in Central Appalachian areas that were notsuitable to conventional techniques � places where coal reservesare located within steep mountaintops or ridges, often at depths ofhundreds of feet or in multiple thin beds. The approach developedto reach these coals is a form of surface mining called mountaintopmining or mountaintop removal mining.

Mountaintop removal (MTR) occurred on small scales inAppalachia as early as the 1960s, but it became much moreprevalent in the 1990s (Copeland, 2015; Szwilski et al., 2000), andby the current century it had become the largest driver of land-cover alterations in the Central Appalachians (Lindberg et al.,

* Corresponding author.E-mail addresses: hendryx@indiana.edu (M. Hendryx), bendholl@umail.iu.edu

(B. Holland).

http://dx.doi.org/10.1016/j.envsci.2016.04.0211462-9011/ã 2016 Elsevier Ltd. All rights reserved.

2011). In part, the increase in MTR beginning in the 1990s wasbecause the CAA amendments encouraged the use of low sulfurcoal predominant in Central Appalachia (Copeland, 2015; Milici,2000). Speaking with respect to the 1990CAA amendments, theVice-President of the West Virginia Coal Association stated, “It isbecause of the (Environmental Protection Agency’s) action withrespect to the acid rain provisions of the act that allowed for theselarge mountaintop mines to develop and flourish.” (State Journal,2013)

Mountaintop removal mining involves clearcutting forests andusing explosives and heavy machinery to remove up to hundredsof feet of rock and soil above and between coal layers. Theexcavated material creates an “immense quantity of excess spoil”(Copeland, 2015) that is dumped into adjacent valleys, buryingheadwater streams. A single valley fill may be over a 1000 feetwide and a mile long. As early as 1992 the EnvironmentalProtection Agency (EPA) had estimated that 1200 miles ofAppalachian streams had been buried by surface mining. MTRoccurs in close proximity to human settlements and takes place inhundreds of sites over a land area in Central Appalachia roughlyequal in size to the states of New Hampshire and Vermontcombined. The negative impacts of MTR are both socioeconomic(Bell and York, 2010; Hendryx, 2011) and environmental(Bernhardt et al., 2012; Bernhardt and Palmer, 2011; Lindberget al., 2011; Palmer et al., 2010), both of which may contribute to

2 M. Hendryx, B. Holland / Environmental Science & Policy 63 (2016) 1–6

poor public health outcomes for nearby populations. Environ-mental impacts of MTR include impaired air and water quality incommunities proximate to the mine sites (Kurth et al., 2014;Kurth et al., 2015; Orem et al., 2012). MTR sites can be mined withfewer employees per ton of coal extracted relative to other miningforms, and the resulting environmental destruction makes theland unattractive for alternative economic development. Inconsequence, counties where MTR is practiced have lowerincome levels, higher poverty rates, and higher unemploymentrates compared to other parts of the region (Hendryx, 2011;Hendryx and Ahern, 2009).

The 1990 amendments to the CAA have resulted in a number ofbenefits. Since its enactment, reductions in the US have beenobserved for all six of the criteria air pollutants: particulate matter,ozone, lead, carbon monoxide, nitrous oxides and sulfur dioxide.According to the EPA, acid rain decreased 55% between 1990 and2010 (EPA, 2015). These improvements in air quality translate toimprovements in public health, as pollutants from coal combustioncontribute to morbidity and premature mortality (Gohlke et al.,2011; Laden et al., 2000; Lewtas, 2007).

However, well intended public policies sometimes have unin-tended and unanticipatednegative consequences(IOM, 2001; Peterset al., 2013). The CAA itself may have provided unintendeddisincentives to promote development of cleaner power plants (Listet al., 2004). Other instances exist in areas of agriculture (Karp et al.,2015), health care (Naylor et al., 2012; Song et al., 2013) andeducation policy (Metos et al., 2015) where unintended negativeconsequences resulted from well-meaning policy interventions.

Previous research on the public health impacts of mountaintopremoval mining has demonstrated that mortality rates are higher inMTR communities compared to control communities in ways notexplained by age, smoking, obesity, socioeconomic status or otherrisks. Elevated rates have been observed for all-cause mortality(Hendryx, 2011; Hendryx and Ahern, 2009), heart, lung and kidneydisease (Hendryx, 2009), and some types of cancer (Ahern andHendryx, 2012). However, the previous mortality studies werelimited to a narrow range of years and did not examine possible CAAeffects. The current study extends prior research by testing a specifichypothesis regarding possible unintended consequences of the CAA.We employ a panel analysis design to use counties as their owncontrols to examine mortality rates pre- and post-CAA in MTR andcontrol areas. If CAA-dependent mortality differences are detected,then theyare not dueto sociodemographic differences in MTR versusother areas to the extent that the pre-CAA observations in the MTRarea serve as an internal control. We also have group comparisons toexamine CAA effects in the MTR region compared to other regions.We examine mortality rates for a 47 year period from 1968 through2014 to test whether all-cause mortality in MTR areas of CentralAppalachia increased in the post-CAAyears as this method of miningbecame predominant.

2. Methods

2.1. Design

The study is a secondary analysis of publicly available county-level data. Annual age-adjusted mortality rates for 1968–2014 areinvestigated in relationship to mountaintop removal mining inCentral Appalachia and the implementation of the 1990 amend-ments to the Clean Air Act (CAA).

The study area consists of the four states where mountaintopremoval mining has been practiced, including Kentucky, Tennessee,Virginia and West Virginia. Counties within these states wereclassified into three groups: those where any amount of mountain-top removal coal mining had been practiced, other counties inAppalachia without mountaintop removal, and the remaining non-

Appalachiancounties inthosestates(the latterusedasthereferentinstatistical models). The Appalachian non-MTR group provides acontrol for general Appalachian effects. Mountaintop removalcounties were identified using satellite imagery as reported inearlier papers (Esch and Hendryx, 2011) and confirmed using EnergyInformation Administration data on tons of coal mined from surfacemines (EIA, 2016). Appalachian counties were identified based onAppalachian Regional Commission designations in place in 2010.

2.2. Measures

Age-adjusted all-cause mortality rates for the four states wereobtained from the Centers for Disease Control and Prevention(CDC, 2016). Mortality rates are per 100,000 persons and are age-adjusted to the 2000 US population. Mortality rates werereported for each county on an annual basis for the years1968–2014.

In addition to total mortality, we also examined age-adjustedmortality rates for two diagnostic classes, including respiratorycancer, and all other non-cancer respiratory disease. These classeswere selected because of prior evidence that MTR activitiesgenerate local air pollution (Kurth et al., 2014; Kurth et al., 2015)and may promote poor health outcomes for these conditions(Christian et al. 2011; Hendryx et al., 2008; Hendryx, 2009;Hendryx and Luo, 2015). Toxicological data suggest ultrafineparticulate matter, a chief air pollutant from MTR mining, maypromote pulmonary inflammation, oxidative stress, and Ca++

influx within lung cells (Donaldson et al., 2004). These may act asa mechanism for long-term, delayed, neoplastic promotion. Ininstances of small numbers of cases in counties within singleyears, the CDC suppresses the data values to protect patientconfidentiality. For this reason, we aggregated mortality rates forthese diagnostic groups into five-year blocks to increase casenumbers and eliminate suppressed values.

Data on covariates were obtained from the County HealthRankings data for 2015 (County Health Rankings, 2016). Eachcounty had single cross-sectional measures for the adult smokingrate, obesity rate, child poverty rate, and per capita supply ofprimary care physicians. In a few instances where covariate datawere missing for the county, the missing observations werereplaced with state averages.

Descriptive summaries of annual age-adjusted mortality rateswere found for the three county groups (MTR, other Appalachian,and other). Then, a panel design analysis was conducted toinvestigate age-adjusted mortality rates in relationship to time,county group, covariates, and implementation of the CAA amend-ments of 1990. The years 1968–1989 were designated as pre-CAAand the years 1990–2014 were designated as post-CAA. Theanalyses were conducted using SAS version 9.4 Proc Mixed,specifying the year as a repeated measure, and the county, countygroup, and CAA dummy variable as class variables. An autore-gressive value of 1 was specified to account for correlated mortalityrates from one year to the next. We first tested a model with maineffects for year, the CAA dummy variable, and county group. Asecond model added covariates. A final model was estimated afteradding an interaction term between county group and the CAAindicator. The final interaction term tests whether age-adjustedmortality rates were significantly higher in the MTR areas in thepost-CAA period while controlling for covariates and year-to-yeartrends.

3. Results

Data for a total of 404 counties were available for the analysis,including 37 MTR counties, 149 other Appalachian counties, and

600

700

800

900

1000

1100

1200

1300

1400

150019

6819

7019

7219

7419

7619

7819

8019

8219

8419

8619

8819

9019

9219

9419

9619

9820

0020

0220

0420

0620

0820

1020

1220

14

MTR Other Appalachian Non-Appalachian

Pre -CAA Post-CAA

Fig. 1. Age-adjusted total mortality rates per 100,000 by county group, 1968–2014.

M. Hendryx, B. Holland / Environmental Science & Policy 63 (2016) 1–6 3

218 remaining counties in the four states. Annual all-cause age-adjusted mortality rates for the three groups are shown in Fig. 1.The figure shows that mortality rates have been higher in MTRcounties throughout the time period, but that the differencebetween MTR and other county groups appears to have increasedin more recent years. A reduction in annual mortality rates for allthree groups is evident from 1968 to about 1980, consistent withnational trends (Hoyert, 2012). The period roughly from 1980 tothe early 1990s shows little change for all groups, and in the latteryears the rates for the MTR counties remained flat while rates inthe other groups declined.

The rates in Fig. 1 correspond to Model 1 results with maineffects and no covariates summarized in the left hand column ofTable 1. Model 1 shows a significant yearly decline in mortalityrates overall, and a significant positive overall effect during the

Table 1Summary of model results, dependent variable is total age-adjusted mortality per 100

Model 1

Independent variables Coefficient (SE) P<

Year �9.41 (0.22) 0.0001

Post-CAA 34.12 (5.45) 0.0001

Pre-CAA (referent) –

MTR counties 134.40 (6.96) 0.0001

Other Appalachian counties 13.65 (4.19) 0.0001

Other counties (referent) –

Smoking rate –

Obesity rate –

Poverty rate –

Per capita primary care physicians –

Year*MTR*Post-CAA –

Year*MTR*Pre-CAA

Year*Other App.*Post-CAA

Year*Other App.*Pre-CAA

Year*Other counties*Post-CAA

Year*Other counties*Pre-CAA (referent) –

AIC fit 230314.3

Model 1: Main effects and no covariates. Null model likelihood ratio test chi-square = 7Model 2: Covariates added. Null model likelihood ratio test chi-square = 5502.0 (df = 1)Model 3: Interaction term added. Null model likelihood ratio test chi-square = 4117.7 (dfwith chi-square = 981.4 (df = 5), p < 0.0001.

post-CAA period. The model also shows significant effects for theMTR and other Appalachian groups compared to the referent, witha much larger mortality coefficient for the MTR area.

Model 2 adds covariates, and results show that the covariateshad significant effects but did not change the pattern for the othereffects, with one exception. The coefficient for the non-MTRAppalachian counties switched from positive to negative. Thisindicates that the covariates (smoking, poverty, obesity, andphysician supply) could account for higher mortality in the non-MTR Appalachian region, but could only partially account for thehigher mortality in the MTR area.

Model 3 shows the effects of adding the interaction term. Themain effect for the post-CAA variable changed from positive tonegative. The main effects for both the MTR and non-MTRAppalachian regions were also negative. The effects of thecovariates remained essentially unchanged. Compared to thereferent category in the interaction, four of the five groups showedsignificantly elevated mortality rates. The largest estimate was forthe post-CAA MTR group, followed by the post-CAA non-MTRAppalachian group. Although not shown in the Table, the 95%confidence interval for the post-CAA, MTR estimate (16.30, 95%CI = 15.03–17.57) was completely above the confidence intervals forall of the remaining groups, including the post-CAA non-MTRAppalachian estimate (13.72, 95% CI = 12.59–14.83), and the pre-CAA MTR estimate (3.47, 95% CI = 1.65–5.28).

Analyses were repeated for the two diagnostic classes,including respiratory cancer and other respiratory disease. Similarsignificant model results were found for both groups. Table 2shows the final Model 3 results adjusting for covariates for eachdiagnostic class. Results show significant positive effects for higheradjusted mortality rates in the MTR counties post-CAA. Forrespiratory cancer, effects for the remaining county groupscompared to the non-Appalachian referent are negative. For othernon-cancer respiratory disease, the other county groups some-times also show significant positive effects, but the coefficient forthe MTR-post CAA group (38.07) is and its 95% confidence interval(34.86–41.27) falls completely above the CIs for the other countygroups (confidence intervals not shown in table).

We conducted a final analysis to estimate the number of excessadjusted all-cause deaths that occurred in the MTR region in thepost-CAA period. To do this we found least squares means for thetwo way county group*CAA interaction, controlling for main

,000.

Model 2 Model 3

Coefficient (SE) P< Coefficient (SE) P<�9.28 (0.20) 0.0001 �17.24 (0.41) 0.000134.15 (5.20) 0.0001 �214.9 (10.10) 0.0001– –

53.42 (6.93) 0.0001 �43.57 (12.32) 0.001�24.10 (4.13) 0.0001 �61.66 (7.67) 0.0001– –

303.4 (36.07) 0.0001 302.1 (24.03) 0.0001459.8 (61.94) 0.0001 466.8 (52.69) 0.0001369.2 (28.25) 0.0001 378.2 (24.03) 0.00010.16 (0.06) 0.02 0.18 (0.05) 0.001– 16.30 (0.65) 0.0001– 3.47 (0.93) 0.0002– 13.72 (0.57) 0.0001– 1.05 (0.58) 0.07– 12.00 (0.5) 0.0001– –

199573.8 198720.2

031.4 (df = 1), p < 0.0001., p < 0.0001. = 1), p<0.0001. The overall effect of the three-way interaction term was significant

Table 2Final model results for respiratory cancer and for other non-cancer respiratory disease.

Respiratory Cancer Non-Cancer Respiratory Disease

Independent variables Coefficient (SE) P< Coefficient (SE) P<Year �1.13 (0.36) 0.002 1.21 (0.26) 0.0001Post-CAA 13.72 (2.09) 0.0001 8.02 (1.50) 0.0001Pre-CAA (referent) – –

Smoking rate 58.69 (9.81) 0.0001 63.97 (7.04) 0.0001Obesity rate 88.95 (16.84) 0.0001 12.50 (12.09) 0.31Poverty rate 50.59 (7.66) 0.0001 33.18 (5.50) 0.0001Per capita primary care physicians 0.10 (0.02) 0.0001 �0.01 (.01) 0.43MTR*Post-CAA 11.75 (2.28) 0.0001 38.07 (1.64) 0.0001MTR*Pre-CAA �15.54 (2.86) 0.0001 24.58 (2.06) 0.0001Other App.*Post-CAA �6.92 (1.38) 0.0001 6.49 (0.99) 0.0001Other App.*Pre-CAA �17.68 (1.76) 0.0001 0.07 (1.26) 0.96Other counties (referent) – –

4 M. Hendryx, B. Holland / Environmental Science & Policy 63 (2016) 1–6

effects of year, CAA, county group and the covariates. The estimatesprovided in Table 3 show annual age-adjusted deaths per100,000 by county group and CAA period.

The result for the MTR region in the post-CAA period,1158.7 deaths per year, is approximately 98 deaths more per100,000 per year compared to the pre-CAA MTR effect, andapproximately 101 deaths more per 100,000 per year compared tothe other Appalachian counties in the post-CAA period. Given thatthe population of the 37 counties where MTR is practiced totals1,204,469 people as provided in the County Rankings Data, thisequates to an estimate of 1180 to 1217 additional deathsexperienced every year in the MTR region in the post-CAA period,controlling for age and other covariates.

4. Discussion

The results of the study indicate that mortality rate disparitiesin mountaintop removal coal mining areas of Appalachia becamesignificantly more pronounced in the years after the introductionof the 1990 amendments to the Clean Air Act. Counties outside ofthe MTR zone in these Central Appalachian states experienced adecline in total age-adjusted mortality rates that was not matchedby declines in the MTR zone. This divergence becomes mostapparent about 10 years after the CAA revisions encouraged thegrowth of MTR mining. The 10 year delay may reflect longer termconsequences of exposures over time.

Death rates were higher in the MTR region throughout theentire study period. Even before this area of Central Appalachiabegan to engage in large scale MTR, it was already characterized byrelatively heavy amounts of coal mining using older techniques,including underground and other surface mining methods. Prior tothe widespread practice of MTR, these areas were experiencinghealth disadvantages related perhaps to the poor socioeconomicconditions that characterize mining-dependent economies, or toenvironmental impacts from other mining techniques.

We also observed a smaller deleterious post-CAA effect for thenon-MTR Appalachian region. These areas also experiencedrelatively higher mortality rates compared to the non-Appalachian

Table 3Least squares means for age-adjusted deaths per year per 100,000 population,controlling for year, county group, CAA period, smoking, obesity, poverty, andprimary care physician supply.

Effect Least square means (standard error)

MTR*Post-CAA 1158.7 (7.8)MTR*Pre-CAA 1060.1 (8.3)Other App.*Post-CAA 1057.8 (4.3)Other App.*Pre-CAA 1009.2 (4.6)Other counties*Post-CAA 1065.1 (3.8)Other counties*Pre-CAA (referent) 1052.8 (4.1)

referent group. This finding may reflect the effects of non-MTRmining that occurs in portions of this region, or it may reflect MTRspillover effects from environmental or socioeconomic impair-ments that do not obey county lines and impact neighboringpopulations.

Significant post-CAA effects were also observed for respiratory(mostly lung) cancer and other non-cancer respiratory disease. Theeffects appear to be most pronounced for respiratory cancer. Otherrespiratory diseases include chronic conditions such as bronchitisand emphysema that may be related to long term air pollutionexposures but also includes acute illness such as pneumonia thatmay reflect other etiologies.

This study does not assess environmental conditions in miningcommunities. Other research has documented that air and waterpollution are impaired in communities proximate to mountaintopremoval coal mining (Kurth et al., 2014; Kurth et al., 2015; Oremet al., 2012). Air contaminates in mining communities includeelevated levels of silica, other inorganics, and polycyclic aromatichydrocarbons. Elevations in ultrafine counts have also beenobserved. Subsequent laboratory-based studies have documentedbiological impairments from exposure to mountaintop miningparticulate matter (Knuckles et al., 2013; Luanpitpong et al., 2014;Nichols et al., 2015). The unintended impacts that may haveresulted from the CAA likely include environmental harm causedby this aggressive form of surface mining, but in addition mayinclude health consequences secondary to the social and economiccosts associated with this land use choice. These social andeconomic costs include reductions in alternative employmentopportunities with corresponding increases in socioeconomicdisadvantage.

It is worth noting that large scale surface coal mining occurs inmany countries around the world. Studies have identified airquality problems around surface mining in India, Columbia,Australia, China, and Great Britain (Ghose 2007; Reynolds et al.,2003; Higginbotham et al., 2010; Huertas et al., 2012; Liu et al.,2012). Public health problems near surface mining have also beennoted outside the US (Temple and Sykes, 1992; Yapici et al., 2006;Liao et al., 2010). Possible unintended consequences of coal miningactivities globally should be considered in the development ofenergy policy, as described later in the paper.

Limitations of the study include the ecological design,imperfect capture of covariates, and uncertain temporal relation-ships between possible exposures and health outcomes. Thecounty level ecological data prohibit drawing conclusions aboutexposure-mortality consequences for individuals; we can onlycomment on county aggregate relationships. Covariates weremeasured from a single time point, although county differences onpopulation sociodemographic and behavioral variables tend to bestable over time. The relationship between the implementation of

M. Hendryx, B. Holland / Environmental Science & Policy 63 (2016) 1–6 5

the CAA amendments and observed mortality differences wasimperfect. In particular, there was a period even before theintroduction of the amendments when mortality rates hadflattened in MTR areas; the difference only becomes apparentabout 10 years after the amendments, when rates began to divergemore noticeably between MTR counties and other places. Thisdivergence may reflect delayed effects of chronic exposure overtime, but cannot be known with certainty. We cannot concludedefinitively with this non-experimental design that the CAAamendments caused the additional mortality disparity, but giventhe evidence that the CAA promoted MTR activity, and theenvironmental and public health evidence for MTR’s impacts, it is apossible contributing factor. Other socioeconomic and environ-mental factors pre-dating or co-occurring with the CAA periodcontribute to health disparities in MTR areas too. Despite theuncertainty, the results provide an illustration of the importance ofconsidering full production costs from fossil fuels in the develop-ment of energy policy.

4.1. General lessons

Although the issue of mountaintop removal may appear to beconfined to Central Appalachia, there are general lessons that canbe learned regarding the need for energy policies to address theentire fossil fuel sequence of extraction, processing, transporta-tion, combustion and disposal. These lessons have relevance tothe US but also to other nations where coal mining and otherfossil fuel extraction activities are practiced. When consideringimpacts from pollution and responses to climate change, energypolicy in the US has often focused exclusively on combustion,largely ignoring other components of production and use. Recentplans by the US Environmental Protection Agency (EPA) toreduce carbon emissions to fight climate change focus on coal-fired power plants, and do not consider how coal reaches thosepower plants in the first place (EPA, 2016). Epstein et al. (2011),however, have shown that the entire production process of usingcoal adds considerably to its full environmental costs. Fox andCampbell (2010) estimated that mountaintop mining adds theequivalent of 7–17% of conventional power plant carbonemissions to the atmosphere. Reducing reliance on coal in favorof natural gas has been promoted as a cleaner approach to powergeneration, but again, policy attention to the environmental andpublic health impacts of natural gas extraction and processing,not just its consumption, has been missing. Hydraulic fracturingfor natural gas in shale formations harms local environments andmay impair public health (Rabinowitz et al., 2015); furthermore,it has been estimated that shale gas has a greater greenhouseeffect than conventional gas or coal when considering the totalproduction cycle (Howarth et al., 2011). Future policies in the USand globally that are designed to reduce climate change andpromote energy sustainability must consider the completeproduction cycles of various fuels and not focus only on fuelcombustion.

References

Ahern, M., Hendryx, M., 2012. Cancer mortality rates in Appalachian mountaintopcoal mining areas. J. Occup. Environ. Sci. 1, 63–70.

Bell, S.E., York, R., 2010. Community economic identity: the coal industry andideology construction in West Virginia. Rural Sociol. 75, 111–143.

Bernhardt, E.S., Palmer, M.A., 2011. The environmental costs of mountaintop miningvalley fill operations for aquatic ecosystems of the Central Appalachians. Ann. N.Y. Acad. Sci. 1223, 39–57.

Bernhardt, E.S., Lutz, B.D., King, R.S., Fay, J.P., Carter, C.E., Helton, A.M., Campagna, D.,Amos, J., 2012. How many mountains can we mine? assessing the regionaldegradation of Central Appalachian rivers by surface coal mining. Environ. Sci.Technol. 46, 8115–8122.

CDC, 2016. Compressed Mortality File. Centers for Disease Control and Prevention. .(accessed 01.19.16.) http://wonder.cdc.gov/mortSQL.html.

Christian, W.J., Huang, B., Rinehart, J., Hopenhayn, C., 2011. Exploring geographicvariation in lung cancer incidence in Kentucky using a spatial scan statistic:elevated risk in the Appalachian coal-mining region. Public Health Rep. 126,789–796.

Copeland, C., 2015. Mountaintop Mining: Background on Current Controversies.Congressional Research Service, Washington DC.

County Health Rankings, 2016. County Health Rankings and RoadmapsUniversity ofWisconsin Population Health Institute and Robert Wood Johnson Foundation. .(accessed 02.01.16.) http://www.countyhealthrankings.org/.

Donaldson, K., Brown, D., Clouter, A., Duffin, R., MacNee, W., Renwick, L., Tran, L.,Stone, V., 2004. The pulmonary toxicity of ultrafine particles. J. Aerosol Med. 15,213–220.

EIA, 2016. Coal. US Department of Energy, Energy Information Administration. .(accessed 02.01.16.) http://www.eia.gov/coal/.

EPA, 2015. Progress Cleaning the Air and Improving People’s HealthUSEnvironmental Protection Agency. . (accessed 02.03.16.) http://www.epa.gov/clean-air-act-overview/progress-cleaning-air-and-improving-peoples-health.

EPA, 2016. The Clean Power Plan. Environmental Protection Agency. . (accessed02.03.16.) http://www.epa.gov/cleanpowerplan.

Epstein, P., Buonocore, J.J., Eckerle, K., Hendryx, M., Stout, B.M., Heinberg, R.,Glustrom, L., 2011. Full cost accounting for the life cycle of coal. Ann. N. Y. Acad.Sci. 1219, 73–98.

Fox, J.F., Campbell, J.E., 2010. Terrestrial carbon disturbance from mountaintopmining increases lifecycle emissions for clean coal. Environ. Sci. Technol. 44,2144–2149.

Ghose, M.K., 2007. Generation and quantification of hazardous dusts from coalmining in the Indian context. Environ. Monit. Assess. 130, 35–45.

Gohlke, J.M., Thomas, R., Woodward, A., Campbell-Lendrum, D., Pruss-Ustun, A.,Hales, S., Portier, C.J., 2011. Estimating the global public health implications ofelectricity and coal consumption. Environ. Health Perspect. 119, 821–826.

Hendryx, M., Ahern, M., 2009. Mortality in Appalachian coal mining regions: thevalue of statistical life lost. Public Health Rep. 124, 541–550.

Hendryx, M., Luo, J., 2015. An examination of the effects of montaintop removal coalmining on respiratory symptoms and COPD using propensity scores. Int. J.Environ. Health Res. 25, 265–276.

Hendryx, M., 2009. Mortality from heart, respiratory, and kidney disease in coalmining areas of Appalachia. Int. Arch. Occup. Environ. Health 82, 243–249.

Hendryx, M., 2011. Poverty and mortality disparities in central Appalachia:mountaintop mining and environmental justice. J. Health Dispar. Res. Pract. 4,44–53.

Hendryx, M., O’Donnell, K., Horn, K., 2008. Lung cancer mortality is elevated in coal-mining areas of Appalachia. Lung Cancer 62, 1–7.

Higginbotham, N., Freeman, S., Conner, L., Albrecht, G., 2010. Environmentalinjustice and air pollution in coal affected communities, Hunter Valley,Australia. Health Place 16, 259–266.

Howarth, R.W., Ingraffea, A., Engelder, T., 2011. Natural gas: should fracking stop?Nature 477, 271–275.

Hoyert, D.L., 2012. 75 Years of Mortality in the United States, 1935–2010, NCHS DataBrief, No. 88. Centers for Disease Control and Prevention. National Center forHealth Statistics, Hyattsville, MD.

Huertas, J.I., Huertas, M.E., Izquierdo, S., Gonzalez, E.D., 2012. Air quality impactassessment of multiple open pit coal mines in northern Colombia. J. Environ.Manage. 93, 121–129.

IOM, 2001. Unintended consequences of health policy programs and policies:workshop summary. Institute of Medicine (US) Robert Wood Johnson HealthPolicy Fellowship Program, National Academies Press Washington DC.

Karp, D.S., Baur, P., Atwill, E.R., DeMaster, K., Gennet, S., Iles, A., Nelson, J.L., Sciligo, A.R., Kremen, C., 2015. The unintended ecological and social impacts of food safetyregulations in California’s central coast region. BioScience 65, 1173–1183.

Knuckles, T., Stapleton, P., Minarchick, V., McCawley, M.A., Hendryx, M., Esch, L.,Nurkeiwicz, T., 2013. Air pollution particulate matter collected from anAppalachian mountaintop mining site induces microvascular dysfunction.Microcirculation 20, 158–169.

Kurth, L.M., McCawley, M.A., Hendryx, M., Lusk, S., 2014. Atmospheric particulatematter size distribution and concentration in West Virginia coal mining andnon-mining areas. J. Expo. Sci. Environ. Epidemiol. 24, 405–411.

Kurth, L., Kolker, A., Engle, M., Geboy, N., Hendryx, M., Orem, W., McCawley, M.,Crosby, L., Tatu, C., Varonka, M., DeVera, C., 2015. Atmospheric particulatematter in proximity to mountaintop coal mines: sources and potentialenvironmental and human health impacts. Environ. Geochem. Health 37, 529–544.

Laden, F., Neas, L.M., Dockery, D.W., Schwartz, J., 2000. Association of fine particulatematter from different sources with daily mortality in six U.S. cities. Environ.Health Perspect. 108, 941–947.

Lewtas, J., 2007. Air pollution combustion emissions: characterization of causativeagents and mechanisms associated with cancer, reproductive, andcardiovascular effects. Mutat. Res. 636, 95–133.

Liao, Y., et al., 2010. Spatial analysis of neural tube defects in a rural coal mining area.Int. J. Environ. Health Res. 20, 439–450.

Lindberg, T.T., Bernhardt, E.S., Bier, R., Helton, A.M., Merola, R.B., Vengosh, A., DiGiulio, R.T., 2011. Cumulative impacts of mountaintop mining on anAppalachian watershed. PNAS 108, 20929–20934.

List, J.A., Millimet, D.L., McHone, W., 2004. The unintended disincentive in the cleanair act. Adv. Econ. Anal. Policy 4.

6 M. Hendryx, B. Holland / Environmental Science & Policy 63 (2016) 1–6

Liu, J., Liu, G., Zhang, J., Yin, H., Wang, R., 2012. Occurrence and risk assessment ofpolycyclic aromatic hydrocarbons in soil from the Tiefa coal mine district,Liaoning, China. J. Environ. Monit. 14, 2634–2642.

Luanpitpong, S., Chen, M., Knuckles, T., Wen, S., Luo, J., Ellis, E., Hendryx, M.,Rojanasakul, Y., 2014. Appalachian mountaintop mining particulate matterinduces neoplastic transformation of human bronchial epithelial cells andpromotes tumor formation. Environ. Sci. Technol. 48, 12912–12919.

Metos, J., Owen, R., Utz, R., 2015. High school nutrition policy: case study of intendedand unintended consequences. FASEB J. 29 (Suppl. 1) 273.2.

Milici, R.C., 2000. Depletion of Appalachian coal reserves—how soon? Int. J. CoalGeol. 44, 251–266.

Naylor, M.D., Kurtzman, E.T., Grabowski, D.C., Harrington, C., McClellan, M.,Reinhard, S.C., 2012. Unintended consequences of steps to cut readmissions andreform payment may threaten care of vulnerable older adults. Health Aff.(Millwood) 31, 1623–1632.

Nichols, C.E., Shepherd, D.L., Knucles, T.L., Stricker, J.C., Stapleton, P.A., Minarchick, V.C., Erdely, A., Zeidler-Erdely, P.C., Alway, S.E., Nurkeiwicz, T.R., Hollander, J.M.,2015. Cardiac and mitochondrial dysfunction following acute pulmonaryexposure to mountaintop removal mining particulate matter. Am. J. Physiol.Heart Circ. Physiol. 309, H2017–H2030.

Orem, W., Tatu, C., Crosby, L., Varonka, M.S., Bates, A., Engle, M., Geboy, N.J., Hendryx,M., 2012. Water chemistry in areas with surface mining of coal. GeologicalSociety of America Annual Meeting and Exposition, Charlotte,NCace Mining ofCoal Geological Society of America Annual Meeting and Exposition,Charlotte, NC.

Palmer, M.A., Bernhardt, E.S., Schlesinger, W.H., Eshleman, K.N., Foufoula-Georgiou,E., Hendryx, M., Lemly, A.D., Likens, G.E., Loucks, O.L., Power, M.E., White, P.S.,Wilcock, P.R., 2010. Mountaintop mining consequences. Science 327, 148–149.

Peters, E., Klein, W., Kaufman, A., Meilleur, L., Dixon, A., 2013. More is not alwaysbetter: intuitions about effective public policy can lead to unintendedconsequences. Soc. Issues Policy Rev. 7, 114–148.

Rabinowitz, P.M., Slizovskiy, I.B., Lamers, V., Trufan, S.J., Holford, T.R., Dziura, J.D.,Peduzzi, P.N., Kane, M.J., Reif, J.S., Weiss, T.R., Stowe, M.H., 2015. Proximity tonatural gas wells and reported health status: results of a household survey inWashington County Environmental Pennsylvania. Health Perspect. 123, 21–26.

Reynolds, L., Jones, T.P., BeruBe, K.A., Wise, H., Richards, R., 2003. Toxicity of airbornedust generated by opencast coal mining. Mineral. Soc. Great Br. Ireland 67, 141–152.

Song, Z., Ayanian, J.Z., Wallace, J., He, Y., Gibson, T.B., Chernew, M.E., 2013.Unintended consequences of eliminating Medicare payments for consultations.JAMA Intern. Med. 173, 15–21.

State Journal, 2013. Appalachian basin coal seems to be struggling more than coalfrom other regions., State Journal, Charleston, WV. March 2, 2013. Available athttp://www.statejournal.com/story/20881915/appalachian-basin-coal-seems-to-be-struggling-more-than-coal-from-other-regions (accessed 01.20.16.).

Szwilski, T.B., Dulin, B.E., Hooper, J.W., 2000. Mountaintop coal mining in WestVirginia: environmental issues. In: Singhal, R.K., Mehrotra, A.K. (Eds.),Environmental Issues and Management of Waste in Energy and MineralProduction. A.A. Balkemal, Rotterdam, pp. 93–98.

Temple, J.M.F., Sykes, A.M.,1992. Asthma and open cast mining. Br. Med. J. 305, 396–397.

Yapici, G., Can, G., Kiziler, A.R., Aydemir, B., Timur, I.H., Kaypmaz, A., 2006. Lead andcadmium exposure in children living around a coal-mining area in Yatagan,Turkey. Toxicol. Ind. Health 22, 357–362.