Climate Amenities, Climate Change, and American Quality of Life

72 Pages·2016·3.58 MB·English

Climate Amenities, Climate Change,
and American Quality of Life
DavidAlbouy,WalterGraf,RyanKellogg,HendrikWolff
Abstract: We present a hedonic framework to estimate US households’ prefer-
encesoverlocalclimates,usingdetailedweatherand2000Censusdata.Wefindthat
Americansfavoradailyaveragetemperatureof65degreesFahrenheit,thattheywill
paymoreonthemargintoavoidexcessheatthancold,andthatdamagesincrease
lessthanlinearlyoverextremecold.Thesepreferencesvarybylocationduetosort-
ingoradaptation.Changesinclimateamenitiesunderbusiness-as-usualpredictions
implyannualwelfarelossesof1%–4%ofincomeby2100,holdingtechnologyand
preferencesconstant.
JELCodes:H49,I39,Q54,R10
Keywords: Amenityvaluation,Climatechange,Hedonicmodel,Householdsorting,
Impactassessment,Qualityoflife
THE CHEMISTRY OF THE HUMAN BODY makes our health and comfort sensitive
toclimate.Everyday,climateinfluenceshumanactivity,includingdiet,chores,recre-
ation,andconversation.Householdsspendconsiderableamountsonhousing,energy,
clothing, and travel to protect themselves from extreme climates and to enjoy com-
DavidAlbouyisattheUniversityofIllinois,DepartmentofEconomics,andNBER(albouy
@illinois.edu).WalterGrafisattheUniversityofCalifornia,Berkeley,DepartmentofAgri-
cultural and Resource Economics ([email protected]). Ryan Kellogg (corresponding
author) is at the University of Michigan, Department of Economics, and NBER (kelloggr
@umich.edu). Hendrik Wolff is at Simon Fraser University, Department of Economics
([email protected]).AlbouyacknowledgesfinancialsupportfromNSFgrantSES-0922340.
We thank Kenneth Chay, Don Fullerton, Philip Haile, Kai-Uwe Kühn, Matthew Turner,
WolframSchlenker,theeditorDanielPhaneuf,andthreeanonymousrefereesforparticularly
helpful criticisms and suggestions. We are also grateful for comments from seminar partici-
pants at Boston, Brown, Calgary, Columbia, CU-Boulder, Duke, EPA, Harvard, Houston,
Illinois, Maryland, Michigan, Michigan State, Minnesota, NYU, Olin, Oregon, Princeton, RFF,
Rice,Stanford,TexasA&M,TREE,UCBerkeley,UCLA,UNR,Washington,WesternMichigan,
andWisconsin,aswellasconferenceattendeesattheAmericanEconomicAssociation,Cowles,
ReceivedJuly23,2014;AcceptedOctober16,2015;PublishedonlineJanuary22,2016.
JAERE,volume3,number1.©2016byTheAssociationofEnvironmentalandResourceEconomists.
Allrightsreserved.2333-5955/2016/0301-0006$10.00 http://dx.doi.org/10.1086/684573
205
206 JournaloftheAssociationofEnvironmentalandResourceEconomists March2016
fortable moderation. Geographically, climate affects the desirability of different loca-
tions and the quality of life they offer; few seek to live in the freezing tundra or op-
pressively hot deserts. Given the undeniable influence climate has on economic de-
cisions and welfare, we seek to estimate the dollar value American households place
onclimateamenities,withafocusontemperature.
Valuingclimateamenitiesnotonlyhelpsustounderstandhowclimateaffectswel-
fareandwherepeoplelivebutalsohelpstoinformpolicyresponsestoclimatechange.
Globalclimatechangethreatenstoalterlocalclimates,mostobviouslybyraisingtem-
peratures.Apriori,thewelfareimpactsofhighertemperaturesareambiguous:house-
holds may suffer from hotter summers but benefit from milder winters. Ultimately,
theseimpactsdependonwherehouseholdsarelocated,thechangesinclimateameni-
tiestheyexperience,andhowmuchtheyvaluethesechanges.
In this paper, we estimate the value of climate amenities in the United States by
examininghowhouseholds’willingnesstopay(WTP)toliveindifferentareasvaries
with climate in the cross-section. Following the intuition laid out by Rosen (1974,
1979)andRoback(1982),andlaterrefinedbyAlbouy(2012),wemeasureWTPby
developingalocalquality-of-life(QOL)indexbasedonhowmuchhouseholdspayin
costsoflivingrelativetotheincomestheyreceive.TheUnitedStatesisaparticularly
appropriate setting in which to use this method as it has a large population that is
mobileoverareaswith diverseclimates.Globally,the UnitedStatesliesinatemper-
ate zone, with some areas that are quite hot (Arizona) while others are quite cold
(Minnesota), and some with extreme seasonality (Missouri) while others are mild
yearround(coastalCalifornia).Thisvariationallowsustoidentifypreferencesovera
broadrangeofhabitableclimates.
Weadoptthishedonicapproachastherearenoexplicitmarketsforclimateame-
nities,onlyanimplicitmarketbasedonhouseholdlocationchoices.Ourestimatesof
amenity values primarily reflect impacts of exposure to climate on comfort, activity,
and health, including time use (Graff Zivin and Neidell 2012) and mortality risk
(DeschênesandGreenstone2011;Barrecaetal.2015).Theyexcludecostsfromres-
idential heating, cooling, and insulation. As such, the value of the climate amenities
we estimate does not appear in national income accounts and so neither would the
impactofclimatechangeontheseamenities.Ourstudythereforecomplementswork
that assesses how climate directly affects national income through agricultural and
urban productivity (for a survey of the climate and productivity literature, see Tol
2002,2009).
We adopt a cross-sectional estimation strategy in the tradition of Mendelsohn,
Nordhaus, and Shaw (1994) rather than a time-series panel approach for several
NorthAmericansummermeetingoftheEconometricSociety,NorthAmericanRegionalSci-
enceCouncil,NationalBureauofEconomicResearch,EnvironmentalandEnergyEconomics,
ResearchSeminarinQuantitativeEconomics,andWorldCongressofEnvironmentalandRe-
sourceEconomists.
ClimateChangeandAmericanQualityofLife Albouyetal. 207
reasons. First, yearly changes in weather are unlikely to affect households’ WTP to
liveinanarea:WTPshoulddependonlong-runclimateratherthantheoutcomeof
theweatherinthemostrecentyear. Second,lowfrequencychangesare notveryin-
formative as long-run secular climate changes so far have been slight, particularly
relativetochangesintechnology—especiallyairconditioning(Barrecaetal.2015)—
and local economic conditions. Third, Kuminoff and Pope (2014) have shown that
temporalchangesinthecapitalizationofamenitiesdonottypicallytranslatetomea-
sures of WTP. Finally, households can mitigate potential damages from climate
throughadaptation—say,byinsulatinghomes,changingwardrobes,oradoptingnew
activities—whichcross-sectional methodsaccountfor,therebymakingourestimates
morerelevantforassessingtheimpactofclimatechange.
An unavoidable drawback of our estimation strategy is that it requires climate
amenities to be uncorrelated with the influence unobserved local amenities have on
QOL. This untestable assumption is hard to circumvent, as there do not appear to
be any viable instrumental variables for climate. Instead, we examine the potential
foromittedvariablebiasbytestingtherobustnessofourestimatestoanarrayofspec-
ificationsandpowerfulcontrols,followingtheintuitionlaidoutbyAltonji,Elder,and
Taber (2005) and the literature on agricultural yields and farmland values (Schlen-
ker, Hanemann, and Fisher 2006; Deschênes and Greenstone 2007; Schlenker and
Roberts2009;DeschênesandGreenstone2012;Fisheretal.2012).Whilethestabil-
ityofourestimatesacrossthesespecificationsisreassuring,wenonethelessacknowl-
edge that our research design cannot conclusively rule out the possibility that unob-
servedfactorsareinfluencingourestimates.
PriorhedonicstudiesinvestigatingUSclimatepreferenceshavebeenfewandcon-
flicting.EstimatesofWTPforincrementalwarmingrangefrompositive (Hochand
Drake1974;Moore1998),toapproximatelyzero(Nordhaus1996),tonegative(Cragg
and Kahn 1997, 1999; Kahn 2009; Sinha and Cropper 2013).1 Our paper makes
three contributions to the literature on climate amenities by drawing from recent in-
novations in the literatures on climate damages to agriculture and health, quality of
life measurement, and hedonic estimation under preference heterogeneity. First, we
characterize climates using the full distribution of realized daily temperatures rather
thanseasonalormonthlyaverages, allowing us toexplore how householdsvalue pro-
gressivelyextremetemperatures.Priorresearchintoclimateamenitieshasignoredthe
importance of flexibly modeling temperature profiles, even though other research has
1. HedonicstudiesfocusingoncountriesotherthantheUnitedStatesincludeMaddison
andBigano(2003)inItaly,Rehdanz(2006)inGreatBritain,RehdanzandMaddison(2009)
in Germany, and Timmins (2007) in Brazil. In addition, alternative nonhedonic approaches
have been used to estimate the impact of climate change on amenities. Shapiro and Smith
(1981) and Maddison (2003) use a household production function approach, Fritjers and
Van Praag (1998) use hypothetical equivalence scales, and Rehdanz and Maddison (2005)
linkapanelofself-reportedhappinessacross67countriestotemperatureandprecipitation.
208 JournaloftheAssociationofEnvironmentalandResourceEconomists March2016
shown that extreme temperatures, not averages, are especially harmful to crop yields
(Schlenker et al. 2006; Schlenker and Roberts 2009) and health (Deschênes and
Greenstone 2011; Barreca 2012). Second, following Albouy and Lue (2015), our
QOLestimatesaccountforcommutingcosts,localexpendituresbeyondhousing,and
federal taxes on wages. We show that previous work, by ignoring these factors, pro-
duced noisy and misleading estimates of climate valuations, while our results are ro-
busttomanyalternativespecifications,suchasincludingcontrolsforeverystate.Third,
weapplymethodsbyBajariandBenkard(2005)tomodelunobservedheterogeneityin
households’preferences,therebyallowingforsortingbasedondifferencesin(dis)taste
forcoldorheatandforadaptationtolocalclimates.
Our estimates of climate preferences yield four main results. First, we find that
Americansmostpreferdailyaveragetemperatures—theaverageofthedailyhighand
dailylow—near65degreesFahrenheit(°F),agreeingwithstandarddegree-daymod-
elsthatpredict littleneedforheating orcoolingatthistemperature. Second,onthe
margin,householdspaymoretoavoidadegreeofexcessheatthanadegreeofexcess
cold.Third,wefindthatthemarginalWTP(MWTP)toavoidextremecoldisnot
substantially greater than the MWTP to avoid moderate cold. Put another way,
householdswillpaymoretoaturnamoderately colddayintoaperfectdaythanto
turn a bitterly cold day into a moderately cold day. This finding is consistent with
evidencethathouseholdsreducetheirtimeoutdoorsastemperaturesbecomeuncom-
fortable, reducing their sensitivity to further temperature changes (Graff Zivin and
Neidell2012).Fourth,wefindthathouseholdsintheSouthareparticularlyaverseto
cold.Thisresultisconsistentwithmodelsofbothresidentialsortingandadaptation.
Conversely, we do not find evidence that southern households are less heat averse
thannorthernhouseholds.
Weapplyourestimatedclimatepreferencesbysimulatinghowclimatechangemay
affect welfare by improving or reducing the value of climate amenities. For our cli-
matechangepredictions,werelyprimarilyonthebusiness-as-usualA2scenarioused
intheIntergovernmentalPanelonClimateChangefourthassessmentreport(IPCC
2007),whichpredictsa7.3°FincreaseinUStemperaturesby2100.Oursimulated
welfare effects are predicated on technology and preferences remaining constant and
are therefore best interpreted as a benchmark case. This assumption is common to
most estimates of climate change damages, including the agricultural and health lit-
eratures referenced above. We view endogenizing technology and preferences as be-
yondthescopeofthispaper’sclimatechangeapplication,andweleavethisissuefor
futurework.
Weprojectthatonaverage,Americanswouldpay1%–4%oftheirannualincome
toavoidpredictedclimatechange.WhiledamagesarerathersevereintheSouth,we
findthatmostareasintheNorthalsosufferbecause(1)theylosemanypleasantsum-
merdaysinexchangeforonlymoderatelywarmerwinterdaysand(2)northernersare
willingtopaylesstoreducecoldthanaresoutherners. Welfare impacts arereduced
slightlywhenwemodelmigrationresponses.
ClimateChangeandAmericanQualityofLife Albouyetal. 209
1. HEDONIC ESTIMATES OF QUALITY OF LIFE
The intuition underlying our approach is that households pay higher prices and ac-
cept lower wages to live in areas with desirable climate amenities. Below we discuss
the hedonic framework underlying this intuition, howwecalculate wage and costof
living differences across areas, and how we combine these to create a single-index
QOL measure for each location. Our approach is rooted in the conceptual frame-
work of Rosen (1974, 1979) and Roback (1982) and adopts important recent con-
tributions to this framework from the hedonic literature. In particular, we follow
Albouy (2012) and Albouy and Lue (2015) to properly weight wages and housing
priceswhencreatingtheQOLmeasure,andweadoptBajariandBenkard(2005)to
allow for preference heterogeneity. These innovations ultimately prove to be conse-
quentialinobtainingpreferencesforclimate,aswediscussfurtherbelow.
1.1. AModel of QOLUsing Local Cost of Livingand Wage Differentials
The US economy consists of locations, indexed by j, which trade with each other
andshareapopulationofperfectlymobile,price-takinghouseholds,indexedbyi(we
discussestimates thatrelaxthe perfect mobilityassumptioninsec. 6).Thesehouse-
holdshavepreferencesovertwoconsumptiongoods:atradednumerairegoodxand
a nontraded “home” good y, with local price p that determines local cost of living.
j
Households earn wage income wi that is location-dependent and own portfolios of
j
land and capital that pay a combined income of Ri. Gross household income is
ð Þ
mi=Riþwi, out of which households pay federal taxes of τ mi . Federal revenues
j j j
arerebatedlumpsum.2
Each location is characterized by a K-dimensional vector of observable amenities
Zj, including climate, and a scalar characteristic ξj that is observable to households
but not econometricians.3 We assume a continuum of locations so that the set of
availablecharacteristics(Z,ξ)isacomplete,compactsubsetofℝKþ1.Eachhousehold
ð Þ
i seeks out the location j that maximizes its utility, given by ui=Vi p;wi;Z;ξ .
j j j j j
Thisindirectutilityfunctionisassumedtobecontinuousanddifferentiableinallits
arguments, strictly increasing in wi and ξ, and strictly decreasing in p. Households
j j
arepermittedtohaveheterogeneouspreferencesover(Z,ξ).
2. Wealsoapplyadjustmentsforstatetaxesandtaxbenefitstoowneroccupiedhousing,
discussedinAlbouy(2012),whichprovetobeminorinpractice.
3. This set up omits an idiosyncratic unobserved preference shock ε from households’
ij
utilityfunction.RelaxingthisassumptionimpliesthattheQOLmeasuremaydependnotjust
onpricesandwagesbutalsoonpopulationlevelsorchanges.Weallowforsuchdependence
insomeofourspecificationsasdiscussedbelow(forexample,byadjustingourwageestimates
formigrationperDahl[2002]),noneofwhichsubstantiallyaltersourconclusions.SeeAlbouy
(2012)foramoregeneraldiscussionofaddressingidiosyncraticpreferencesinQOLestimates.
210 JournaloftheAssociationofEnvironmentalandResourceEconomists March2016
Onthesupplyside,weassumethatfirmsfaceperfectlycompetitiveinputandout-
putprices and earn zeroprofits, offering higher wagelevels in more productive loca-
tions.Wemodeleachhouseholdi’swageinlocationjasϕiw,whereϕiishousehold-
j
specificskillandw isthelocalwagelevel.4
ð Þ j ð Þ
Let p Z;ξ and w Z;ξ denote the functions relating wages and prices to
j j j j
local characteristics. These functions are determined in equilibrium by households’
demandsforlocalamenities,firms’locationdecisions,andtheavailabilityofland.On
the demand side, household utility maximization implies the following first-order
conditionforeachcharacteristick:
1 ∂Viðpj;wj;Zj;ξjÞ=pjyij∂lnpðZj;ξjÞ–wijð1–τ0Þ∂lnwðZj;ξjÞ; ð1Þ
miλ(cid:2)i ∂Z mi ∂Z mi ∂Z
j k j k j k
where λi is the marginal utility of income and τ0 is the average marginal tax rate on
labor income.This equation relateshousehold i’s marginal valuation of characteristic
k, as a fractionofincome, todifferentialchangesinthelogarithmsofthe costofliv-
ingandwagedifferentialsatj.
Operationally,wedevelopaQOLindextoindicatethewillingnesstopayofhouse-
holds,averagedbyincome,fromtheright-handsideof(1).Thismeasureatj,denoted
Q^ , isa weighted combination of^p and w^,the differentials inlog housing costs and
j j j
wagesrelativetotheUSincome-weightedaverage,accordingtotheformula
Q^j=s^p –ð1–τ0Þs w^ =0:33^p –0:50w^; ð2Þ
y j w j j j
wheres denotestheaverageshareofincomespentonlocalgoodsands theaverage
y w
share of income from wages. The second equality substitutes in values for these pa-
rameters of s =0:33, s =0:75, and τ0=0:33. For additional details, including the
y w
incorporationoflocalnonhousingexpendituresintos,seeAlbouy(2012).Section6
y
considersestimatesallowingforheterogeneityins ands .
Let Q^ðZ;ξÞ denote QOL as a function of lyocal chwaracteristics, per (2) and the
ðj j Þ ð Þ
functionsp Z;ξ andw Z;ξ .Then,bycondition(1),foranyhouseholdiinj,the
j j j j
marginal willingness to pay (MWTP) for characteristic k is equal to the derivative of
Q^ðZ;ξÞwithrespecttok:
j j
1 ∂Viðpj;wj;Zj;ξjÞ=∂Q^ðZj;ξjÞ: ð3Þ
miλi ∂Z ∂Z
j k k
Condition (3) is illustrated in figure 1 in the case of a single characteristic, aver-
age summer temperature T. The bold line denotes a hypothetical function Q^ðTÞ
s s
4. Byusingasingleindexofskill,weabstractawayfromthepossibilitythatsomehouse-
holdshaveacomparativeadvantageincertainlocations.Relaxingthisassumptionhasimplica-
tionssimilartothoseforallowinganidiosyncraticunobservedpreferenceshock.
ClimateChangeandAmericanQualityofLife Albouyetal. 211
Figure 1. Illustrativehedonicpricefunctionwithdemand-sideequilibriumfirst-ordercon-
ditionssatisfied.“MWTP”denotesmarginalwillingnesstopay.
that is decreasing in T, indicating that milder weather is “paid for” through either
higherhousingpricesors lowerwages.TheslopeofQ^ðTÞatanygivenlocationisthe
s
hedonic price for a marginal increase in temperature at that location. This hedonic
priceisequaltohouseholds’MWTPforT atthatlocation,asshownforlocationsA
s
andBonthefigure.Asshown,householdsatAhaveahigherMWTPtoavoidheat
thandohouseholdsatB,consistentwithsortingoradaptation.
1.2. Estimates ofWage and Housing Cost Differentials, and QOL
We follow Albouy (2012) to estimate wage and housing-cost differentials using the
5% sample of Census data from the 2000 Integrated Public Use Microdata Series
(IPUMS). Geographically, these data are available by Public Use Microdata Areas
(PUMAs), which contain populations of 100,000–300,000 and form the main unit
of our analysis. We summarize this procedure below; for more details, see Albouy
(2012).
212 JournaloftheAssociationofEnvironmentalandResourceEconomists March2016
WecalculatewagedifferentialsbyPUMA,w^,usingthelogarithmofhourlywages
j
forfull-timeworkersaged25–55andcontrollingforobservableskillandoccupation
differencesacrossworkers.Specifically,weregressthelogwageofworkerionPUMA
indicators μw and extensive controls Xw (each interacted with gender) for education,
j i
experience,race,occupation,andindustry,aswellasveteran,marital,andimmigrant
status,inanequationoftheformlnw =Xwβwþμwþεw.Theestimatesoftheμware
ij i j ij j
usedasthePUMAwagedifferentialsw^ followingarefinement,perAlbouyandLue
j
(2015), so that they reflect wages by place of work rather than place of residence,
nettingoutdifferencesincommutingcosts.
Ourmodelinterpretsthew^ asthecausaleffectofaPUMA’scharacteristicsona
j
worker’s wage, while the observable and unobservable skill differences across work-
ers, the Xw and εw, are an analogue to the ϕi factors in the model. This interpreta-
i ij
tion requires that anysorting ofworkers across locations based on unobserved skills
or a spatial match component of wages does not substantially affect observed wage
premia. Thisassumptionreceives mixedsupportintheliterature. Glaeserand Maré
(2001) and Baum-Snow and Pavan (2012) find that unobserved skill and match-
based sorting contribute little to city-size wage premia; however, Gyourko, Mayer,
andSinai(2013)findthataselectgroupof“superstarcities”maydisproportionately
attract high-skilled workers. Dahl (2002) finds that selective migration biases esti-
matesofthereturnstoeducation(thoughnottherangeofreturnsacrossstates),and
KennanandWalker(2011)findsaroleforlocation-specificjobmatchesinmigration
decisions.
In light oftheuncertainty intheliterature, weaddress thepotential forskill and
match-basedsortinginaseriesofalternativespecifications.First,weadoptthemethod
usedinDahl(2002)toadjustourwageestimatesw^ forselectivemigrationbyinclud-
j
ingaflexiblecontrolfunctionofmigrationprobabilitiesinourwageequation.Second,
inacloselyrelatedspecification,wedirectlyadjustourQOLestimatesforeachPUMA
usingthePUMA’srateofnetin-migrationbetween1990and2000(inpercent)and
a mobility cost estimate from Notowidigdo (2013). Finally, we guard against effects
from “superstar cities” by estimating a specification in which superstar metropolitan
areas(asdefinedinGyourkoetal.[2013])aredroppedfromthesample.Thesespec-
ifications are described in more detail in section 6 and ultimately yield estimates of
climate preferences and welfare effects that are not qualitatively different from our
baselineestimates.
Tocalculatehousing costdifferentials, weusehousingvaluesandgrossrents,in-
cluding utilities. Following previous studies, we convert housing values to imputed
rentsatadiscountrateof7.85%(PeiserandSmith1985)andaddinutilitycoststo
make them comparable to gross rents. This approach follows the standard practice
intheQOLliteraturefromBlomquist,Berger,andHoehn(1988)toChenandRo-
senthal(2008)andisrequiredbythedatabecauseutilitycostsareincludedingross
ClimateChangeandAmericanQualityofLife Albouyetal. 213
rents.Wethencalculatehousing-costdifferentialswitharegressionofrentsonflex-
iblecontrolsY (eachinteractedwithrenterstatus)forsize,rooms,acreage,commer-
ij
cialuse,kitchenandplumbingfacilities,typeandageofbuilding,andthenumberof
residents per room. This regression takes the form lnp =Y βpþμpþεp.The esti-
ij ij j ij
mates of the μp are then used as PUMA-level housing cost differentials ^p. Proper
j j
identification of housing-cost differences requires that they not vary systematically
withunobservedhousingqualityacrosslocations.
We incorporate energy and insulation costs in our housing-cost measure because
doingsoallowsustointerpretourQOLdifferentialsassolelyreflectingthevalueof
nonmarketclimateamenitiesratherthantheeffectofclimateonutilitycosts.Hence,
ourQOLdifferentialswillreflectthedisamenityofoutdoorexposuretoclimateand
the disamenity of adverse indoor temperatures to the extent that they are not com-
pletely mitigated by insulation and energy use. In addition, the QOL estimates will
incorporate any disamenity from spending more time indoors to avoid uncomfort-
ableoutdoortemperatures.
Descriptive statistics for QOL are given at the bottom of table 1,5 and QOL
differentials across PUMAs for the year 2000 are mapped in figure 2. These esti-
matesshowthathouseholdsfindtheamenitiesinurbanareas,coastallocations,and
certainmountainareastobequitedesirable.Areasinthemiddleofthecountry,where
seasonsaremoreextreme,tendtobelessdesirable,althoughthevariationisconsider-
able.AsdiscussedinAlbouy(2012),ourQOLestimatescorrelatewellwithnoneco-
nomicmeasuresofQOL,suchasthe“livability”rankingsinthePlacesRatedAlmanac
(Savageau 1999). Moreover, the QOL model correctly predicts the relationship be-
tweenhousingcostsandwages,controllingforobservableamenities.
2. DATA
We estimate our main specifications at the PUMA level using 2,057 PUMAs cov-
eringthecontiguous48statesasofthe2000census.6Inthissection,wesummarize
ourdataset,coveringrecenthistoricalclimate,climate-changeprojections,andother
variables.Additionaldetailsareprovidedinappendix2(availableonline).
5. The meanQOL differential is notexactly zero in table 1 because thetable shows un-
weighteddata,whileQOLdifferentialsaredefinedsothattheincome-weightedmeaniszero.
6. We have also aggregated our data to the MSA-level and run some of our empirical
specifications at an MSA-level resolution. The point estimates for preferences and climate
changewelfareimpactsaresimilartothosediscussedbelow,withmodestlylargerstandarderrors
(see appendix table A1.1, col. R13). We believe that the MSA-level results are less precise
because MSAs are frequently too large to capture important micro-climates, particularly in
denselypopulatedcoastalareassuchasSanFrancisco.
Table1.DescriptiveStatisticsforPrimaryDataSet
10th 90th
Mean SD Percentile Percentile
Temperaturedata,1970–99average:
Averageannualheatingdegreedays
(1000s) 4.384 2.204 1.326 7.009
Averageannualcoolingdegreedays
(1000s) 1.290 .929 .374 2.762
Temperaturedata,2070–99projected
(CCSMA2):
Projectedaverageannualheating
degreedays(1000s) 2.974 1.729 .527 5.024
Projectedaverageannualcooling
degreedays(1000s) 2.547 1.190 1.304 4.413
Otherclimatedata,1970–99average:
Averageannualprecipitation
(inches) 39.26 14.09 16.25 53.85
Averageannualrelativehumidity(%) 63.58 8.10 53.31 70.52
Averageannualsunshine(%of
availabledaylight) 60.18 8.64 49.74 73.08
Otherclimatedata,2070–99
projected(CCSMA2):
Projectedaverageannual
precipitation(inches) 41.54 15.23 15.28 56.40
Projectedaverageannualrelative
humidity(%) 62.66 8.74 51.18 69.63
Projectedaverageannualsunshine
(%ofavailabledaylight) 61.37 8.56 51.31 72.85
Geographicdata:
Areainsquaremiles 1,436 4,255 17 3,355
DistancefromcentroidofPUMA
toocean(miles) 250.1 272.3 4.3 729.2
DistancefromcentroidofPUMA
toGreatLake(miles) 763.2 715.4 54.0 2,128.4
Averagelandslope(degrees) 1.677 2.131 .191 4.270
Demographicdata(2000census):
Weightedpopulationdensity
(peoplepersquaremile) 5,466 11,997 360 9,981
Percenthighschoolgraduates 83.9 8.8 72.1 93.3
Percentofpopulationwith
bachelor’sdegree 24.1 12.4 11.3 41.0
Percentofpopulationwith
graduatedegree 8.7 5.6 3.7 20.3