Improving energy poverty measurement in Southern European Regions through equivalisation of theoretical energy costs

sus ainabili y
A icle
Imp o ing Ene gy Po e y Measu emen in Sou he n
Eu opean Regions h ough Equi aliza ion o
Modeled Ene gy Cos s
Iñigo An epa a 1,* , Le ko hea Papada 2, João Ped o Gou eia 3, Nikolas Ka soulakos 2and
Dimi is Kaliampakos 4
1Alokabide, Technical Depa men , Po al de Gama a, 1A—2a plan a (Edi icio el Boule a d),
01013 Vi o ia-Gas eiz, Spain
2Me so ion In e disciplina y Resea ch Cen e , Na ional Technical Uni e si y o A hens,
9 He oon Poly echneiou S ., Zog aphos, 15780 A hens, G eece; [email p o ec ed] (L.P.);
[email p o ec ed] (N.K.)
3CENSE Cen e o En i onmen al and Sus ainabili y Resea ch, NOVA School o Science and Technology,
NOVA Uni e si y Lisbon, 2829-516 Capa ica, Po ugal; [email p o ec ed]
4School o Mining and Me allu gical Enginee ing, Na ional Technical Uni e si y o A hens,
9 He oon Poly echneiou S ., Zog aphos, 15780 A hens, G eece; [email p o ec ed]
*Co espondence: [email p o ec ed]; Tel.: +34-945-000-565
Recei ed: 21 June 2020; Accep ed: 14 July 2020; Published: 16 July 2020


Abs ac :
In many Eu opean coun ies, ene gy po e y is measu ed on he basis o eal ene gy bills,
as heo e ical ene gy cos s a e ha d o calcula e. The UK is an excep ion— he da a inpu s o he Low
Income-High Cos (LIHC) indica o a e based on easonable ene gy cos s, hese da a a e collec ed
h ough specially designed su eys, o en an in ensi e and cos ly p ocedu e. App oaches which
calcula e ene gy needs a e alid when ene gy bill da a a e un eliable o whe e households es ic
consump ion. In his analysis, ene gy po e y le els a e e alua ed o G eece, he municipali y o
É
o a (Po ugal), and he Basque Coun y (Spain): ene gy bills a e modeled based on building ene gy
pe o mance da a and o he ene gy uses, and adjus ed acco ding o socio-demog aphic a iables.
To his end, equi aliza ion weigh s a e calcula ed using socio-economic da a om he a o emen ioned
sou he n Eu opean coun ies/ egions. Da a a e analyzed o compa e measu emen s wi h ac ual e sus
modeled bills using he Ten-Pe cen Rule (TPR) and Hidden Ene gy Po e y (HEP) agains wice he
median (2M) indica o , enhancing he iden i ica ion o households wi h low ene gy consump ion.
In conclusion, heo e ical ene gy needs can be combined wi h socio-demog aphic da a ins ead o
ac ual ene gy bills o measu e ene gy po e y in a simpli ied way, a oiding he p oblem o a ge ing
households ha unde consume.
Keywo ds:
ene gy po e y; modeled ene gy cos s; TPR; HEP; equi aliza ion coe icien s; unde
consump ion; G eece; Po ugal; Spain
1. In oduc ion
Be ween 54 and 141 million people a e unable o keep hei homes adequa ely wa m in Eu ope;
hese ci izens a e he e o e conside ed o be in ene gy po e y [
1
]. Day e al. [
2
] de ine ene gy po e y
as “insu icien access o a o dable, eliable and sa e ene gy se ices”.
The Eu opean Commission (EC) is awa e o he si ua ion wi hin many Eu opean households.
In 2016, The EC published he Win e Package, which aimed o shape he clean ene gy ansi ion,
whils a oiding he ma ginaliza ion o ulne able consume s. The second o eigh legisla i e ac s
app o ed in 2019 was he Eu opean Pa liamen legisla i e esolu ion o 13 No embe 2018 on he
Sus ainabili y 2020,12, 5721; doi:10.3390/su12145721 www.mdpi.com/jou nal/sus ainabili y
Sus ainabili y 2020,12, 5721 2 o 21
p oposal o a di ec i e o he Eu opean Pa liamen and o he Council amending Di ec i e 2012/27/EU
on ene gy e iciency, which s a es ha “ he Union’s ene gy e iciency policies should be inclusi e
and should he e o e ensu e accessibili y o ene gy e iciency measu es o consume s a ec ed by
ene gy po e y”.
The EU Ene gy Po e y Obse a o y (EPOV) [
3
] ini ia i e was launched by he EC o assis
Membe S a es (MS) in hei e o s o mi iga e ene gy po e y. One o i s asks is o de ine use ul
indica o s o ack he ene gy po e y si ua ion ac oss Eu ope. The common indica o s p oposed by
he EPOV a e [3]:
•
A ea s on u ili y bills, sha e o (sub-) popula ion ha ing a ea s on u ili y bills, based on
sel - epo ed expe iences o limi ed access o ene gy se ices (i.e., EU-S a is ics on Income and
Li ing Condi ions (SILC) da a);
•
Low absolu e ene gy expendi u e (M/2), sha e o households whose absolu e ene gy expendi u e
is below hal he na ional median, based on household income and/o ene gy expendi u e da a
(i.e., Housing Budge Su ey (HBS) da a);
•
High sha e o ene gy expendi u e in income (2M), p opo ion o households whose sha e o
ene gy expendi u e in income is mo e han wice he na ional median sha e, based on HBS da a;
•
Inabili y o keep home adequa ely wa m, sha e o (sub)popula ion no able o keep hei home
adequa ely wa m, based on EU-SILC da a.
As acknowledged by p e ious esea ch [
4
–
6
], measu ing and moni o ing ene gy po e y is
challenging. The mul idimensionali y o he condi ion means ha i is no cap u ed h ough a single
indica o , and he e is also a lack o ele an and de ailed key da a ha could be used o c oss-coun y
compa isons and o de ailed egional analysis.
The i s a emp o measu e ene gy po e y was applied in he UK, using he Boa dman [
7
]
me hodology. This me hod, known as he Ten-Pe cen Rule (TPR), calcula es he a io be ween income
and ene gy cos s, whe e ene gy cos s should no exceed 10% o household income. Ene gy cos s a e
modeled a he han based on ac ual bills, and include space hea ing, domes ic ho wa e (DHW),
cooking, ligh ing, and addi ional elec ic appliances [
8
]. C i ics o he me hod ques ioned he means o
calcula ing ene gy cos s [
6
] and iden i ied an unde ep esen a ion o he impac o o he a iables,
such as he imp o emen o ene gy e iciency [
9
]. Addi ionally, Boa dman’s 10% app oach was
designed o he UK, so i s ele ance may a y be ween di e en coun ies and income le els [6].
In 2012, he UK e ised he de ini ion and adop ed he me hodology de eloped by Hills [
9
],
known as he Low Income-High Cos (LIHC), based on easonable ene gy cos s. Hills conside ed
wo addi ional c i e ia o measu e ene gy po e y [
10
]: o iden i y low-income households li ing in
ine icien dwellings as uel-poo , and a sel - epo ed, subjec i e index. Hills [
9
] inally ecommended
he ene gy po e y indica o was based on easonable ene gy cos s, and in doing so, he LIHC is
ega ded as a benchma k o ‘objec i e’ expendi u e- and income-based indica o s [
11
]. The main
poin is ha ene gy po e y cen e s on a lack o su icien wa m h and space hea ing [
12
], a guably
he mos impo an domes ic ene gy se ice, as a consequence o he well-known implica ions o
li ing in a cold home o physical and men al heal h. The basis o he Hills me hodology [
9
] is ha
well-insula ed dwellings equi e less ene gy o main ain a p esc ibed indoo he mal com o and his
can be accoun ed o in building ene gy modeling echniques [10].
Al hough he Boa dman me hod also uses modeled ene gy cos s ins ead o ac ual bills, in he wo ds
o Hills, his is he majo ad an age o his ype o me hodology [
10
]. The main di e ence be ween bo h
me hods is ha Hills p oposed o ix he ene gy po e y h eshold a he median ene gy expendi u e.
The BRE Domes ic Ene gy Model (BREDEM) modeling ool, using da a om he English Housing
Su ey (EHS), is applied o calcula e he h eshold o he median modeled bill [
13
], a oiding ene gy
cos s ela ed o unhea ed spaces, e.g., due o excessi e size o he building, using a ious hea ing
egimes dependen on he age o he household and employmen s a us, and adjus ing o he numbe
o occupan s. A s anda d hea ing pa e n, i.e., assuming ha du ing usual wo king hou s he house
Sus ainabili y 2020,12, 5721 3 o 21
is emp y, does no apply, in pa icula o he ulne able, e.g., he elde ly o hose ca ing o young
child en [
13
]. Fo hese cases, he ull hea ing egime was in oduced in he UK me hodology. Size and
numbe o occupan s is also impo an ; he e o e, o unde -occupied dwellings, pa ial s anda d/ ull
hea ing egimes we e included. Ini ially in he LIHC de ini ion, he addi ion o non-hea ing ene gy
uses was no clea [
14
]; ollowing he unpacking o he me hodology used o gene a e he UK’s annual
uel po e y s a is ics hese measu es became mo e p ominen . The s eps in ol ed in his me hodology
can be summa ized as ollows: income is measu ed wi hou conside ing housing cos s, he so-called
a e housing cos s [
15
], he dynamic 2M is used o he a o dabili y h eshold. Then, he ‘ uel po e y
gap’ is added as an addi ional indica o in o de o conside he di e ence, in mone a y e ms, be ween
he modeled expendi u e o an ene gy poo household and he expendi u e equi ed o emo e ha
household om ene gy po e y. Subsequen ly, once modeled ene gy (and housing) cos s a e deduc ed,
hose whose income is no wi hin he 60 pe cen median ange a e iden i ied as ene gy poo [10].
Hills [
9
] included bo h ene gy bills and income de ails, and iden i ied ha hey should be adjus ed
so ha compa ison o a single h eshold is possible o households wi h di e en numbe s o occupan s,
known as ‘equi aliza ion’ o income measu emen . Equi aliza ion allows compa ison on he same
scale be ween households o di e en composi ion and size: dec easing he expenses o mul iple
pe son households and inc easing he expenses o single pe son households in o de o make hem
compa able. Income equi aliza ion o he O ganiza ion o Economic Co-ope a ion and De elopmen
(OECD) coun ies was depic ed by [
16
]: as 1 head, 0.5 o he adul , 0.3 child en unde 14. In he UK
me hodology, income is equi alized a e he discoun ing o housing cos s [13].
To secu e a good le el o capabili ies, he necessa y amoun o ene gy se ices is dependen on
household size and he clima e, as p e iously desc ibed, bu also speci ic indi iduals’ needs and
ci cums ances—e.g., whe he hey a e olde , disabled, e y young, o ill [
2
]. To accoun o his e ec ,
no wi hs anding hea ing egimes, Hills included he equi aliza ion o ene gy cos s. Ini ially, using
he same ac o s used o income (adap ed om he modi ied OECD scale), bu as he ela ionship o
household o size ene gy cos s and gene al li ing cos s a e no he same, he idea was la e ejec ed [
9
].
Fo ins ance, i he numbe o occupan s inc eases by one pe son in an iden ical space, hea ing use
would a y e y li le, while he cos o li ing would inc ease [
9
]. Fu he mo e, he ene gy demands o
speci ic consume s can exceed he a e age as a esul o socio-demog aphic ac o s, which can lead
o hei being iden i ied as ulne able, i.e., i hey expe ience di icul ies when accessing ene gy [
17
].
I hose ulne able cha ac e is ics a e o be included, addi ional adjus men s a e needed, esul ing
in he equi alized heo e ical ene gy cos s. The equi aliza ion ac o s o ene gy cos s a e e iewed
pe iodically, based on h ee yea s o equi ed uel cos da a om he EHS, bu no on an annual
basis [13].
On he di e ences be ween indica o s, TPR is an absolu e indica o , while LIHC is ela i e.
Bo h TPR and LIHC ha e hei d awbacks:
•TPR is a ixed a io and e y sensi i e o ene gy p ices [9];
•
LIHC—smalle homes a e excluded om he de ini ion, whe e inc eases in ene gy p ices no
longe ha e an impac on he indica o [
18
]. The indica o also has a doubly- ela i e cha ac e ,
which makes i e y di icul o isola e causes and e ec s in he p ocess o analysis [19].
The endency o he Hills LIHC indica o o iden i y low income households li ing in smalle
dwellings was desc ibed by Moo e e al. [
20
]. The e is a disp opo iona e ep esen a ion o households
expe iencing di icul y paying ene gy bills, and consequen ly unde -spending on ene gy, esul ing in
lowe indoo empe a u es. Accoun ing o dwelling size is he e o e c i ical, and a sligh ly di e en
equi aliza ion me hod was subsequen ly p oposed, his me hod is dependen on occupancy and he
dwelling size-useable loo a ea (m2) [20].
Dwelling size is a key ac o con ibu ing o high ene gy bills, he e o e, in an a emp no o
unde es ima e his e ec , he adjus men o ene gy cos s o dwelling size was deemed inapp op ia e
by he UK Go e nmen [
13
]. This is cohe en wi h o he app oaches, e.g., L’Obse a oi e Na ional de
Sus ainabili y 2020,12, 5721 4 o 21
la P
é
ca i
é É
ne g
é
ique (ONPE) s a is ics [
21
], whe e he e ec o occupancy on ene gy cos s is less
han he impac o size o li ing space.
Hills a gues ha , as ene gy p ices change, he TPR esul s in housands o households mo ing
in and ou o ene gy po e y [
15
]. In compa ison, he LIHC gi es s able esul s o e ime, and also
ac oss coun ies. This does no seem o be ealis ic wi h ega d o he e ec o ene gy p ices and o he
a iables: ene gy p ices a e compa able h oughou he EU, ene gy e iciency is only d ama ically
supe io in No dic coun ies, wi h clima e being he only ac o which makes a signi ican di e ence
ega ding ene gy needs. To ace hei mos se e e wea he , Finland, No way, and Sweden impose
highe ene gy e iciency s anda ds [
22
], consequen ly hea ing cos s a e usually lowe , and gi en ha
hese cos s a e he mos signi ican con ibu o o ene gy expenses, ene gy po e y indexes should be
lowe , as a e age income is highe . In Finland and Sweden, he “keep wa m” subjec i e index is lowe
han 3% [
23
], whe eas 2M e ains he 15% o ene gy poo calcula ed o o he Eu opean coun ies.
In sou he n Eu opean coun ies, hea ing cos s a e high despi e compa a i ely good wea he condi ions,
and he “keep wa m” indica o is g ea e han 20% in G eece and Po ugal due o lowe mean income,
as shown in Table 1 o 2M. Unexpec edly, 2M emains a ound 15%.
The LIHC has been conside ed in o he EU Membe S a es as a benchma k indica o o o icial
ene gy po e y s a is ics [
11
]. Rademaeke s e al. [
5
], howe e , s a e ha ins ead o easonable/modeled/
heo e ical ene gy cos s, eal/ac ual ene gy cos s a e usually used as he p oxy o calcula e hea ing
expenses om o icial s a is ics. The da a needed o he Hills me hod/LIHC indica o a e p o ided by
specially designed UK su eys; his esou ce-in ensi e and cos ly p ocedu e can in ol e da a which a e
di icul o ob ain, making he whole p ocess expensi e. In Po ugal, Spain, and G eece, o example,
he cu en me hodologies using ene gy po e y measu emen s a e he ollowing:
•
G eece: no o icial me hod o iden i y ene gy po e y. The “inabili y o keep home adequa ely
wa m” index is moni o ed annually by he Hellenic S a is ical Au ho i y;
•Po ugal: no o icial me hod o iden i y ene gy-poo households;
•
Spain: he only a ailable s a is ics a a na ional le el a e no p o ided by an o icial ins i u ion,
bu hanks o he Associa ion o En i onmen al Sciences (Asociaci
ó
n de Ciencias Ambien ales,
ACA) [24], al hough using ac ual bills.
The EPOV pe o ms ene gy po e y measu emen s in hose h ee coun ies, and 23 mo e [
25
],
based on HBS da a. So, using ac ual ene gy bills.
App oaches using heo e ical ene gy cos s can be ound in o he coun ies. In F ance, whe e he
p oduc ion o s a is ics by ONPE is a esou ce in ensi e p ocess, he indica o s do no eply on modeled
in oices. Moo e [
18
] ound he use o ac ual ene gy bills con o e sial, conside ing his a iable as
a poo indica o o ene gy po e y. A e analyzing da a on ac ual ene gy use, mos households
do no li e a he empe a u es assumed in modeling [
15
]. In pa icula , he lowes income decile
appea o be li ing a empe a u es below he “adequa e” h eshold [
15
]. Moo e [
18
] iden i ied ha
low-income householde s s ay in cold dwellings because hey spend signi ican ly less on uel han
equi ed. Di icul y iden i ying households which unde -consume ene ge ically is widesp ead among
mos Membe S a es; acco ding o Flo io and Teissie [
26
], i is di icul o iden i y he p esence o
ene gy es ic ions, indica ing lowe eal expendi u e. Palma e al. [
27
] also add essed his issue o
unde consump ion h ough a de ailed analysis o egional ene gy pe o mance gaps o he ci y
o É o a.
The use o modeled ene gy cos s, howe e , led esea che s o include no ma i e c i e ia in ene gy
po e y cha ac e iza ion, esul ing in he idea o ac o ing o “access o adequa e ene gy se ices”,
gi en ha adequacy is a no ma i e concep [
6
]. The jus i ica ion o his app oach is ha modeled
ene gy bills a e ela ed o indoo empe a u es, and hese empe a u es a e ela ed o heal h p oblems.
Howe e , while i is di icul o know wha indoo household empe a u e condi ions a e, calcula ing
ene gy cos s wi h building e iciency da a is he bes way o a oid his a ge ing p oblem. I is
Sus ainabili y 2020,12, 5721 5 o 21
he e o e assumed ha his p oblem does no exis in he UK, and consequen ly, he LIHC me hodology
p io i izes ene gy e iciency measu es as he app op ia e solu ion [28].
As an al e na i e o sol e his p oblem o a ge ing, he Hidden Ene gy Po e y (HEP) index
is used in Belgium [
29
]: households wi h an ene gy expendi u e lowe han 50% o he na ional
median a e also conside ed o be in ene gy po e y. In 2013, he index iden i ied a no able 4.6% o
he popula ion as ene gy-poo . The applica ion o his index highligh s a new a ge ing p oblem,
which can be clea ly iden i ied in he ollowing wo cases:
•
Highly e icien buildings cause hei inhabi an s (be hey low- o high-income households)
o be conside ed as ene gy-poo by he HEP due o he low ene gy cos s esul ing om a
well-insula ed building;
•When compa ing a eas wi h di e en ene gy demands due o clima e, hose who eside in a eas
wi h a less se e e clima e may be conside ed as being in ene gy po e y.
Thus, al hough he HEP add esses he p oblem o a ge ing, i causes a p oblem o i s own.
A d i e in he sea ch o al e na i es is he a ious d awbacks associa ed wi h he Hills LIHC indica o ;
i.e., Moo e [
20
] c i icized i s complexi y and p oposed se e al imp o emen s, e.g., no excluding low
ene gy cos s. No wi hs anding he me hodology in UK, he e a e ew s udies dealing wi h he use
o heo e ical ene gy cos s. Wi h he aim o add essing his gap, Imbe e al. [
21
] used he F ench
3CL-me hod o modeling ene gy cos s o in es iga e whe he i is possible, in coun ies whe e da a
a ailabili y is limi ed, o ans e ene gy po e y indica o s based on equi ed ene gy needs.
In G eece, as in o he Eu opean coun ies, he use o ac ual ene gy expenses unde es ima es he
measu e o ene gy po e y. Mo eo e , sel -decla ed da a a e no eliable, as hey o en also include
uel consumed o anspo a ion. To sol e hese p oblems, Papada and Kaliampakos [
30
,
31
] used
modeled ene gy consump ion as a a iable o assessing ene gy po e y. The e we e 12 pa ame e s
ha we e aken in o accoun in he S ochas ic Model o Ene gy Po e y (SMEP) model used, e.g.,
hea ing-deg ee-days, coe icien o pe o mance o hea ing sys ems, hea ing p ice, hea ans e
coe icien . In Po ugal, Gou eia e al. [
4
,
32
] showcased o all Po uguese ci il pa ishes how a
mul idimensional ene gy po e y ulne abili y index could be used, while combining in o ma ion
om buildings’ ene gy pe o mance (i.e., ene gy needs), inal ene gy consump ion, clima e a iables,
and ep esen a i e socio-economic indica o s.
In F ance, based on he su ey o Housing Pe o mance, Equipmen , Needs, and Uses o
Ene gy PHEBUS (L’enqu
ê
e Pe o mance de l’Habi a ,
É
quipemen s, Besoins e USages de l’
é
ne gie),
a simpli ied app oach by income and he F ench ene gy pe o mance label DPE (Diagnos ic Pe o mance
Ene g
é
ique) was explained by Flo io and Teissie [
26
] o ene gy po e y e alua ion. This app oach
employs he ela i ely simple me hod o se ing a minimum ene gy pe o mance s anda d o he
buildings (and wo king wi h he i s i e deciles o income, o example), simila o he Hills al e na i e
p oposal explained abo e. Theo e ically, i he ene gy pe o mance o all homes was a maximum
e iciency, and i ene gy p ices we e he same o all households, ene gy po e y and income po e y
assessmen would achie e he same esul s [
10
]. DPE, howe e , is only compulso y when en ing
o selling a p ope y, wi h DPE being calcula ed o only 3.6 million buildings in F ance in 2018,
and no o he emaining building s ock. Non-hea ing expenses a e no included in his me hod,
excep pe haps DHW, despi e he e being a s ong case o e alua ing he signi icance o non-hea ing
ene gy-uses [14].
This p oposal has he po en ial o con ibu e o he de elopmen o an al e na i e app oach o
he cu en ene gy po e y measu es wi h ac ual ene gy expenses; i ene gy e iciency s anda ds o
buildings could be made uni o m ac oss coun ies, his me hodology could be adap ed o he na ional
indica o s o each coun y. Theo e ical ene gy needs a e he ocus o his pape , whe e modeled ene gy
bills will be calcula ed based on building ene gy pe o mance da a and non-hea ing ene gy-uses,
and subsequen ly adjus ed acco ding o socio-demog aphic a iables, by using equi aliza ion weigh s
using he Hills [
9
] me hodology. Da a om h ee sou he n EU egions, i.e., G eece,
É
o a municipali y,

Sus ainabili y 2020,12, 5721 6 o 21
and he Basque coun y, will be used as case s udies, wi h econome ic analysis applied o calcula e
equi aliza ion coe icien s. The p oposed me hodology o calcula ing ene gy cos s and calcula ions
will be applied o he TPR indica o , simpli ying he p ocess o a ce ain ex en , bu no o he deg ee o
he F ench DPE app oach. The same app oach applied o he HEP indica o acili a es he iden i ica ion
o households ha unde -consume. In o de o a oid accoun ing o highe incomes, a cap on incomes
ha all wi hin he i s i e deciles can be se . This app oach will a oid inaccu acies in he iden i ica ion
o ene gy-poo households.
In he emaining sec ions, d awing on da a om di e en Eu opean egions, his pape will
ocus on measu ing ene gy po e y ollowing he me hodology abo e. The emainde o he a icle
is s uc u ed as ollows: in he ollowing sec ion he me hodology will be ou lined and da a se s
om G eece, and he Po uguese and Spanish egions will be p esen ed. The me hodology o he
econome ic analysis will hen be desc ibed, including an explana ion o why he p esen wo k
is ca ied ou using he 2M, TPR, and HEP indica o s. The esul s o he analysis a e in oduced,
and subsequen ly he key indings and discussion a e p esen ed, highligh ing how he TPR and HEP
indica o s can imp o e he iden i ica ion o ulne able homes h ough he use o heo e ical ene gy
cos s. The a icle concludes wi h some implica ions o policymake s.
2. Me hodology
In his sec ion, de ails o he coun ies selec ed o he case s udy a e p o ided, he me hodology
used o measu ing ene gy po e y is explained, and da ase s ha will be es ed a e cha ac e ized.
2.1. Case S udies Desc ip ion
Th ee case s udies we e used om sou he n Eu ope (i.e., G eece, he ci y o
É
o a in Po ugal, and
he Basque egion in Spain) o showcase he me hodology, p esen ing in e es ing a ia ions in clima e
cha ac e is ics, income le els, and ene gy p ices. Case s udy selec ion was based on he p esence o
unde -consuming and a ailabili y, in some cases e en hose wi h no hea ing we e included. The da a
om G eece o igina ed om a su ey pe o med by Papada and Kaliampakos [
33
] co e ing he whole
coun y. Fo Po ugal, a doo - o-doo su ey in he municipali y o
É
o a was used. In o ma ion on he
su ey conduc ed, ype o da a collec ed, and key insigh s a e a ailable in Gou eia e al. [
34
]. Spanish
da a o igina ed om social housing in he Basque egion, accessible h ough he public company’s
da abases. Used he e o he i s ime, i was collec ed in 2019, comp ising da a om 2018.
A summa y o he mos ele an a iables can be ound in Table 1. Popula ion da a we e sou ced
om Eu os a [
35
] o G eece, Po ugal, Spain, and he Basque coun y, and PORDATA o
É
o a
municipali y [
36
]. The numbe o households was aken om Eu os a [
35
] (l s _hhnh ych) o G eece,
Po ugal, and Spain, Eus a [
37
] o he Basque coun y, and PORDATA o
É
o a municipali y [
36
].
Clima e a iables we e hea ing deg ee days (HDD) and cooling deg ee days (CDD) calcula ed a 15.5
◦
C.
Ene gy p ices, including axes o households (band o elec ici y be ween 2500 and 5000 kWh/y . and
o gas be ween 20 and 200 GJ/y .) we e aken om he Eu os a [
35
]; a e age gas (n g_pc_202) and
elec ici y p ices (n g_pc_204) o he yea s 2016–2018. The income a iable o he coun ies ep esen ed
he adjus ed g oss disposable income o households pe capi a aken o Eu os a , while da a o he ci y
o
É
o a we e aken om Po uguese Na ional S a is ics O ice (INE) [
38
], and om Spanish INE o
he Basque coun y [
39
]. The mos ele an ene gy po e y indica o s o G eece, Po ugal, and Spain
we e aken om he EPOV da abase o la es da a a ailable [
25
], and e e ed o a pe cen age o he
popula ion. The e we e no o icial da a a ailable o ene gy po e y le els in he ci y o
É
o a, bu ACA
included disagg ega ed esul s o Spanish egions when calcula ing ene gy po e y indica o s o
Spain wi h da a o 2016 [24].
Sus ainabili y 2020,12, 5721 7 o 21
Table 1.
Summa y o mos ele an a iables in G eece,
É
o a municipali y (Po ugal), and Basque
egion (Spain).
G eece Po ugal Ci y o É o a Spain Basque Region
Popula ion 2018 110,741,165 10,291,027 52,664 246,658,447 2,170,868
Numbe o households 2018 14,348,100 3,910,800 29,812 217,384,300 1,054,610 3
HDD (inland) 2108 41390 (Ioannina) - 1050 1500 (Mad id) 1980 (Vi o ia)
HDD (coas ) 2018 4650 (A hens) 675 (Lisbon) - 970 (Ba celona) 955 (Bilbao)
CDD (inland) 2018 41060 (Ioannina) - 1185 1355 (Mad id) 535 (Vi o ia)
CDD (coas ) 2018 41675 (A hens) 970 (Lisbon) - 1340
(Ba celona) 875 (Bilbao)
Elec ici y p ice, €pe kWh,
2016–2018 10.168 0.228 0.228 50.23 0.23 5
Gas p ice, €pe kWh,
2016–2018 10.059 0.081 0.081 50.077 0.077 5
Adjus ed g oss disposable
income o households pe
capi a, 2016, €1
14,622 17,686 16,644 719,216 25,121 6
High sha e o ene gy
expendi u e in income (2M)
Popula ion (%), 2015 8
16.3 15.1 n/a 14.2 9.0 9
Low absolu e ene gy
expendi u e (M/2), Popula ion
(%), 2015 8
12.8 6.8 n/a 13.0 5.0 9
Inabili y o keep home
adequa ely wa m, Popula ion
(%), 2016 8
29.1 22.5 n/a 10.1 6.0 9
A ea s on u ili y bills,
Popula ion (%), 2016 842.2 7.3 n/a 7.8 6.0 9
1
Eu os a 2018 [
35
],
2
PORDATA [
36
],
3
Eus a [
37
],
4
h ps://www.deg eedays.ne /calcula ed a 15.5
◦
C,
5
Ma ke a
na ional le el, 6www.ine.es [39], 7www.ine.p [38], 8EPOV [25], 9ACA [24].
Wi h ega d o he di e ences ac oss coun ies, HDDs a e highe in Spain, bo h inland and coas al
a eas. When compa ing coun ies and egions, he Basque inland is colde han Mad id, and he ci y
o
É
o a is conside ed a e age, as he Po uguese inland egions do no a y signi ican ly in e ms
o al i ude. The CDDs a e g ea es o he coas al a eas o G eece and Spain, as he Spanish inland
egion is as ho as he coas al a ea du ing summe and he G eek inland a ea much colde . The Basque
coun y is much colde han he es o Spain du ing summe . O e all, wi h ega d o clima e da a,
he ci y o
É
o a is ep esen a i e o he Po uguese inland egions, while he Basque egion is one o
he coldes in Spain.
A e age income is highes in Spain and lowes in G eece; he same applies o household ene gy
p ices. The disposable income in he ci y o
É
o a is only sligh ly below he na ional a e age.
Compa a i ely, howe e , income in he Basque egion is he highes .
The ene gy po e y indica o s ake simila alues when using ela i e me hodologies, i.e., o 2M
a ound 15%, bu he sel -pe cei ed “inabili y o keep home adequa ely wa m” is di e en ac oss
coun ies, ollowing he in e se o de o mean income. The “a ea s on u ili y bills” indica o is
disp opo iona ely high in G eece. In he absence o o icial da a o
É
o a, he only analysis on ene gy
po e y analysis o Po uguese egions was conduc ed by Gou eia e al. [
4
]. In his s udy, he Ene gy
Po e y Vulne abili y Index (EPVI) was used o map and ank ulne abili y o ene gy po e y in
all Po uguese egions, bo h in he win e and in he summe . Thei la es esul s e ealed ha he
municipali y o É o a is anked 271s ( o win e ulne abili y) and 287 h ( o summe ulne abili y)
o a o al o 308 municipali ies. These esul s show ha
É
o a is on he lowe end o he ulne abili y
spec um in he hea ing and cooling seasons, his can be explained by a highe adap i e capaci y and
lowe ene gy pe o mance gaps in compa ison wi h o he Po uguese egions. Finally, o he Basque
coun y, e y p obably as a esul o highe disposable household incomes, all he ene gy po e y
indica o s a e lowe .
Sus ainabili y 2020,12, 5721 8 o 21
To summa ize, while he ci y o
É
o a is ep esen a i e o Po ugal ega ding clima e and income,
he ene gy po e y le els a e lowe . The esul s o he Basque Coun y a e no ep esen a i e o he
es o Spain, due o he colde clima e, highe incomes, and lowe ene gy po e y le els.
2.2. Resea ch Me hod
The esea ch hypo hesis es ed in he ollowing sec ion con es s ha modeled ene gy bills can
be adjus ed o ac ual ene gy expenses acco ding o socio-demog aphic a iables. Ce ain socio-
demog aphic a iables inc ease ene gy needs, o example, householde s main aining highe indoo
empe a u es and/o spending longe pe iods o ime a home, he make-up o such households usually
includes: (a) child en unde 18 yea s old, and (b) elde ly and/o pensione s [
13
]. These households
also e y o en include unemployed people and/o a dependen householde .
Fo his eason, analysis o ind co ela ions be ween ac ual ene gy bills and socio-demog aphic
a iables was i s pe o med using da a om G eece, and wo egions in Po ugal and Spain.
The co ela ions be ween each o he a iables and ac ual ene gy bills we e accep ed o ejec ed using
Spea man and Pea son analyses.
Subsequen ly, equi aliza ion weigh s we e calcula ed h ough an econome ic analysis, so as o
adjus socio-demog aphic a iables wi h he ac ual ene gy expendi u e and heo e ical ene gy cos s
based on he a ailable da a, explained in he nex sec ion. A posi i e co ela ion be ween a a iable and
he ac ual ene gy consump ion, e.g., a highe occupancy usually means highe ene gy consump ions,
should inc ease he modeled ene gy bills. The e o e, h ough he econome ic analysis, equi aliza ion
weigh s o socio- demog aphic cha ac e is ics we e calcula ed.
The p oposed econome ic model o he eg ession needed o calcula e hose equi aliza ion
ac o s is desc ibed in Equa ion (1):
Log(Ac ual ene gy bills) =β1×Log(Modeled ene gy bills) +Σiβixi+εi, (1)
whe e “Ac ual ene gy bills” is he dependen a iable ac ual ene gy expenses, “Modeled ene gy bills”
ep esen s he ene gy cos s calcula ed heo e ically, x
i
a e he explana o y a iables (occupancy, a ea.
and he socio-demog aphic a iables as bina y a iables, namely, he p esence o child en and/o
elde ly, also including pensione s/ e i ed and unemployed whe e e possible), and
εi
is he e o e m.
In o de o adjus he model,
βi
coe icien s a e a ained om he eg ession. This app oach is he
model commonly p oposed, e.g., by [40].
Hea ing/cooling ene gy needs will be explained in he nex sec ion, and hey a e in oduced as
kWh pe dwelling. Sunikka-Blank and Gal in [
41
] p oposed using kWh pe dwelling/yea ins ead
o using ene gy consump ion pe squa e me e (kWh/m
2
y ): hey a i med ha in his way mo e
in o ma ion is gi en abou consume s, as hea ing expenses depend on bo h he consump ion pe
squa e me e and he size o he dwelling. ONPE also p oposes bo h uni s o exp ess he ene gy
cos s o he households, and o he case o LIHC/UC he equi aliza ion conside s household size and
composi ion. Howe e , he equi aliza ion scale is no men ioned by ONPE [21].
A ea was included as a a iable in he analysis, as i is a common explana o y a iable o ene gy
consump ion [
41
]. Occupancy is he second common a iable when analyzing ene gy consump ion,
o example, h ough he BREDEM modeling ool, as explained in he in oduc ion.
The socio-demog aphic a iables we e bina y a iables, he e o e mul i a ia e eg ession was
used. I was possible o calcula e he pe cen age inc ease esul ing om each o he a iables, acili a ing
he calcula ion o equi aliza ion weigh s o o each o he socio-demog aphic a iables. In o de no o
calcula e biased coe icien s, since he objec i e was o ob ain a modeled ene gy bill, his econome ic
analysis was pe o med o an a e age consume o each o he coun ies, so ha speci ic cha ac e is ics
we e no o e ep esen ed, e.g., excessi ely low/high ene gy consump ions due o low/high income.
In doing his, he aim o building modeled ene gy bills adjus ed o ulne abili y a iables was ealized.
Sus ainabili y 2020,12, 5721 9 o 21
In o de o compa e ene gy po e y le els using ac ual ene gy bills agains easonable ene gy
cos s, he ob ained coe icien s we e inco po a ed in he modeled ene gy bills. The compa ison be ween
modeled ene gy bills wi h and wi hou coe icien s was also ca ied ou . Wi h ega d o he indica o
applied, al hough bo h he TPR and he LIHC ha e disad an ages, use o he o me is jus i iable.
The in oduc ion highligh ed ha , in in e -coun y compa ison, 2M me hodologies —including LIHC—
do no cap u e he impac o ene gy p ices on hea ing bills o e ime, o he impac s, such as di e ences,
in clima e da a a e also omi ed. The TPR was conside ed a good al e na i e gi en ha i is in ui i ely
unde s ood by many. Subsequen ly, he analysis o es he unc ionali y o he cu en me hodology
was ca ied ou by calcula ing he TPR using ac ual expenses e sus modeled ene gy bills, wi h and
wi hou equi aliza ion coe icien s.
Addi ionally, his me hodology o calcula ing easonable ene gy cos s can be used o
iden i y hose unde consuming by compa ing ene gy po e y measu emen s using ac ual
expenses wi h measu emen s using heo e ical ene gy needs adjus ed wi h socio-economic a iables.
Fo
Rademaeke s e al.
[
5
], unde -consump ion occu ed when households did no spend enough o
achie e a ce ain s anda d o ene gy se ices. He e, he de ini ion assumed o unde consump ion
was simila o he HEP indica o , his is M/2, bu e e ed o modeled ene gy cos s, and in oducing
a cap on income, his is conside ing only he i s i e deciles. I HEP unde lined he exis ence o
sel - a ioning p ac ices [
29
], de ined in HEP me hodology as hose consuming less han hal he median
(M/2), in an analogous manne , in his a icle i was assumed ha a household was unde consuming
when consuming 50% o he heo e ical ene gy cos o less. Conside a ions abou using heo e ical
ene gy cos s ins ead o ac ual expenses hen applied equally o he HEP indica o .
2.3. Da a Se
Fi s ly, he da a used in he co ela ion analysis a e p esen ed.
Rega ding ac ual expenses, occupancy, size o he dwelling, and socio-economic a iables, da a
sou ces o each coun y/ egion a e p o ided below:
•
Fo G eece, su ey da a om Papada and Kaliampakos [
33
] comp ise 400 households ac oss all o
G eece, wi h da a a ailable on ac ual and equi ed ene gy cos s, income, building cha ac e is ics
( ype and size, main hea ing sys em, yea o cons uc ion, egion, and al i ude), and some social
a iables (occupancy and age anges, gende and p esence o pensione s and/o unemployed).
Raw da a we e used o he analysis pe o med, as ecalcula ion o ene gy cos s wi h supplemen a y
in o ma ion was unnecessa y. All da a we e comple e. The co ela ions be ween ac ual ene gy
expenses and he ollowing a iables we e analyzed: young unde 18 yea s old, pensione s,
people o e 60, and unemployed.
•
In Po ugal, a doo - o-doo su ey o 388 households was pe o med in he municipali y o
É
o a [
34
]. DHW was included as well as cooking, ligh ing, and elec ical de ices. Daily egis y
in o ma ion om sma me e s (2014–2017) was also a ailable o his household da ase , he e o e
elec ici y consump ion was e ie ed o be e unde s anding o he le els o consump ion.
A e age elec ici y consump ion in he sampled households was 3682 kWh [
42
], which ep esen s
44% o o al ene gy consump ion, i.e., he a e age o al ene gy consump ion is 8350 kWh; om his
o al, DHW is usually non-elec ic and eaches 2613.5 kWh on a e age, app oxima ely hal o he
3081 kWh o cooking comes om elec ici y, elec ical appliances ep esen 1361 kWh in elec ici y,
and ligh ing an addi ional 560 kWh. A e age da a we e used o ill he in o ma ion gaps in ac ual
ene gy expenses, in cases whe e no eal da a we e a ailable. The sample o 388 households was
educed o a inal sample o 219 households, as some da a om he su ey we e incomple e.
F om hose 219 households, occupancy in he ci y o
É
o a was unknown o h ee da a poin s
and m
2
o one da a poin . The unde 18 a iable was a ailable in his case s udy, as was he
unde 4 ca ego y, bo h we e he e o e analyzed bu only he i s could be compa ed wi h G eek
da a. The e we e no da a a ailable o he o e 60 ca ego y, wi h da a ins ead being a ailable o
Sus ainabili y 2020,12, 5721 16 o 21
consequence hei TPR and HEP indexes esul ed in highe pe cen ages o ene gy po e y compa ed
o indexes calcula ed wi h equi alized modeled ene gy bills.
Table 7. Compa ison be ween ene gy po e y measu emen me hodologies.
Index G eece É o a (Po ugal) Basque Coun y (Spain)
2M Ra io ac ual bills/median 33 ou o 400 (8.2%) 4 ou o 145 (2.8%) 39 ou o 1205 (3.2%)
TPR
Ac ual bills
190 ou o 400 (47.5%)
60 ou o 145 (41.4%) 428 ou o 1205 (35.5%)
Modeled bills 228 ou o 400 (57%) 116 ou o 145 (80%) 762 ou o 1205 (63.2%)
Modeled bills wi h equi . coe . 192 ou o 400 (48%) 99 ou o 145 (68.3%) 715 ou o 1205 (59.3%)
HEP
M/2 32 ou o 400 (8%) 5 ou o 145 (3.4%) 62 ou o 1205 (5.1%)
Modeled bills 64 ou o 400 (16%)
112 ou o 145 (77.2%)
232 ou o 1205 (19.3%)
Modeled bills wi h equi . coe . 33 ou o 400 (8.3%) 12 ou o 145 (8.3%) 197 ou o 1205 (16.3%)
Households
wi h no hea ing
To al 6 ou o 400 1 ou o 145 17 ou o 1205
Non ene gy-poo o 2M 6 ou o 6 Yes 17 ou o 17
Non ene gy-poo o TPR ac ual 3 ou o 6 No 15 ou o 17
Non ene gy-poo o TPR model 2 ou o 6 No 5 ou o 17
Non ene gy-poo o TPR model +coe 3 ou o 6 No 4 ou o 17
Non ene gy-poo o HEP median 3 ou o 6 Yes 5 ou o 17
Non ene gy-poo o HEP model 4 ou o 6 No 1 ou o 17
Non ene gy-poo o HEP model +coe
5 ou o 6 Yes 2 ou o 17
In he case o he TPR indexes, he ob ained esul s when using ac ual bills we e lowe o all he
cases s udied. This means some ac ual bills a e lowe han hei co esponding equi alized modeled
bills. This was also e iden when he ac ual bills we e compa ed wi h hei median (M/2) o wi h
hei co esponding equi alized modeled bills (HEP). Fo each case s udy, he ollowing analysis will
discuss i households wi h no hea ing we e co ec ly iden i ied by he indica o s.
In G eece, all he indica o s had high alues. This a icle used households wi h no hea ing as an
example o compa e di e en indexes, 2M did no iden i y any o hese households. I is clea ha he
2M indica o ailed o iden i y hem as ene gy-poo . Bo h TPR and HEP calcula ed wi h ac ual ene gy
bills we e complemen a y, wi h TPR iden i ying he high ac ual ene gy bills and HEP iden i ying he
lowe ene gy bills.
The ene gy po e y indica o s calcula ed o he ci y o
É
o a we e also high. The e was one
household wi h no hea ing sys em, which should be ca ego ized as ene gy-poo . Once again, 2M did
no iden i y he household wi h no hea ing.
In he Basque egion, he a ailable da a o incomes we e mo e accu a e han in he p e ious wo
case s udies. The indica o s ook high alues o he 1205 households, as low-income households we e
o e ep esen ed. Wi h ega d o he iden i ica ion o households wi h no hea ing, he e we e di e en
hea ing sys ems and di e en easons o no ha ing hea ing; o a mul i-occupancy building wi h
indi idual na u al gas boile s, 6 ou o 222 dwellings had no gas con ac , in he o he 4 buildings
wi h p epaymen cen al hea ing 6 ou o he 361 dwellings had ze o ene gy consump ion, and o he
emaining 8 buildings wi h cen al hea ing (no p epaymen ) 5 ou o 636 dwellings had ze o ene gy
expenses. The 2M was once again he indica o wi h he poo es abili y o iden i y hese households.
In con as , he esul s o indica o s wi h equi alized modeled bills we e ela i ely accu a e.
Fo u u e wo k, gi en ha wo o he samples used o his s udy we e om municipali ies o
egions, i migh be in e es ing o epea he s udy in o de o ex apola e he esul s o he es o he
coun y/ies. Addi ional socio-demog aphic a iables may also be in e es ing o analysis, such as
long- e m illness, disabili ies, and hou s spen a home. I will be in e es ing o compa e empe a u e
da a, o gain insigh in o indoo he mal com o le els. The issue o unde -consump ion is also wo h
in es iga ing by esea ch g oups, i.e., in non-hea ing ene gy-uses [
14
], including e ec s on people’s
quali y o li e.
5. Conclusions
In his a icle he aim was o use an index ha conside ed heo e ical ene gy cos s ins ead o eal
cos s in coun ies/ egions ou side he UK. The alue o de eloping a me hodology o calcula ing

Sus ainabili y 2020,12, 5721 17 o 21
hose heo e ical ene gy needs combined wi h socio-demog aphic a iables is ha he p oblem
o a ge ing households ha consume less ene gy due o es ic ions [
26
] and expe ience cold a
home as a consequence is a oided [
18
]. The in oduc ion highligh ed ha , J. Hills also in oduced
socio-demog aphic a iables in his me hodology, known as ene gy equi aliza ion coe icien s,
highligh ing hese as impo an ac o s o conside .
In o de o calcula e hose heo e ical ene gy cos s, an econome ic analysis was ca ied ou .
A e compa ing ene gy consump ion le els in h ee sou he n EU egion case s udies—G eece, ci y o
É
o a (Po ugal), and he Basque coun y (Spain)— he e ec s o ce ain socio-economic a iables we e
assessed. The esul s a e compa able o hose ob ained by J. Hills.
These esul s we e applied o he calcula ion o indexes using modeled ene gy bills in Po ugal,
Spain, and G eece. In se e al coun ies, ene gy po e y s a is ics a e assessed using ac ual bills.
Households consuming less han 50% o heo e ical ene gy needs due o ha ing low incomes a e
usually no conside ed as ene gy-poo acco ding o he TPR using ac ual expenses. This is a common
p oblem o mos Eu opean coun ies. The HEP was de eloped o iden i y hese households, bu when
using he median o he ac ual bills, his index inaccu a ely cap u es low ene gy cos s due o high
building ene gy e iciency. This can be a oided using easonable ene gy cos s, as in Hills [
9
], o modeled
equi ed ene gy consump ion, as in Papada and Kaliampakos [
30
], he me hodology used he ein.
The conclusion eached o cu en ene gy po e y indica o s, in line wi h he opinion o esea che s
e e enced in he in oduc ion such as Moo e [
18
] and Flo io and Teissie [
26
], among o he s, is ha
ac ual expenses is no a good p oxy o ene gy po e y measu emen ; he ac ha unde -consuming
households and e en hose wi h no hea ing a e no conside ed as ene gy-poo leads o a a ge ing
p oblem. The e o e, heo e ical ene gy cos s should be used in ene gy po e y indexes, including
in hose modeled bills he equi aliza ion ac o s, as p oposed by Hills [
9
]. This is also e ec i e in
in e -coun y compa ison, as heo e ical expenses can be excessi ely high in coun ies wi h low mean
incomes, such as hose in he sou he n EU egions s udied.
Rega ding compa ison ac oss coun ies, building an index ha does no e lec he impac o
ene gy o e ime o calcula es simila pe cen ages o ene gy-poo households ac oss coun ies appea s
un ealis ic. This is he case o indica o s based on 2M. A pi o al ole should be played by he ene gy
cos s- o-income a io; his means ha ade-o s can occu in homes ying o keep hei houses wa m,
e.g., cu ing ood expenses, o hey can all in o deb . This is he so called “hea o ea ” dilemma when
ene gy cos s in uel-poo households compe e wi h o he basic expenses, which can lead o heal h
consequences [
15
]. TPR may seem o be a bi a y, se ing he amoun o ene gy expenses a 10%, o a
no ma i e way o sha ing he income among he di e en expenses. E alua ing ene gy needs should
be pe o med o he desi ed ou come, calcula ing he sha e o he income des ined o ene gy bills
ac oss Eu opean coun ies h ough he ene gy cos s- o-income a io app oach.
Fo policymake s, a he han ene gy p ices and ene gy e iciency, he impo an a iable is ene gy
cos s, namely heo e ical ene gy cos s. Wi h ac ual bills, he index does no cap u e he whole pic u e.
Wi h modeled ene gy cos s, in o ma ion on bo h ene gy e iciency and ene gy p ices is cap u ed:
uel po e y is dis inc om income po e y [
9
], being caused p ima ily by poo ene gy e iciency
and a ailabili y o a o dable ene gy ca ie s. In he end, he ocus should be placed on a o ing
ene gy e iciency measu es. This also has implica ions o ca bon educ ion: in he case o households
unde -using ene gy, aking in o accoun only ac ual bills, in es men s o educe ca bon emissions
canno be eco e ed. O he ypes o bene i s—heal h, employmen , indoo ai quali y, and he mal
com o —should be conside ed o make hose ene gy e iciency measu es ‘p o i able’ [51].
A me hodology like his has also implica ions o u u e legisla ion o o icial s anda ds [
6
].
The da a ex ac ed om HBS a e hose ela ed o he ac ual ene gy bills. I modeled ene gy bills a e o
be used, he p oblem is ha he exis ing building s anda ds o ene gy e iciency a e di e en ac oss EU
coun ies. The Belgium app oach, whe e he o e all hea ans e coe icien o he building en elope
(U- alue) was ad oca ed by Sunikka-Blank and Gal in [
41
] o his pu pose, his app oach is also
compa able o he Ge man EPR. Howe e , inc easing esea ch on he ole o EPCs in ene gy po e y
Sus ainabili y 2020,12, 5721 18 o 21
assessmen and e o i ing o buildings (e.g., Gou eia and Palma [
32
] o
Cha alambides e al.
[
52
]),
as well as new ways o accessing Eu opean building cha ac e is ic da a a e now a ailable, as he
EU Commission launched an ini ia i e o collec da a on ene gy e iciency s anda ds om building
s ock [
53
]. The EU SILC da a can be used o adjus hose socio-economics a iables in all EU coun ies,
as well as he Household Budge Su ey (HBS).
Au ho Con ibu ions:
Concep ualiza ion, I.A.; me hodology, L.P. and I.A.; da a access, L.P., J.P.G. and I.A.;
w i ing—o iginal d a p epa a ion, I.A.; w i ing— e iew and edi ing, L.P. and J.P.G.; Sho Te m Scien i ic Mission
adminis a ion, N.K. and D.K. All au ho s ha e ead and ag eed o he published e sion o he manusc ip .
Funding:
Iñigo An epa a hanks he ENGAGER Ac ion CA16232 “Eu opean Ene gy Po e y: Agenda Co-C ea ion
and Knowledge Inno a ion” o he Sho Te m Scien i ic Mission schola ship awa ded ha was comple ed
du ing Ma ch–Ap il 2018 a NTUA, A hens (G eece). Jo
ã
o Ped o Gou eia acknowledge and hank he suppo
gi en o CENSE by he Po uguese Founda ion o Science and Technology (FCT) h ough he s a egic p ojec
UIDB/04085/2020. The pape s ems om collabo a i e wo k wi hin COST Ac ion ‘Eu opean Ene gy Po e y:
Agenda Co-C ea ion and Knowledge Inno a ion’ (ENGAGER 2017–2021, CA16232) unded by Eu opean
Coope a ion in Science and Technology—www.cos .eu.
Acknowledgmen s:
The au ho s would like o hank Ka he ine Mahoney o he English language e iew. Iñigo
An epa a would like o hank Alokabide o allowing access o da a.
Con lic s o In e es : The au ho s decla e no con lic o in e es .
Appendix A
Table A1. Co ela ion analysis o G eek da a.
Va iable Numbe o Obse a ions Spea man’s ho Pea son
Ac ual ene gy expenses (log) s occupancy 400 0.2964 ** 0.3028 **
Ac ual ene gy expenses (log) s m2400 0.4577 ** 0.47 **
Socio-demog aphic a iables
Ac ual ene gy expenses (log) s u18 y . old 400 0.1386 ** 0.1381 **
Ac ual ene gy expenses (log) s Pensione s 400 −0.1554 ** −0.1599 **
Ac ual ene gy expenses (log) s o e 60 400 −0.0873 ns −0.0822 ns
Ac ual ene gy expenses (log) s Unemployed 400 −0.0866 ns −0.1019 *
* signi ican a he 0.05 le el, ** signi ican a he 0.01 le el, and ns no s a is ically signi ican .
Table A2. Co ela ion analysis o Po uguese da a.
Va iable Numbe o Obse a ions Spea man’s ho Pea son
Ac ual ene gy expenses (log) s occupancy 216 0.4047 ** 0.3744 **
Ac ual ene gy expenses (log) s m2218 0.3419 ** 0.2666 **
Socio-demog aphic a iables
Ac ual ene gy expenses (log) s u4 y . old 219 0.1135 ns 0.0686 ns
Ac ual ene gy expenses (log) s u18 y . old 219 0.2531 ** 0.2323 **
Ac ual ene gy expenses (log) s Re i ed 219 −0.1763 ** −0.1537 *
Ac ual ene gy expenses (log) s o e 65 219 −0.1749 ** −0.1476 *
* signi ican a he 0.05 le el, ** signi ican a he 0.01 le el, and ns no s a is ically signi ican .
Table A3. Co ela ion analysis o Basque da a.
Va iable Numbe o Obse a ions Spea man’s ho Pea son
Ac ual ene gy expenses (log) s occupancy 1216 0.1137 ** 0.1097 **
Ac ual ene gy expenses (log) s m21219 0.0389 ns 0.0454 ns
Socio-demog aphic a iables
Ac ual ene gy expenses (log) s u4 y . old 489 −0.1215 ** −0.12 **
Ac ual ene gy expenses (log) s u18 y . old 489 −0.1376 ** −0.1234 **
Ac ual ene gy expenses (log) s o e 65 505 −0.0323 ns −0.0629 ns
** signi ican a he 0.01 le el, and ns no s a is ically signi ican .
Sus ainabili y 2020,12, 5721 19 o 21
Appendix B
Table A4. Equi aliza ion ac o s p oposed by Hills [9].
Household Type Equi aliza ion Fac o
Couple wi h dependen child en 1.15
Couple wi hou dependen child en 1.00
Lone pa en 0.94
Single pe son 0.82
O he mul i-pe son household 1.07
Table A5. P oposed ene gy cos equi aliza ion ac o s by Moo e e al. [20].
Household Size
Dwelling Size-Useable Floo A ea (m2)
<50 50–69 70–89 90–109 110+
Single pe son 0.63 0.77 0.90 0.96 1.22
Two people 0.70 0.84 1.00 1.08 1.28
Th ee people 0.78 0.86 1.01 1.17 1.44
Fou people 0.84 0.92 1.07 1.20 1.53
Fi e+people 0.87 1.00 1.14 1.31 1.62
Re e ences
1.
Da abase–Eu os a . A ailable online: h ps://ec.eu opa.eu/eu os a /da a/da abase (accessed on
29 Ma ch 2020).
2.
Day, R.; Walke , G.; Simcock, N. Concep ualising ene gy use and ene gy po e y using a capabili ies
amewo k. Ene gy Policy 2016,93, 255–264. [C ossRe ]
3.
Eu opean Commission EU Ene gy Po e y Obse a o y 2018. A ailable online: h ps://www.ene gypo e y.eu/
(accessed on 21 June 2020).
4.
Gou eia, J.P.; Palma, P.; Simoes, S.G. Ene gy po e y ulne abili y index: A mul idimensional ool o iden i y
ho spo s o local ac ion. Ene gy Rep. 2019,5, 187–201. [C ossRe ]
5.
Rademaeke s, K.; Yea wood, J.; Fe ei a, A.; Pye, S.; Hamil on, I.; Agnolucci, P.; G o e , D.; Ka
á
sek, J.;
Anisimo a, N. Selec ing Indica o s o Measu e Ene gy Po e y; T inomics: Ro e dam, The Ne he lands, 2016.
6.
Schuessle , R. Ene gy Po e y Indica o s: Concep ual Issues-Pa I: The Ten-Pe cen -Rule and Double Median/
Mean Indica o s; Discussion Pape 14-037; ZEW-Cen e o Eu opean Economic Resea ch: Mannheim,
Ge many, 2014.
7.
Boa dman, B. Fuel Po e y: F om Cold Homes o A o dable Wa m h; Belha en P ess: London, NY, USA, 1991;
ISBN 1-85293-139-6.
8. DECC. Fuel Po e y Me hodology Handbook; BRE: London, UK, 2010.
9.
Hills, J. Ge ing he Measu e o Fuel Po e y: Final Repo o he Fuel Po e y Re iew; Cen e o Analysis o Social
Exclusion; The London School o Economics and Poli ical Science; Commissioned by Depa men o Ene gy
and Clima e Change (DECC): London, UK, 2012.
10.
P obe , L.; Haines, V.; Lo eday, D. The Hills Fuel Po e y Re iew In e im Repo : Assessing P oposals and
Implica ions. People Place Policy Online 2011,5, 161–173. [C ossRe ]
11.
Sa een, S.; Thomson, H.; He e o, S.T.; Gou eia, J.P.; Lippe , I.; Lis, A. Eu opean ene gy po e y me ics:
Scales, p ospec s and limi s. Glob. T ansi . 2020,2, 26–36. [C ossRe ]
12. Liddell, C.; Mo is, C. Fuel po e y and human heal h: A e iew o ecen e idence. Ene gy Policy 2010,38,
2987–2997. [C ossRe ]
13. DECC. Fuel Po e y Me hodology Handbook; BRE: London, UK, 2016.
14.
Simcock, N.; Walke , G.; Day, R. Fuel po e y in he UK: Beyond hea ing. People Place Policy
2016
,10, 25–41.
[C ossRe ]
15.
Hills, J. Fuel Po e y: The P oblem and I s Measu emen . In e im Repo o he Fuel Po e y Re iew; Cen e o
Analysis o Social Exclusion, LSE: London, UK, 2011.
Sus ainabili y 2020,12, 5721 20 o 21
16.
Hagenaa s, A.J.; De Vos, K.; Asgha Zaidi, M. Po e y S a is ics in he La e 1980s: Resea ch Based on Mic o-Da a;
O ice o O icial Publica ions o he Eu opean Communi ies: Luxembou g, 1994.
17.
Bouza o ski, S.; Pe o a, S. A global pe spec i e on domes ic ene gy dep i a ion: O e coming he ene gy
po e y– uel po e y bina y. Ene gy Res. Soc. Sci. 2015,10, 31–40. [C ossRe ]
18. Moo e, R. De ini ions o uel po e y: Implica ions o policy. Ene gy Policy 2012,49, 19–26. [C ossRe ]
19.
Heindl, P.; Schuessle , R. Dynamic p ope ies o ene gy a o dabili y measu es. Ene gy Policy
2015
,86,
123–132. [C ossRe ]
20.
Moo e, R.; B idgeman, T.; C o , D.; Gue le , P.; P es on, I. Imp o ing he Hills App oach o Measu ing Fuel
Po e y; Cen e o Sus ainable Ene gy: London, UK, 2012.
21.
Imbe , I.; Nogues, P.; Se ene , M. Same bu di e en : On he applicabili y o uel po e y indica o s ac oss
coun ies—Insigh s om F ance. Ene gy Res. Soc. Sci. 2016,15, 75–85. [C ossRe ]
22.
Healy, J.D. Excess win e mo ali y in Eu ope: A c oss coun y analysis iden i ying key isk ac o s. J. Epidemiol.
Communi y Heal h 2003,57, 784–789. [C ossRe ] [PubMed]
23.
Bouza o ski, S. ENGAGER Policy B ie 1, COST Ac ion CA16232; Uni e si y o Manches e : Manches e ,
UK, 2018.
24.
De Ciencias Ambien ales, A. Pob eza Ene g
é
ica en España: Hacia un Sis ema de Indicado es y una Es a egia de
Ac uaci
ó
n Es a al [Ene gy Po e y in Spain: Towa ds a Sys em o Indica o s and a S a e Ac ion S a egy]; Asociaci
ó
n
de Ciencias Ambien ales: Mad id, Spain, 2018.
25.
EU Ene gy Po e y Obse a o y Indica o s & Da a. A ailable online: h ps://www.ene gypo e y.eu/
indica o s-da a (accessed on 6 June 2020).
26.
Flo io, P.; Teissie , O. Es ima ion o he Ene gy Pe o mance Ce i ica e o a housing s ock cha ac e ised ia
quali a i e a iables h ough a ypology-based app oach model: A uel po e y e alua ion ool. Ene gy Build.
2015,89, 39–48. [C ossRe ]
27.
Palma, P.; Gou eia, J.P.; Simoes, S.G. Mapping he ene gy pe o mance gap o dwelling s ock a high- esolu ion
scale: Implica ions o he mal com o in Po uguese households. Ene gy Build.
2019
,190, 246–261. [C ossRe ]
28.
Middlemiss, L. A c i ical analysis o he new poli ics o uel po e y in England. C i . Soc. Policy
2017
,37,
425–443. [C ossRe ]
29.
Meye , S.; Lau ence, H.; Ba , D.; Middlemiss, L.; Ma
é
chal, K. Cap u ing he mul i ace ed na u e o ene gy
po e y: Lessons om Belgium. Ene gy Res. Soc. Sci. 2018,40, 273–283. [C ossRe ]
30.
Papada, L.; Kaliampakos, D. A S ochas ic Model o ene gy po e y analysis. Ene gy Policy
2018
,116, 153–164.
[C ossRe ]
31.
Papada, L.; Kaliampakos, D. De elopmen o ulne abili y index o ene gy po e y. Ene gy Build.
2019
,183,
761–771. [C ossRe ]
32.
Gou eia, J.P.; Palma, P. Ha es ing big da a om esiden ial building ene gy pe o mance ce i ica es:
Re o i ing and clima e change mi iga ion insigh s a a egional scale. En i on. Res. Le .
2019
,14, 095007.
[C ossRe ]
33.
Papada, L.; Kaliampakos, D. Measu ing ene gy po e y in G eece. Ene gy Policy
2016
,94, 157–165. [C ossRe ]
34.
Gou eia, J.P.; Seixas, J.; Long, G. Mining households’ ene gy da a o disclose uel po e y: Lessons o
Sou he n Eu ope. J. Clean. P od. 2018,178, 534–550. [C ossRe ]
35.
Eu opean Commission Eu os a . A ailable online: h ps://ec.eu opa.eu/eu os a /(accessed on 16 No embe 2019).
36.
PORDATA PORDATA—Es a
í
s icas, g
á
icos e Indicado es de Munic
í
pios, Po ugal e Eu opa. A ailable
online: h ps://www.po da a.p /(accessed on 16 No embe 2019).
37.
EUSTAT Da os es ad
í
s icos de la C.A. de Euskadi. A ailable online: h p://es.eus a .eus/indice.h ml (accessed
on 17 No embe 2019).
38.
INE Po al do INE. A ailable online: h ps://www.ine.p /xpo al/xmain?xpgid=ine_main&xpid=INE
(accessed on 6 June 2020).
39. INE. Ins i u o Nacional de Es adís ica. A ailable online: h ps://www.ine.es/(accessed on 6 June 2020).
40.
Bai , S.; Belaid, F.; Teissie , O. Quels enseignemen s i e de l’enque e Phebus su la ques ion de l’e e ebond?
Les Ménages e La Consomm. d’Éne gie 2017, 101–113.
41.
Sunikka-Blank, M.; Gal in, R. In oducing he p ebound e ec : The gap be ween pe o mance and ac ual
ene gy consump ion. Build. Res. In . 2012,40, 260–273. [C ossRe ]
42. INE. Consumo de Ene gia no Sec o Domés ico 2010; INE; DGEG: Lisboa, Po ugal, 2011.
43.
De Minis e io, V. C
ó
digo T
é
cnico de la Edi icaci
ó
n. Documen o B
á
sico HE; Aho o de ene g
í
a: Mad id, Spain, 2017.
Sus ainabili y 2020,12, 5721 21 o 21
44. EVE. Cla es ene gé icas del sec o domés ico en Euskadi; En e Vasco de la Ene gía: Bilbao, Spain, 2013.
45.
Uni e si y o Ox o d Ou Wo ld in Da a 2019. A ailable online: h ps://ou wo ldinda a.o g/(accessed on
21 June 2020).
46.
Raynaud, M.; Bou ges, B.; Osso, D.; Duplessis, B.; Adno , J. Bo om-up S a is ical Analysis o he Ene gy
Consump ion o F ench Single-Family Dwellings; hal-01133743; A chi e ou e e HAL: Lyon, F ance, 2014.
47.
Denjean, M. Compo emen s e consomma ions
é
ne g
é
iques. Les M
é
nages e La Consomm. d’
É
ne gie
2017
,
60–67.
48.
Ca ailh
è
s, J.; Joly, D.; B ossa d, T.; Ca do , H.; Hilal, M.; Wa esky, P. La Consomma ion d’
É
ne gie des M
é
nages
en F ance; INRA/CNRS o he CGDD: Dijon, F ance, 2011.
49.
Risch, A.; Salmon, C. Wha Ma e s in Residen ial Ene gy Consump ion? E idence om F ance; hal-01081953;
A chi e Ou e e HAL: Lyon, F ance, 2013.
50.
Ti ado He e o, S. Ene gy po e y indica o s: A c i ical e iew o me hods. Indoo Buil En i on.
2017
,26,
1018–1031. [C ossRe ]
51.
Uga e, S.; an de REE, B.; Voog , M.; Eichhamme , W.; O doñez, J.A.; Reu e , M.; Schlomann, B.;
Llo e Gallego, P.; Villa a ila Robles, R. Ene gy E iciency o Low-Income Households; Policy Depa men A:
Economic and Scien i ic Policy; Eu opean Pa liamen : B ussels, Belgium, 2016.
52.
Cha alambides, A.G.; Maxoulis, C.N.; Ky iacou, O.; Blakeley, E.; F ances, L.S. The impac o Ene gy
Pe o mance Ce i ica es on building deep ene gy eno a ion a ge s. In . J. Sus ain. Ene gy
2019
,38, 1–12.
[C ossRe ]
53.
Eu opean Commission EU Building S ock Obse a o y. A ailable online: h ps://ec.eu opa.eu/ene gy/ opics/
ene gy-e iciency/ene gy-e icien -buildings/eu-bso_en (accessed on 3 June 2020).
©
2020 by he au ho s. Licensee MDPI, Basel, Swi ze land. This a icle is an open access
a icle dis ibu ed unde he e ms and condi ions o he C ea i e Commons A ibu ion
(CC BY) license (h p://c ea i ecommons.o g/licenses/by/4.0/).