Inequali y ac oss coun ies in ene gy
in ensi ies: an analysis o he ole o ene gy
ans o ma ion and inal ene gy consump ion.
Juan An onio Du o
Emilio Padilla
10.08
Facul a d'Economia i Emp esa
De
p
a amen d'Economia A
p
licada
Aques documen pe any al Depa amen d'Economia Aplicada.
Da a de publicació :
Depa amen d'Economia Aplicada
Edi ici B
Campus de Bella e a
08193 Bella e a
Telè on: (93) 581 1680
Fax:(93) 581 2292
E-mail: [email p o ec ed]
h p://www.ecap.uab.es
Desemb e 2010
Inequali y ac oss coun ies in ene gy in ensi ies: an analysis o he ole o
ene gy ans o ma ion and inal ene gy consump ion
Juan An onio Du oa and Emilio Padillab
a Depa amen o de Economía, Uni e sidad Ro i a i Vi gili. A . de la Uni e si a ,
1. 43204 Reus, España. E-mail: [email p o ec ed]
b Depa amen o de Economía Aplicada, Uni e sidad Au ónoma de Ba celona
08193 Bella e a, España. E-mail: [email p o ec ed]
Abs ac
This pape analyzes he ole o he ene gy ans o ma ion index and o inal
ene gy consump ion pe GDP uni in he dispa i ies in ene gy in ensi y ac oss
coun ies. In ha ein, we use a Theil decomposi ion app oach o analyze
global p ima y ene gy in ensi y inequali y as well as inequali y ac oss di e en
egions o he wo ld and inequali y wi hin hese egions. The pape i s
demons a es he p e-eminence o di e gence in inal ene gy consump ion pe
GDP uni in explaining global p ima y ene gy in ensi y inequali y and i s
e olu ion du ing he 1971–2006 pe iod. Secondly, i shows he lowe (albei non
negligible) impac o he ans o ma ion index in global p ima y ene gy
inequali y. Thi dly, he ele ance o egions as uni o analysis in s udying c oss-
coun y ene gy in ensi y inequali y and hei explana o y ac o s is highligh ed.
And inally, how egions a ound he wo ld di e as o he ele ance o he
ene gy ans o ma ion index in explaining p ima y ene gy in ensi y inequali y.
Keywo ds: Ene gy e iciency, ene gy in ensi ies, ene gy ans o ma ion, Theil
index.
JEL code: C69, D39, Q43.
1
1. In oduc ion
Ene gy consump ion and CO2 emissions pe capi a di e la gely ac oss
coun ies and egions a ound he wo ld. Consequen ly, people be ween
coun ies and egions con ibu e a di e en deg ees o he in ensi y o he
g eenhouse e ec . Se e al s udies ha e analyzed such di e ences using
dis ibu i e analysis ools and ha e d awn ene gy and clima ic policy
implica ions (see among o he s, Heil and Wodon, 1997, 2000; Millime and
Slo je, 2002; Hedenus and Aza , 2005; Padilla and Se ano, 2006; Du o and
Padilla, 2006, 2008; Can o e and Padilla, 2010a, 2010b; G oo , 2010).
One o he mos commonly used ools o analyze he d i ing o ces o emissions
and hei e olu ion is he so-called Kaya iden i y (Kaya, 1989; Yamaji e al.,
1991; Alcán a a and Padilla, 2005). Acco ding o he Kaya iden i y, he ac o s
explaining he e olu ion o pe capi a emissions a e he ca bon in ensi y o
ene gy, he ene gy in ensi y o GDP, and he a luence, which is usually
measu ed in e ms o GDP pe capi a. In a p e ious a icle in his jou nal, Du o
and Padilla (2006) used a Theil index decomposi ion ha allows decomposing
inequali y in o di e en Kaya ac o s and wo in e ac ion e ms o analyze he
de e minan s o c oss-coun y inequali y in CO2 emissions pe capi a be ween
1971 and 1999. The au ho s highligh ed he g ea e impo ance (al hough
dec easing o e ime) o income inequali y in explaining di e ences in
emissions pe capi a. Ne e heless, hey also highligh ed he impo ance o
ca boniza ion index inequali ies and ene gy in ensi ies in globally educing
inequali y in pe capi a emissions. As ega ds ene gy in ensi ies, he au ho s
poin o he exis ence o conside able di e gence wo ldwide. Howe e , due o
he educ ion o ene gy in ensi y in some de eloping coun ies such di e gence
has been educed. Ano he inding om hese au ho s is he ele ance o
con e gence in ene gy in ensi ies in educing CO2 inequali y in he Tempe a e
zone g oup o coun ies, which is basically composed o ich coun ies. Sun
(2002) and Alcán a a and Du o (2004) also show a down end in ene gy
in ensi y inequali y among he OECD coun ies. Sun (2002) uses mean
de ia ion as a dispe sion measu e o analyze his inequali y. Alcán a a and
2
Du o (2004) analyze his inequali y by means o he Theil index. In a ecen
pape , Du o e al. (2010) use a Theil index decomposi ion which allows he
inequali y in ene gy consump ion pe capi a o be decomposed in o explana o y
ac o s. They demons a e ha , al hough di e ences in a luence a e he mos
signi ican ac o in explaining inequali y in ene gy consump ion pe capi a, he
educ ion o he inequali y in ene gy in ensi y le els plays a p ominen ole in
educing he inequali y in ene gy consump ion pe capi a be ween 1980 and
2006 among OECD coun ies.
The educ ion in di e ences in p ima y ene gy in ensi y ound in he ci ed
s udies may ha e a ious causes. I may s em om a mo e e icien way o
ans o ming p ima y ene gy in o inal ene gy. O , i may in con as be he
consequence o con e gence in inal ene gy use pe GDP uni among
coun ies1. Tha is, such a decline may be a ibu able o changes in he ene gy
sec o o o changes in inal ene gy consump ion in he economy.2 De e mining
he ela i e impo ance o bo h ac o s is use ul bo h o analy ical pu poses and
he o mula ion o policy ecommenda ions. Acco dingly, a signi ican weigh o
he ene gy ans o ma ion index would sugges a la ge scope o imp o emen
in coun ies no e icien in ans o ming ene gy. In ha ein, ene gy policies
could success ully educe ene gy in ensi y inequali ies by con e ging o g ea e
e iciency le els in ene gy ans o ma ion —ei he imp o ing he e iciency o
ene gy con e sion p ocesses o changing he ene gy mix. In con as , a
signi ican weigh o he inal ene gy in ensi y componen , would sugges ha
measu es implemen ed o achie e g ea e e iciency in inal ene gy
1 This second componen can in u n be due o wo di e en ac o s. I could be caused by
ene gy consump ion e iciency con e gence among coun ies, o may be due o con e gence in
p oduc ion composi ion. Du o e al. (2010) de elop a shi -sha e me hodology o analyze hese
ac o s o 16 OECD coun ies. They conclude ha he con e gence in inal ene gy consump ion
pe GDP uni could be a ibu ed o a con e gence in he e iciency in he use o inal ene gy
ac oss coun ies. Meanwhile, he di e ence in p oduc ion composi ion could ha e inc eased
wi hou o se ing he i s e ec . Mike a and Mulde (2005) un an econome ic analysis o he
inal ene gy p oduc i i y con e gence ac oss 56 coun ies in 10 manu ac u ing sec o s. They
ound ha he di e ences in he inal ene gy in ensi y le els o hese sec o s diminished ac oss
some coun ies.
2 The impo ance o aking in o accoun con e sion-e iciency was showed by Hamil on and
Tu on (2002) who employed a decomposi ion o mula ha sepa a ed ou he e ec s on
emissions g ow h o changes in popula ion, economic g ow h, ene gy in ensi y, ene gy
ans o ma ion index, sha e o ossil uels and ca bon in ensi y o ossil uels. They ound ha
changes in ene gy ans o ma ion con ibu ed o inc ease emissions in OECD coun ies o e he
pe iod 1982–1997.
3

consump ion ( om a be e use o he di e en sec o s) o o con e gence
owa d sec o s ha use less ene gy would be he bes ways o educe
di e ences in ene gy in ensi ies and lowe ene gy consump ion pe GDP uni .
This pape complemen s he li e a u e and makes an o iginal con ibu ion
in ending o disce n he weigh o di e ences wi hin he ene gy sec o and hose
ela i e o inal ene gy in ensi y in de e mining he di e ences in ene gy
in ensi ies in e na ionally and i s e olu ion. In ha espec , we use he Theil
inequali y index as he syn he ic benchma k index since i allows decomposi ion
by pa s. Two ypes o decomposi ions will be done: he mul iplica i e
decomposi ion pionee ed by Du o (2003) and implemen ed in he ene gy
analysis in Du o and Padilla (2006), and g oup decomposi ion (Sho ocks,
1980).
The pape is o ganized as ollows. Sec ion 2 lays ou he me hodology used. In
Sec ion 3, we p esen he empi ical esul s on c oss-coun y ene gy in ensi y
inequali ies o he pe iod 1971–2006. Finally, Sec ion 4 p esen s some
concluding ema ks.
2. Me hodology
Assessing he ole he ene gy ans o ma ion index and inal ene gy
consump ion pe GDP uni ha e played in he e olu ion o ene gy in ensi y
inequali ies s a s i s om a simple bi ac o ial b eakdown o ene gy in ensi y in
he ollowing way:
*
i i i
i i i
PE PE FE
GDP FE GDP
= (1)
ei = i * wi (2)
whe e PEi is p ima y ene gy consump ion o coun y i in pe iod , FEi is inal
ene gy consump ion, GDPi is he g oss domes ic p oduc . Then, ei is ene gy
in ensi y, i is he ans o ma ion index, which measu es he e iciency o he
4
ene gy sec o in ans o ming ene gy3, while wi cap u es he inal ene gy
consump ion pe GDP uni ( inal ene gy in ensi y index).
Secondly, o cla i y he ole o bo h ac o s in explaining ene gy in ensi y
inequali ies ac oss coun ies, we de ine wo hypo he ical ec o s o p ima y
ene gy consump ion pe GDP uni and we le jus one o he alues o he
ac o s included in (2) di e ge om he mean. Acco dingly, we ob ain he
ollowing ic i ious ac o s4:
*
i i
e w= (3)
*
w
i i
e w= (4)
whe e y w a e wo ld a e ages o he ac o being conside ed.
Reso ing o Du o’s (2003) me hodology and Du o and Padilla (2006), using he
Theil index (Theil, 1967)5 as he benchma k inequali y index allows a syn he ic
decomposi ion o global ene gy in ensi y inequali ies in o h ee ac o s:
()
()()
,
,,,1
w
w
Tep Te p Te p log e
σ
⎛⎞
=+++
⎜
⎝⎠
⎟
(5)
T(e,p) = T + Tw + in e ,w (6)
whe e
σ
is he co a iance (weigh ed) be ween he wo ac o s and
eis he
wo ld a e age o he i s ic i ious ac o ; e is he ene gy in ensi y and p he
weigh in global GDP.
3 Ne e heless, he index will no only depend on how e icien coun ies a e in he con e sion o
one o o he ype o ene gy bu also on hei di e en ene gy mix. Fo ins ance, acco ding o he
In e na ional Ene gy Agency, he hea gene a ed by nuclea powe plan s is conside ed p ima y
ene gy while o hyd o-elec ic s a ions, wind o pho o ol aic sola powe sys em, only he
ene gy alue o he elec ici y gene a ed is aken in o accoun . Consequen ly, he e iciency in
ans o ming nuclea ene gy is less han in he case o ossil uels, while i is always g ea e o
enewable ene gy.
4 This me hodology as de eloped by Du o (2003) o analyze spa ial income inequali y.
5 The Theil index has been used in di e en wo ks on en i onmen al dis ibu ion (Alcán a a and
Du o, 2004; Du o and Padilla, 2006; Padilla and Se ano, 2006; Padilla and Du o, 2009;
Can o e and Padilla, 2010a; Du o e al., 2010). Cowell (1995) highligh s i s analy ical
ad an ages, which include i s abili y o decompose addi i ely a se ies o mul iplica i e ac o s.
This is due o he ac ha i is a loga i hmic unc ion.
5
Thus, o al inequali y is pe ec ly decomposed in o wo indices ha cap u e he
pa ial con ibu ion o each mul iplica i e ac o accoun ed o in global inequali y
(T cap u es he con ibu ion o ene gy ans o ma ion index and Tw he
con ibu ion o inal ene gy in ensi y index), and an in e ac ion e m ep esen ing
he in e ac o ial co ela ion (in e ,w)6. A posi i e alue o his las componen
would sugges ha coun ies ha a e no e icien in ene gy ans o ma ion
would also end o be ine icien in ene gy use. So, he wo inequali ies would be
sel - ein o cing. In u n, a nega i e alue would mean ha less e icien
coun ies in ene gy ans o ma ion end o be mo e ine icien in inal ene gy
consump ion.
I should be no ed ha , as he ac o s ha e been o mula ed in equa ions (3)
and (4), he impo ance o each ac o in he decomposi ion exe cise can be
seen as he a ia ion ac oss coun ies o he ac o unde analysis, while he
emaining ac o s a e se equal o mean.
On he o he hand, his ac o ial decomposi ion me hodology can be ex ended
o subg oup componen s o inequali y. Tha is, he p e ious mul iplica i e
ac o ial decomposi ion can be combined wi h subg oup decomposi ion. This
would di ide global inequali y in o an elemen o in e -g oup inequali y and
ano he o in a-g oup inequali y. The well known Theil index can be easiliy
decomposed in o popula ion subg oup. We adap i he e o he s udy o
inequali ies in ene gy in ensi y ac oss coun ies (Theil, 1967; Sho ocks, 1980)
() ()
11
,
GG
gg
g
gg
g
e
Tep pTe p e
==
⎛⎞
=+
⎜
⎜
⎝⎠
∑∑
*ln
⎟
⎟
(7)
T(e,p) = Twi hin + Tbe ween (8)
6 Mind ha i in addi ion
e
w,
σ is su icien ly small, he decomposi ion could be app oached as:
()
e
w,
p,
w
eTp,
eTp,eT σ
+
⎟
⎠
⎞
⎜
⎝
⎛
+
⎟
⎠
⎞
⎜
⎝
⎛
≈
6
Whe e pg is he weigh o g oup g in global GDP; Tg deno es ene gy in ensi y
inequali y wi hin g oup g; eg ep esen s g oup g ene gy in ensi y; ē is he wo ld
a e age ene gy in ensi y.
The i s e m — he in a-g oup componen (Twi hin)— is a weigh ed a e age o
in a-g oups Theil indexes. The e o e, a mul iplica i e decomposi ion is
s aigh o wa d acco ding o (5). The second e m — he in e -g oup componen
(Tbe ween)— is simply a Theil index e alua ing di e ences among g oups. So, he
p e ious mul iplica i e bi ac o ial decomposi ion o equa ion (5) is also
s aigh o wa d. The e o e, o he egional analysis we will i s apply a
decomposi ion acco ding o equa ion (6) and a a second s age a
decomposi ion on he basis o equa ion (5) o each o he inequali y
componen s ob ained in he i s s age.
3. Empi ical indings
The da a used o he analysis a e he ene gy balances om he In e na ional
Ene gy Agency (IEA, 2009a, 2009b). The selec ed sample includes 116
coun ies and basically co e s he en i e wo ld GDP and p ima y and inal
ene gy. Fo consis ency eason, o e he en i e pe iod we g oup he coun ies
o he o me USSR and also hose o he o me Yugosla ia. Fo he coun ies
conside ed in he analysis pe g oup (whose esul s a e con ained in Table 5),
we use he IAE classi ica ion as a e e ence. The classi ica ion is done
acco ding o economic and geog aphic c i e ia and conside s he ollowing nine
wo ld egions: OECD Eu ope, non-OECD Eu ope, No h Ame ica, OECD
Paci ic, A ica, La in Ame ica, Middle Eas , Asia and China. Annex 1 gi es a
de ail o he coun ies included in each g oup. Table 1 below gi es an o e iew
o he s a is ics o he sample used and o he di e en egions conside ed in
he analysis.
7
he pe iod. Two hi ds o he o al di e ences in ans o ma ion indexes occu
wi hin he egions conside ed. Tha is, he g ea e con ibu ion o global
di e ences in ene gy ans o ma ion e iciency occu s pa icula ly wi hin he
ela i ely homogenous g oups o coun ies aken in o accoun .
The in e ac ion ac o plays a e y impo an ole in in a-g oup global
inequali ies o ene gy in ensi y, pa icula ly a he end o he pe iod. The
nega i e sign on his ac o sugges s ha coun ies ha a e mo e e icien in
ene gy ans o ma ion a e mos likely he mo e in ensi e ones in inal ene gy
consump ion, and his ends o o se inequali ies.
Table 5. Mul iplica i e ac o ial ene gy in ensi y inequali y decomposi ion
by egion o 1971, 1990 and 2006.
Ene gy
in ensi y
inequali y
weigh
T ans o ma ion
componen
T
Final use
componen
Tw
In e ac ion
componen
in e ,w
OECD
Eu ope
1971
0.0635 30.5%
0.0048
(7.5%)
0.0477
(75.2%)
0.0110
(17.3%)
1990
0.0372 26.8%
0.0029
(7.8%)
0.0297
(79.9%)
0.0046
(12.3%)
2006
0.0148 21.9%
0.0030
(20.4%)
0.0107
(71.8%)
0.0011
(7.7%)
No h
Ame ica
1971
0.0243 25.8%
0.0002
(0.7%)
0.0232
(95.7%)
0.0009
(3.7%)
1990
0.0052 25.3%
0.0006
(11.1%)
0.0063
(121.2%)
-0.0017
(-32.3%)
2006
0.0038 23.2%
0.0005
(12.7%)
0.0072
(189.7%)
-0.0039
(-102.4%)
14

OECD
Paci ic
1971
0.0036 9.6%
0.0004
(10.6%)
0.0019
(51.0%)
0.0014
(38.5%)
1990
0.0141 11.3%
0.0001
(0.6%)
0.0132
(93.7%)
0.0008
(5.6%)
2006
0.0119 9.2%
0.0003
(2.2%)
0.0117
(98.4%)
-0.0001
(-0.6%)
Non-OECD
Eu ope
1971
0.0078 10.6%
0.0010
(12.4%)
0.0096
(122.3%)
-0.0027
(-34.7%)
1990
0.0135 8.3%
0.0005
(3.9%)
0.0157
(116.8%)
-0.0028
(-20.7%)
2006
0.0313 4.8%
0.0009
(2.8%)
0.0309
(99.0%)
-0.0005
(-1.8%)
A ica
1971
0.2491 4.4%
0.0183
(7.3%)
0.3127
(125.6%)
-0.0819
(-32.9%)
1990
0.1320 4.0%
0.0230
(17.4%)
0.1975
(149.5%)
-0.0884
(--66.9%)
2006
0.1151 3.8%
0.0270
(23.5%)
0.1873
(162.7%)
-0.0993
(-86.2%)
La in
Ame ica
1971
0.0346 6.6%
0.0192
(55.6%)
0.0305
(88.1%)
-0.0151
(-43.7%)
1990
0.0504 6.3%
0.0051
(10.1%)
0.0339
(67.3%)
0.0114
(22.6%)
2006
0.0641 6.0%
0.0033
(5.1%)
0.0599
(93.4%)
0.0010
(1.5%)
Middle Eas
1971 0.0802 2.6% 0.0643 0.1267 -0.1107
15
(80.1%) (157.9%) (-138.0%)
1990
0.0543 2.3%
0.0086
(15.9%)
0.0762
(140.3%)
-0.0305
(-56.2%)
2006
0.0803 2.5%
0.0210
(26.2%)
0.0952
(118.6%)
-0.0360
(-44.8%)
Asia
1971
0.1088 7.2%
0.0084
(7.7%)
0.1246
(114.5%)
-0.0242
(-22.2%)
1990
0.0467 9.9%
0.0093
(20.0%)
0.0654
(139.9%)
-0.0280
(-59.9%)
2006
0.0350 13.3%
0.0069
(19.6%)
0.0457
(130.3%)
-0.0175
(-49.9%)
China
1971
0.0590 2.7%
0.0006
(1.1%)
0.0663
(112.5%)
-0.0080
(-13.6%)
1990
0.0446 5.9%
0.0005
(1.1%)
0.0507
(113.7%)
-0.0066
(-14.8%)
2006
0.0095 15.5%
0.0000
(0.0%)
0.0096
(101.2%)
0.0095
(-1.2%)
Sou ce: Au ho s’ own calcula ions based on he IEA (2009a, 2009b). The pe cen ages in he
second column show he weigh o GDP ela i e o global GDP while in he emaining columns
he weigh s o he di e en componen s in in a-g oup inequali y a e gi en.
The i s column shows he c oss coun y inequali y wi hin each o he egions
conside ed. The ac o ial decomposi ion analysis o he di e en egions
p o ides a much mo e de ailed and in e es ing in o ma ion. Fo ins ance, i
allows iden i ying in which g oups he weigh o bo h he ans o ma ion index
and he in e ac ion componen a e ele an enough o ha e inc eased hei
impo ance in he in a-g oup componen analyzed p e iously. This way, we a e
able o iden i y ce ain di e gen pa e ns ha a e e lec ed in he beha iou o
he di e en ac o s a egional le el.
16
Fi s ly, he esul s show ha , con a y o he global endency, ene gy in ensi y
inequali ies ha e inc eased in ou among he nine egions conside ed. These
a e OECD Paci ic, non-OECD Eu ope, La in Ame ica —whe e inequali y has
almos doubled— and he Middle Eas . This endency has howe e been mo e
han o se by he downwa d ajec o y o some egions wi h g ea e ela i e
weigh (wi h espec o global GDP).
A common ea u e o all he egions is he p e-eminence o di e ences in inal
ene gy consump ion in ensi y as de e minan s o he di e ences in ene gy
in ensi y. This occu s despi e he signi ican di e ences in he con ibu ion o
his ac o obse ed among he g oups o coun ies analyzed — om 71.8% in
he case o OECD Eu ope o 189.7% in No h Ame ica. Ac oss he boa d, he
e olu ion o he di e ences in inal ene gy consump ion pe GDP uni is wha
de e mines he e olu ion o ene gy in ensi y inequali ies. Tha is, he obse ed
inc eases o declines in inequali ies in inal ene gy consump ion pe GDP uni
p edomina e he e olu ion o he emaining ac o s ha de e mine he ene gy
in ensi y inequali y endency.
The ac o ha cap u es he weigh o he di e ences in he ans o ma ion
index, which ep esen s almos 8% o he explained global di e ences (see
Table 2), is much mo e ele an in some o he egions conside ed. This is he
case o Asia (19.6%), he Middle Eas (26.2%), A ica (23.5%), and e en
OECD Eu ope (20.4%). As o A ica, whe e he weigh o his componen is
g ea e , i has e en inc eased subs an ially du ing he pe iod o analysis. Wi hin
hese egions, among coun ies economically and geog aphically ela i ely
homogenous, he e a e impo an di e ences in e iciency o he ene gy sec o
in ans o ming p ima y ene gy in o inal ene gy.
The in e ac ion ac o also exhibi s a e y di e en pa e n ac oss he egions,
being e en posi i e in he case o OECD Eu ope and La in Ame ica. Howe e ,
in he o me i dec eases subs an ially h oughou he pe iod while in he la e
he beha iou is somewha e a ic, inc easing be ween 1970 and 1990 o
dec ease a e wa ds. Anyway, and con a y o he o he egions, in OECD
Eu ope and La in Ame ica he in e ac ion componen ein o ces global
17
inequali y. Tha is, inequali ies in he ans o ma ion indices and inal ene gy
consump ion pe GDP uni would ein o ce each o he . This is so because he
coun ies wi h wo se e iciency le el in ene gy ans o ma ion end also o be
hose ha consume mo e ene gy pe GDP uni . A he o he ex eme a e
egions like No h Ame ica o A ica, whe e he nega i e and signi ican alue o
he in e ac ion ac o has been inc eased h oughou he pe iod. This la gely
compensa es he inequali ies in inal ene gy consump ion in bo h cases.
4. Conclusions
As shown by p e ious s udies, ene gy in ensi y inequali y is a de e mining ac o
in he unequal ene gy consump ion and emissions pe capi a among coun ies.
The obse ed decline in ene gy in ensi y inequali y in he pas decades has
been one o he main causes o he educ ion in inequali ies in emissions pe
capi a. Se e al s udies ha e ocused on he impo ance o di e en ac o s in
he e olu ion o inal ene gy in ensi y di e ences. Howe e , p ima y ene gy
inequali ies and hei e olu ion can be condi ioned by di e ences in inal ene gy
in ensi y and di e ences in e iciency in ans o ming p ima y in o inal ene gy
alike. The p esen pape con ibu es o he li e a u e by illus a ing he ole o he
di e ences in he in e nal componen o he ene gy sec o and he one o he
di e ences in inal ene gy in ensi y in he e olu ion o p ima y ene gy in ensi y
inequali ies h ough syn he ic indica o s.
We analyzed he e olu ion o ene gy in ensi y inequali ies using he Theil index.
The me hodology, which is de eloped by he au ho s, allows decomposing
inequali y in o h ee componen s. One ha cap u es he pa ial con ibu ion o
ene gy ans o ma ion indexes; one ha calcula es he ole o inal ene gy
consump ion pe GDP uni ; and inally an in e ac ion ac o . The me hodology
also pe mi ed he decomposi ion o he ac o s by g oups o coun ies and
wi hin he g oups conside ed ( ollowing he IEA classi ica ion).
In line wi h p e ious s udies he indings clea ly e eal an impo an decline in
c oss coun y ene gy in ensi y inequali y. Fo he g oups o coun ies
18
conside ed, in e -g oup inequali y is mo e impo an han in a-g oup inequali y
(60%–40%). This shows he ele ance o he g ouping we made. Howe e ,
inequali y is no educed in all he egions. In ou o he nine egions conside ed
inequali y ac ually inc eased.
The ac o ial decomposi ion allowed o iden i y he unequal inal ene gy
consump ion pe GDP uni as he mos ele an ac o in c oss coun y ene gy
in ensi y inequali ies. The obse ed educ ion in inal ene gy in ensi y inequali y
can be a ibu ed o ei he a con e gence in inal ene gy consump ion e iciency
ac oss coun ies o a g ea e simila i y in sec o al p oduc ion s uc u es. In ha
espec , Du o e al. (2010) show ha , o a sample o OECD coun ies, he
educ ion in inal ene gy in ensi y inequali ies is undamen ally due o a
con e gence in consump ion e iciency om sec o o sec o and no o a
g ea e simila i y in p oduc ion composi ion, which ha e become mo e unequal
du ing he pe iod o analysis. Ne e heless, his e idence canno howe e be
ex apola ed o he es o he egions. In ac , he p esen pape shows ha in
some egions he e has been inc easing di e gence in inal ene gy consump ion
pe GDP uni .
The ans o ma ion ac o has been less ele an in de e mining c oss coun y
ene gy in ensi y inequali ies, al hough i s ole is a om being negligible. I s
con ibu ion o in e -g oup inequali y is e en mo e mode a e. Howe e , i is e y
ele an in explaining he exis ing di e ences wi hin some o he egions
conside ed. I ep esen s one ou h o egional inequali ies in ene gy in ensi y in
some egions. Impo an di e ences s ill exis in he e iciency in ans o ming
p ima y ene gy in o inal one. Clea ly, he di e ences in ene gy con e sion
e iciency canno be neglec ed. In some egions (e.g. A ica) such di e ences
go up while he o e all end is downwa d. Howe e , h ough he
implemen a ion o pe inen measu es, he e could s ill be scope o each
g ea e con e gence owa d highe e iciency le el in ene gy ans o ma ion.
The in e ac ion componen is qui e ele an , pa icula ly wi h espec o wi hin
g oup ene gy in ensi y inequali ies o some coun ies. Howe e , in e ms o
global inequali y i s con ibu ion is mode a e and nega i e. This nega i e
19

co ela ion sugges s ha inequali ies in he wo ac o s conside ed end o
compensa e one ano he . Ne e heless, his beha iou changes depending on
whe he we conside in e -g oup o in a-g oup inequali y.
Acknowledgemen s
The au ho s acknowledge suppo om p ojec s ECO2009-10003 (Minis e io de
Ciencia e Inno ación), 2009SGR-600 and XREPP (DGR).
Re e ences
Alcán a a, V., Du o, J.A. (2004) “Inequali y o ene gy in ensi ies ac oss OECD
coun ies”, Ene gy Policy, 32, 1257–1260.
Alcán a a, V., Padilla, E. (2005) “Análisis de las emisiones de CO2 y sus
ac o es explica i os en las di e en es á eas del mundo”, Re is a de
Economía C í ica, 4, 17–37.
Ang, B.W. (1999) “Is he ene gy in ensi y a less use ul indica o han he ca bon
ac o in he s udy o clima e change?”, Ene gy Policy, 27, 943–946.
Can o e, N., Padilla, E. (2010a) "Equali y and CO2 emissions dis ibu ion in
clima e change in eg a ed assessmen modelling", Ene gy, 35 (1), 298–
313.
Can o e, N., Padilla, E. (2010b) "Emissions dis ibu ion in pos – Kyo o
in e na ional nego ia ions: a policy pe spec i e" en Ekekwe, N. (Ed.)
Nano echnology and Mic oelec onics: Global Di usion, Economics and
Policy. IGI Global, He shey, USA.
Cowell, F. (1995) Measu ing Inequali y. Ha es e Whea sheal, Hemel
Hemps ead.
Du o, J.A. (2003) "Fac o decomposi ion o spa ial income inequali y: a
e ision", Wo king Pape wpdea0302, Depa men o Applied Economics,
Uni . Au ónoma de Ba celona.
Du o, J.A., Padilla, E. (2006) “In e na ional inequali ies in pe capi a CO2
emissions: a decomposi ion me hodology by Kaya ac o s”, Ene gy
Economics, 28, 170–187.
20
Du o, J.A. y Padilla, E. (2008) “Analysis o he in e na ional dis ibu ion o pe
capi a CO2 emissions using he pola iza ion concep ", Ene gy Policy, 36,
456–466.
Du o, J.A., Padilla, E., Alcán a a, V. (2010) "In e na ional inequali y in ene gy
in ensi y le els and he ole o p oduc ion composi ion and ene gy
e iciency: An analysis o OECD coun ies”, Ecological Economics,
doi:10.1016/j.ecolecon.2010.07.022
Es eban, J., G adín, C., Ray, D. (1999) “Ex ensions o a Measu e o
Pola iza ion, wi h an Applica ion o he Income Dis ibu ion o i e OECD
Coun ies.” Luxembou g Income S udy Wo king Pape Se ies 218.
Sy acuse, New Yo k: Maxwell School o Ci izenship and Public A ai s,
Sy acuse Uni e si y.
G oo , L. (2010) “Ca bon Lo enz cu es”, Resou ce and Ene gy Economics, 32
(1), 45–64.
Hamil on, C., Tu on, H. (2002) “De e minan s o emissions g ow h in OECD
coun ies”, Ene gy Policy, 30, 63–71.
Hedenus, F., Aza , C. (2005) “Es ima es o ends in global income and
esou ce inequali ies”, Ecological Economics, 55, 351–364.
In e na ional Ene gy Agency (IEA) (2009a) Ene gy Balances o OECD
coun ies. AIE, OECD, Pa is.
In e na ional Ene gy Agency (IEA) (2009b) Ene gy Balances o non-OECD
coun ies. AIE, OECD, Pa is.
Kaya, Y. (1989) “Impac o Ca bon Dioxide Emission Con ol on GNP G ow h:
In e p e a ion o P oposed Scena ios," pape p esen ed o he Ene gy and
Indus y Subg oup, Response S a egies Wo king G oup,
In e go e nmen al Panel on Clima e Change, Pa is, F ance.
Millime , D.L., Slo je, D. (2002) “An en i onmen al Paglin-Gini”, Applied
Economics Le e s, 9 (4), 271–274
Padilla, E., Du o, J.A. (2009) "Analysis o he in e na ional inequali y in CO2
emissions: The con ibu ion o income inequali y and o he ac o s", en
I ing H. Wadell (Ed.) Income Dis ibu ion: Inequali ies, Impac s and
Incen i es. No a Science Publishe s, Nue a Yo k, pp. 171–183.
21
Padilla, E., Se ano, A. (2006) "Inequali y in CO2 emissions ac oss coun ies
and i s ela ionship wi h income inequali y: a dis ibu i e app oach", Ene gy
Policy, 34, 1762–1772.
Sho ocks, A. (1980) “The class o addi i ely decomposable inequali y
measu es”, Econome ica, 48, 613-625.
Sun, J.W. (2002) “The dec ease in he di e ence o ene gy in ensi ies be ween
OECD coun ies om 1971 o 1998”, Ene gy Policy, 30, 631–635.
Theil, H. (1967) Economics and In o ma ion Theo y, Ams e dam: No h Holland.
Yamaji, K., Ma suhashi, R., Naga a, Y., Kaya, Y. (1991) “An In eg a ed Sys ems
o CO2/Ene gy/GNP Analysis: Case S udies on Economic Measu es o
CO2 Reduc ion in Japan. Wo kshop on CO2 Reduc ion and Remo al:
Measu es o he Nex Cen u y”, 19–21 Ma ch 1991. In e na ional Ins i u e
o Applied Sys ems Analysis, Laxenbu g, Aus ia.
22
Annex 1. Coun ies included in he sample:
OECD Eu ope: Aus ia, Belgium, Czech Republic, Denma k, Finland, F ance,
Ge many, G eece, Hunga y, Island, I eland, I aly, Luxembou g, Ne he lands,
No way, Poland, Po ugal, Slo ak Republic, Spain, Sweden, Swi ze land,
Tu key, Uni ed Kingdom.
No h Ame ica: Canada, Mexico, Uni ed S a es.
OECD Paci ic: Aus alia, Japan, Ko ea, New Zealand.
Non-OECD Eu ope: Albania, Bulga ia, Cyp us, Gib al a , Mal a, Rumania,
Fo me USSR, Fo me Yugosla ia
A ica: Alge ia, Angola, Benin, Came oon, Congo, Democ a ic Republic o
Congo, Cô e d'I oi e, Egyp , E hiopia, Gabon, Ghana, Kenya, Libya, Mo occo,
Mozambique, Nige ia, Senegal, Sou h A ica, Sudan, Tanzania, Togo, Tunisia,
Zambia, Zimbabwe, O he s A ica.
La in Ame ica: A gen ina, Boli ia, B azil, Chile, Colombia, Cos a Rica, Cuba,
Dominican Republic, Ecuado , El Sal ado , Gua emala, Hai i, Hondu as,
Jamaica, Nica agua, Panama, Pa aguay, Pe u, T inidad and Tobago, U uguay,
Venezuela, O he s La in Ame ica.
Middle Eas : Bah ain, Islamic Republic o I an, I aq, Is ael, Jo dan, Kuwai ,
Lebanon, Oman, Qa a , Saudi A abia, Sy ia, Uni ed A ab Emi a es, Yemen.
Asia: Bangladesh, B unei Da ussalam, Chinese Taipei, India, Indonesia, Dem.
People's Rep. o Ko ea, Malaysia, Myanma , Nepal, Pakis an, Philippines,
Singapo e, S i Lanka, Thailand, Vie nam, O he s Asia.
China: People's Republic o China, Hong Kong.
23