Fiedle , Nicolas
A icle
Analyzing he e ail gasoline ma ke in Ge many: Impac o
spa ial compe i ion and ma ke concen a ion on p ices
Junio Managemen Science (JUMS)
P o ided in Coope a ion wi h:
Junio Managemen Science e. V.
Sugges ed Ci a ion: Fiedle , Nicolas (2025) : Analyzing he e ail gasoline ma ke in Ge many: Impac
o spa ial compe i ion and ma ke concen a ion on p ices, Junio Managemen Science (JUMS),
ISSN 2942-1861, Junio Managemen Science e. V., Planegg, Vol. 10, Iss. 2, pp. 424-440,
h ps://doi.o g/10.5282/jums/ 10i2pp424-440
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/320454
S anda d-Nu zungsbedingungen:
Die Dokumen e au EconS o dü en zu eigenen wissenscha lichen
Zwecken und zum P i a geb auch gespeiche und kopie we den.
Sie dü en die Dokumen e nich ü ö en liche ode komme zielle
Zwecke e iel äl igen, ö en lich auss ellen, ö en lich zugänglich
machen, e eiben ode ande wei ig nu zen.
So e n die Ve asse die Dokumen e un e Open-Con en -Lizenzen
(insbesonde e CC-Lizenzen) zu Ve ügung ges ell haben soll en,
gel en abweichend on diesen Nu zungsbedingungen die in de do
genann en Lizenz gewäh en Nu zungs ech e.
Te ms o use:
Documen s in EconS o may be sa ed and copied o you pe sonal
and schola ly pu poses.
You a e no o copy documen s o public o comme cial pu poses, o
exhibi he documen s publicly, o make hem publicly a ailable on he
in e ne , o o dis ibu e o o he wise use he documen s in public.
I he documen s ha e been made a ailable unde an Open Con en
Licence (especially C ea i e Commons Licences), you may exe cise
u he usage igh s as speci ied in he indica ed licence.
h ps://c ea i ecommons.o g/licenses/by/4.0/
Junio Managemen Science 10(2) (2025) 424-440
Junio Managemen Science
www.jums.academy
ISSN: 2942-1861
Edi o :
DOMINIK VAN AAKEN
Ad iso y Edi o ial Boa d:
FREDERIK AHLEMANN
JAN-PHILIPP AHRENS
THOMAS BAHLINGER
MARKUS BECKMANN
SULEIKA BORT
ROLF BRÜHL
KATRIN BURMEISTER-LAMP
CATHERINE CLEOPHAS
NILS CRASSELT
BENEDIKT DOWNAR
KERSTIN FEHRE
MATTHIAS FINK
DAVID FLORYSIAK
GUNTHER FRIEDL
MARTIN FRIESL
FRANZ FUERST
WOLFGANG GÜTTEL
NINA KATRIN HANSEN
ANNE KATARINA HEIDER
CHRISTIAN HOFMANN
SVEN HÖRNER
STEPHAN KAISER
NADINE KAMMERLANDER
ALFRED KIESER
ALEKSANDRA KLEIN
NATALIA KLIEWER
DODO ZU KNYPHAUSEN-AUFSESS
SABINE T. KÖSZEGI
ARJAN KOZICA
CHRISTIAN KOZIOL
MARTIN KREEB
WERNER KUNZ
HANS-ULRICH KÜPPER
MICHAEL MEYER
JÜRGEN MÜHLBACHER
GORDON MÜLLER-SEITZ
J. PETER MURMANN
ANDREAS OSTERMAIER
BURKHARD PEDELL
ARTHUR POSCH
MARCEL PROKOPCZUK
TANJA RABL
SASCHA RAITHEL
NICOLE RATZINGER-SAKEL
ASTRID REICHEL
KATJA ROST
THOMAS RUSSACK
FLORIAN SAHLING
MARKO SARSTEDT
ANDREAS G. SCHERER
STEFAN SCHMID
UTE SCHMIEL
CHRISTIAN SCHMITZ
MARTIN SCHNEIDER
MARKUS SCHOLZ
LARS SCHWEIZER
DAVID SEIDL
THORSTEN SELLHORN
STEFAN SEURING
VIOLETTA SPLITTER
ANDREAS SUCHANEK
TILL TALAULICAR
ANN TANK
ANJA TUSCHKE
MATTHIAS UHL
CHRISTINE VALLASTER
PATRICK VELTE
CHRISTIAN VÖGTLIN
BARBARA E. WEISSENBERGER
ISABELL M. WELPE
HANNES WINNER
THOMAS WRONA
THOMAS ZWICK
Volume 10, Issue 2, June 2025
JUNIOR
MANAGEMENT
SCIENCE
Yannick Hohens ein,The Impac o P o i abili y on Scope 1, 2
and 3 GHG Emissions in Eu ope
Gün he Gampe , Small bu Powe ul: The Impac o Shel
Talke Flags on Consume Shopping Beha io
Sebas ian Lüpni z,Un a elling Collec i e Ac ion F ames
Th ough a Tempo al Lens: A Case S udy o an
En i onmen al Mo emen in Ge many
Mohammad Izza Raihan Im on, Sus ainabili y in he
Co po a e Sec o : A News Tex ual Analysis App oach
o Measu ing ESG Pe o mance
Eli Leman Bilgin, Unde s anding Eme gen Leade ship Ac oss
Cul u al Le els: A Theo e ical F amewo k
Nicolas Fiedle , Analyzing he Re ail Gasoline Ma ke in
Ge many: Impac o Spa ial Compe i ion and Ma ke
Concen a ion on P ices
Sebes yén And ás Huszá , The Role o Hie a chical
Di e en ia ion o he E ec i eness o Socce Teams
Hashma ullah Sadid, Wai ing Time Es ima ion o Ride-Hailing
Flee s Using G aph Neu al Ne wo ks
Licia Recke sd ees, Gone, Space Gone -Non-Te i o ial
Wo kplace Models in he Con ex o Hyb id Wo king
F om he Employees' Pe spec i e
Timo And eas Delle , Explo ing Disc epancies in Ene gy
Pe o mance Ce i ica es: Analyzing Ene gy E iciency
P emiums o Buildings Based on Theo e ical Ene gy
Requi emen s Ve sus Ac ual Ene gy Consump ion
292
334
349
369
402
424
441
462
491
522
Published by Junio Managemen Science e.V.
This is an Open Access a icle dis ibu ed unde he e ms o he CC-BY-4.0
(A ibu ion 4.0 In e na ional). Open Access unding p o ided by ZBW.
ISSN: 2942-1861
Analyzing he Re ail Gasoline Ma ke in Ge many:
Impac o Spa ial Compe i ion and Ma ke Concen a ion on P ices
Nicolas Fiedle
Ludwig Maximilian Uni e si y o Munich
Abs ac
Gi en he changing landscape o uel e ailing, his s udy explo es he impac o spa ial compe i ion and ma ke concen a ion
on diesel p ices in Ge many. The ques ion o how popula ion densi y and gas s a ion densi y, i.e. he equilib ium pa e n o
loca ions o i ms, a e ela ed is examined. In addi ion, he impac o gas s a ion densi y, as a spa ial measu e o compe i ion,
and ma ke concen a ion on diesel p ices is in es iga ed. Based on heo y, popula ion densi y should ha e a posi i e impac
on gas s a ion densi y. Gas s a ion densi y should ha e a nega i e and ma ke concen a ion a posi i e in luence on he diesel
p ice in a dis ic . Using 2022 da a on Ge man gas s a ions and diesel p ices, a posi i e e ec o popula ion densi y and on gas
s a ion densi y, a nega i e e ec o gas s a ion densi y on diesel p ice, and a posi i e e ec o ma ke concen a ion on diesel
p ice we e each ound a he dis ic le el. The e ec s o gas s a ion densi y and ma ke concen a ion, howe e , we e ela i ely
small. The esul s show ha uel p ices a gas s a ions a e in luenced by spa ial compe i ion and ma ke concen a ion.
Keywo ds: diesel p ices; gas s a ion densi y; ma ke concen a ion; e ail gasoline ma ke ; spa ial compe i ion
1. In oduc ion
In 2022, uel p ices in Ge many ha e come e y much
in o ocus. Fo he eason ha i was he mos expensi e
yea e e o d i e s in Ge many o ill up hei uel anks. A
li e o diesel was p iced a an a e age o 1.95 eu os, and a
li e o Supe E10 a 1.86 eu os o e he yea 2022. Gasoline
was hus a ound 27 cen s mo e expensi e han in he p e i-
ous eco d yea o 2012, and he p ice pe li e o diesel was
e en almos 47 cen s highe han in 2012 (P ack, 2023). The
mobili y landscape is o g ea impo ance o socie y and econ-
omy. Mobili y is no only a ool o ul ill indi idual needs, bu
also an impo an d i e o economic ac i i ies and ade. In
his con ex , uel p ices play a c i ical ole, in luencing o a
la ge ex en indi idual mobili y cos s and o e all cos s o
he anspo a ion sec o . I is impo an o all popula ion
g oups and economic sec o s o pu chase uels e icien ly and
a easonable p ices. Fo his eason, i makes sense o exam-
ine in de ail he ac o s ha in luence hese p ices and he
unde lying ma ke s uc u e.
Fuels can be conside ed a ai ly homogeneous p oduc
due o hei chemical composi ion. Ne e heless, gasoline
p ices a indi idual gas s a ions a y widely. Many pape s
explain he p ice di e ences by he in ensi y o local compe-
i ion. This compe i ion is de e mined in he e ail gasoline
ma ke no only by he numbe o compe i o s, bu also by
he geog aphic dis ibu ion o gasoline s a ions in he a ea.
Consume s end o buy hei uel a gas s a ions close o hei
homes, as g ea e dis ance o gas s a ions esul s in highe
anspo a ion cos s. I ollows ha compe i ion in he e ail
gasoline ma ke is highly localized (Be gan ino e al., 2020).
The objec i e o his pape is o in es iga e he impac o
spa ial compe i ion and ma ke concen a ion on p ices by
looking a he e ail gasoline ma ke in Ge many. The i s
ques ion ha a ises is wha a ec s he equilib ium pa e n o
company loca ions? And he second ques ion is, wha a e he
cha ac e is ics o equilib ium p ices i he e is spa ial compe-
i ion among i ms? Clemenz and Gugle (2006) ha e also
al eady s udied hese wo ques ions in hei wo k. I has al-
eady been obse ed ha popula ion densi y has a majo im-
pac on he densi y o gas s a ions (Clemenz & Gugle , 2006;
Gö z & Gugle , 2006). Based on his and he model o Salop
DOI: h ps://doi.o g/10.5282/jums/ 10i2pp424-440
© The Au ho (s) 2025. Published by Junio Managemen Science.
This is an Open Access a icle dis ibu ed unde he e ms o he CC-BY-4.0
(A ibu ion 4.0 In e na ional). Open Access unding p o ided by ZBW.
N. Fiedle /Junio Managemen Science 10(2) (2025) 424-440 425
(1979), he i s hypo hesis was de eloped ha e ail s o es
a e mo e likely o be loca ed mo e densely in a eas wi h
highe popula ion densi y. Se e al esea che s such as Ba -
on e al. (2004), Ca doso e al. (2020), Clemenz and Gugle
(2006), and Mee beeck (2003) ha e looked a he ela ion-
ship be ween spa ial compe i ion and p ices a gas s a ions.
Mos o hem ound ha highe spa ial compe i ion leads o
lowe p ices. An impo an ac o is he dis ance be ween
gas s a ions. This leads o he second hypo hesis ha , unde
spa ial compe i ion, p ices end o be lowe as he densi y o
selle loca ions inc eases. The ela ionship be ween ma ke
concen a ion and p ices has also been s udied a ew imes,
o example by Ecke and Wes (2005), Kihm e al. (2016),
and Sen (2003). Thei indings indica e ha highe p ices
a e ela ed o highe ma ke concen a ion. This also o ms
he hi d and las hypo hesis o his pape , ha p ices end o
be highe , as ma ke concen a ion inc eases.
To es hese hypo heses, a da ase o Ge man gas s a ions
and hei diesel p ices om he yea 2022 was used. The po-
li ical dis ic s in Ge many we e de ined as local ma ke s, ol-
lowing Clemenz and Gugle (2006). Based on his, a e age
alues such as diesel p ice, gas s a ion densi y, one and ou
i m concen a ion a io in 2022 we e calcula ed o each dis-
ic . All hypo heses we e es ed using mul iple OLS eg es-
sions wi h di e en combina ions o con ol a iables. In he
i s s ep, a signi ican posi i e e ec o popula ion densi y on
gas s a ion densi y was ound. In he second s ep, a signi i-
can nega i e e ec o gas s a ion densi y on diesel p ice was
ound. A obus ness es was pe o med by emo ing 20%
o he dis ic s wi h he highes popula ion densi y om he
da ase , in o de o exclude mainly me opoli an a eas om
he analysis. The e ec was s ill nega i e bu no longe sig-
ni ican . Las ly, a signi ican posi i e e ec o ma ke concen-
a ion on diesel p ice was ound. This was also he case in
he obus ness es . Bu i should be no ed ha he e ec s o
gas s a ion densi y and ma ke concen a ion a e e y small.
This pape has he ollowing ou line. Sec ion 2 p esen s
he heo e ical backg ound by means o ele an li e a u e
and desc ip ion o impo an concep s and leads a gumen a-
i ely o he hypo heses. Sec ion 3 p o ides an o e iew o
he Ge man e ail gasoline ma ke . Sec ion 4 desc ibes he
me hodology used and sec ion 5 p esen s he empi ical e-
sul s. Sec ion 6 concludes he s udy wi h a discussion and
compa ison o he esul s wi h p e ious li e a u e, limi a-
ions, heo e ical and manage ial implica ions.
2. Theo y and Hypo heses
2.1. Gas S a ion Densi y
In his pa ag aph, ollowing Clemenz and Gugle (2006),
he aspec s o he ci cula ci y model o Salop (1979) ha
a e ele an o his esea ch a e desc ibed. An impo an ea-
u e o spa ial compe i ion is ha consume s shop a he s o e
whe e hei o al cos s a e lowes . These consis o he p ice
(mul iplied by he quan i y) and any anspo cos s a cus-
ome mus ace. This esul s in each s o e ha ing a “local
monopoly”, he geog aphical ex en o which depends on he
p ices o he nea es compe i o s and he espec i e anspo
cos s o each cus ome o he a ious s o es. These ans-
po a ion cos s a e gene ally subs an ially in luenced by he
dis ance be ween s o es, and o some ex en by oad quali y
and he a ailabili y o public anspo a ion, e c. The p ice
a s o e can cha ge, inc eases wi h he dis ance o he nea -
es compe i o as well as wi h he anspo cos s o a cus-
ome . Demand in such a local monopoly depends no only
on he geog aphic size o he ma ke , bu also on he numbe
o cus ome s in he e i o y. In o he wo ds, he popula ion
densi y. A company wan s o be a a loca ion wi h many po-
en ial cus ome s bu ew compe i o s, as he high demand is
dis ibu ed among ela i ely ew companies. The po en ially
high demand is dis ibu ed among ela i ely ew companies,
hus he expec ed p o i is highe . In mo e densely popu-
la ed a eas, he gap be ween compe i o s can be smalle , as
demand pe squa e kilome e is usually much highe .
This ela ionship has al eady been obse ed in he e ail
gasoline ma ke . Gö z and Gugle (2006) looked a he ela-
ionship be ween popula ion densi y and gas s a ion densi y
in Aus ia. Using an O dina y Leas Squa es (OLS) eg es-
sion, hey ound ha i he popula ion densi y in a ma ke
inc eases by one pe cen age poin , hen he densi y o gas
s a ions inc eases on a e age by 0.835 pe cen . Clemenz
and Gugle (2006) ha e de ined he 121 poli ical dis ic s
in Aus ia as local ma ke s. Looking a he same ela ion-
ship, hey an mul iple eg essions, each wi h di e en con-
ol a iables, such as di e en a iables o ma ke concen-
a ion o numbe o mo o ehicles pe capi a. Thei esul s
showed ha in each o hei models, he coe icien o pop-
ula ion densi y is posi i e and signi ican a he 5% le el o
lowe . Thei coe icien s ange om 0.810 o 0.873, depend-
ing on he con ol a iable. This means ha i popula ion
s a ion densi y inc eases by one pe cen age poin , he gas s a-
ion densi y inc eases by a leas 0.8% on a e age. They also
pe o med a obus ness es by changing he ma ke de ini-
ion and de ining each municipali y as a local ma ke . Thei
esul s ha e shown o be obus o he change in ma ke de -
ini ion.
This sugges s ha , on a e age, he highe he popula ion
densi y in a local ma ke , he highe he gas s a ion densi y.
Based on his li e a u e and he p e ious explana ions, he
ollowing i s hypo hesis was de eloped:
Hypo hesis 1: Re ail s o es a e mo e likely o be
loca ed mo e densely in a eas wi h highe popu-
la ion densi y.
2.2. Diesel P ice and Gas S a ion Densi y
Highe gas s a ion densi y, in u n, leads o inc eased spa-
ial compe i ion. This aises he ques ion o he ex en o
which his compe i ion in luences he p ices a gas s a ions.
F om a heo e ical s andpoin , p icing di e ences can be ex-
plained by p oduc di e en ia ion.
In o de o explain he concep o p oduc di e en ia ion,
his pa ag aph e e s o he book “The Theo y o Indus ial
N. Fiedle /Junio Managemen Science 10(2) (2025) 424-440426
O ganiza ion” by Ti ole (1988). Two p oduc s a e almos
ne e pe ec subs i u es, in he sense ha all cus ome s a e
indi e en i he wo goods ha e he same p ice. P oduc s
a e almos always di e en ia ed by a ce ain a ibu e. A
good can be cha ac e ized as a bundle o di e en a ibu es:
quali y, a ailabili y, loca ion, he in o ma ion o he consume
abou i s a ailabili y and quali y, e c. Each consume has a
anking o he di e en cha ac e is ics. Resea che s usually
ocus on a small pa o hese cha ac e is ics and a special de-
sc ip ion o he espec i e cus ome p e e ences. The e a e
wo cases ha a e used commonly. The i s case is e ical
di e en ia ion. In a e ically di e en ia ed p oduc en i on-
men , all consume s ag ee on he p e e ed mix o ea u es.
In mo e gene al e ms, hey also ag ee on he o de o p e e -
ence. The mos ypical example o his is quali y. The majo -
i y ag ee ha a highe quali y is o be a o ed. Gi en equal
p ices, he e is a clea na u al o de ing o e he cha ac e is-
ics space. Fo example, gi en he same p ice, mos people
p e e a powe ul lap op o e a less powe ul lap op, o an
Audi o e a Toyo a. The second case is ho izon al di e en-
ia ion. Fo some a ibu es, gi en equal p ices, he op imal
decision depends on he indi idual consume . P e e ences
and as es a e e y di e en . A common example is colo ,
one pe son p e e s blue, he o he one ed. Ano he impo -
an poin in his con ex is he loca ion. To exagge a e, i is
likely ha Munich esiden s p e e p oduc s ha a e a ailable
in Munich o physically iden ical p oduc s ha a e only a ail-
able in New Yo k. B oken down, cus ome s p e e s o es ha
a e close o hem. One model o ho izon al di e en ia ion is
Salop’s ci cle model, which was desc ibed a he beginning
o he chap e .
In his wo k, Mee beeck (2003) applied he concep o
p oduc di e en ia ion o he gasoline ma ke . He s a es ha
in he case o gas s a ions, bo h ho izon al and e ical di -
e en ia ion can be obse ed. Al hough gasoline can heo e -
ically be conside ed a homogeneous commodi y in e ms o
i s physical and chemical cha ac e is ics, in p ac ice p oduc
speci ica ions a y due o he addi i es used by each b and.
Al hough hese ac o s could play a ole, he ound ha lo-
ca ion has he g ea es impac on p ice di e ences be ween
gas s a ions. Consume s gene ally p e e gas s a ions in hei
“neighbo hood.” Lee (2007) examined he na u e o compe-
i ion in San Diego’s e ail gasoline ma ke using wo yea s o
panel da a on weekly gas s a ion p ices. He was able o ind
ha e ail p ices a e s ongly in luenced by gas s a ion cha -
ac e is ics such as b and name and ameni ies. Bu he also
ound ha he ela i e geog aphic p oximi y o compe ing
gas s a ions is an impo an ac o in explaining p ice di e -
ences among gas s a ions. Gas s a ions compe e mos wi h
gas s a ions ha a e less han 1 mile away. The in ensi y o
compe i ion con inues o dec ease wi h dis ance.
F om hese pape s, i can be in e ed ha spa ial compe-
i ion is an impo an ac o in he o ma ion o p ices in he
gasoline ma ke . Se e al esea che s ha e al eady s udied
how his spa ial compe i ion a ec s equilib ium p ices a gas
s a ions. O en, local gas s a ion densi y is de ined using a
ixed and a bi a y adius. Hosken e al. (2008) de e mined
he numbe o gas s a ions in a 1.5-mile adius and he dis-
ance o he nea es gas s a ion as he compe i i e measu e.
They looked a how his local compe i ion a ec s p ices a
indi idual gas s a ions. Howe e , no consis en ela ionship
was ound be ween local compe i ion and he p ices o ma -
gins o he gas s a ions. Ba on e al. (2004) also analyzed
he e ec o he numbe o gas s a ions in a 1.5-mile adius
on p ice. They obse ed ha a 50% inc ease in he numbe
o gas s a ions in his adius leads o a dec ease in he p ice
o abou 0.5%.
O he s udies ha e de ined poli ical a eas as local ma -
ke s. Mee beeck (2003) has de ined he gas s a ions densi y
in a municipali y as he local compe i ion measu e o he Bel-
gian e ail gasoline ma ke . A highe numbe o gas s a ions
leads o mo e in ense compe i ion. The esul s show ha he
numbe o local compe i o s has a nega i e e ec on p ices.
This indica es ha as he numbe o compe ing s a ions in a
local ma ke inc eases, diesel p ices dec ease. Howe e , he
magni ude o his e ec is e y small, bu signi ican .
Clemenz and Gugle (2006) ha e de ined he 121 poli -
ical dis ic s in Aus ia as local ma ke s. They analyzed he
in luence o gas s a ion densi y on he ma gin a gas s a ions.
Se e al equa ions we e analyzed wi h di e en con ol a i-
ables, such as ma ke concen a ion o ALPS. ALPS desc ibes
he p opo ion o moun ains and o es in each coun y. This
is in ended o se e as an addi ional p oxy o he di e ences
in anspo a ion cos s be ween he espec i e dis ic s. Thei
esul s showed ha in each o hei models, he coe icien o
gas s a ion densi y is nega i e and signi ican a he 5% le el
o lowe . Thei coe icien s ange om -0.035 o -0.045. This
means ha i gas s a ion densi y inc eases by one pe cen age
poin , he ma gin dec eases by a ound 0.04% on a e age.
They also pe o med a obus ness es by changing he ma -
ke de ini ion and de ining each municipali y as a local ma -
ke . Thei esul s ha e shown o be obus o he change in
ma ke de ini ion. This sugges s ha he close compe i o s
a e o each o he on a e age, he lowe he ma gin.
Ca doso e al. (2020) ha e measu ed he e ec on p ices
in he B azilian gasoline ma ke when a new company en e s
a local ma ke . As a esul o he inc eased spa ial compe-
i ion, p ices dec ease. The close he loca ion o he new
compe i o is o exis ing gas s a ions, he mo e p ices a e e-
duced.
In summa y, mos pape s ha e shown ha p ices a gas
s a ions a e a ec ed by spa ial compe i ion. I u ns ou ha
he dis ance be ween gas s a ions is a key ac o . Gas s a ion
densi y e lec s his a e age dis ance as an in e se p oxy and
hus can be used as a measu e o spa ial compe i ion. Based
on his and he model o Salop (1979), he second hypo hesis
was de eloped:
Hypo hesis 2 - Unde spa ial compe i ion, p ices
end o be lowe as he densi y o selle loca ions
inc eases.
As Clemenz and Gugle (2006) ha e al eady poin ed ou , hy-
po hesis 2 has an ob ious consequence o hypo hesis 1: In
N. Fiedle /Junio Managemen Science 10(2) (2025) 424-440 427
he p esence o spa ial compe i ion, he inc ease in s o e den-
si y has o be less han p opo ional o he inc ease in popu-
la ion densi y, since a highe s a ion densi y lowe s he equi-
lib ium p ice.
2.3. Diesel P ice and Ma ke Concen a ion
So a , only gas s a ions ha e been conside ed indi idu-
ally and independen ly o he ope a ing company. In eali y,
howe e , no e e y company ope a es jus one gas s a ion in
each ma ke , bu in a numbe o cases se e al. Fo his ea-
son, some companies could ha e mo e powe han o he s in
ce ain ma ke s. This in u n could ha e an impac on p ices.
In o de o in oduce he subjec o he ma ke powe
and oligopolies, his pa ag aph e e s o he book “Indus ial
O ganiza ion: Ma ke s and S a egies” by Belle lamme and
Pei z (2015). Companies a e expec ed o maximize p o i s,
bu he ma ke condi ions limi hei abili y o exploi con-
sume s. I he pa adigm o pe ec compe i ion is o be be-
lie ed, i ms will in he end sell a a p ice ha equals hei
ma ginal cos . Fi ms do no ha e ma ke powe i com-
pa ed o he indus y hey a e small and p ice ake s. Bu
wha happens when some companies ha e a la ge ma ke
sha e in he espec i e indus y o local ma ke ? They can-
no be desc ibed as p ice ake s. Bu nei he can hey be
desc ibed as pu e p ice make s, compa ed o monopolis s,
as hey s ill compe e wi h o he companies. Al hough hese
la ge companies unques ionably exe cise ma ke powe , hei
smalle compe i o s do as well. Ma ke powe is he abili y
o se p ices abo e he compe i i e le el and hus gene a e
p o i . Indus ies in which a ew companies compe e wi h
each o he and hus ma ke powe is collec i ely sha ed a e
called oligopolies. Mos indus ies a e oligopolies, and so
is he e ail gasoline ma ke . The cha ac e is ic ea u e o
oligopolis ic compe i ion is ha companies canno be igno -
ing he beha io o hei i als. Companies’ p o i s in he end
depend on he combina ion o decisions aken by all compa-
nies in he ma ke . In making decisions, i ms mus ake in o
accoun he likely beha io o hei compe i o s and espond
o hei own decisions. The analysis o his s a egic in e -
ac ion goes back as a as he nine een h cen u y. Augus in
Cou no and Joseph Be and a e conside ed he ounding
a he s o oligopoly heo y. They each de eloped hei own
models o analyzing s a egic in e ac ion in such ma ke s. I
is beyond he scope o his pape o discuss hem.
Acco ding o Co e ill (1986), he e a e se e al oligopoly
models ha p edic ha he p ice le el in a ma ke is posi-
i ely associa ed wi h one o mo e measu es o selle concen-
a ion, such as he ou - i m a io o one i m a io.
Se e al esea che s ha e al eady s udied how ma ke
concen a ion a ec s equilib ium p ices. Co e ill (1986)
looked a he ela ionship be ween ma ke concen a ion
and p ices using supe ma ke s in local ma ke s in Ve mon .
He ound ha p ices a e signi ican ly highe in mo e concen-
a ed ma ke s. Keele e al. (1999) looked a his e ec in
o -p o i and non-p o i hospi als. Rega ding ma ke de i-
ni ion, hey ha e assumed ha he ma ke o each hospi al
co esponds o he bounda ies o he dis ic in which he
hospi al is loca ed. They we e able o ind ha hospi al
p ices we e highe in ma ke s wi h highe concen a ion.
This was e en he case o non-p o i hospi als. Asplund and
Sandin (1999) used 486 d i ing schools and hei p ices o
analyze compe i ion in 235 local ma ke s in Sweden. The
esul s showed ha p ices a e inc easing in i m concen a-
ion wi hin a ma ke . Newma k (2004) also p esen s se e al
s udies in his wo k ha also suppo his ela ionship. He
lis s, o example, he pape s o Conno (1990), Co e ill
(1990), and Kolle and Weiss (1989) and se e al mo e.
This ela ionship has also been s udied by se e al e-
sea che s in he e ail gasoline ma ke . Sen (2003) in es i-
ga ed his e ec o he Canadian e ail gasoline ma ke . To
do his, he analyzed mon hly a e ages in 10 majo Canadian
ci ies o e a se en-yea pe iod. He ound ha in addi ion
o highe a e age mon hly wholesale p ices, inc easing local
ma ke concen a ion is also posi i ely and signi ican ly asso-
cia ed wi h highe e ail p ices. Ecke and Wes (2005) s ud-
ied gas s a ions in Vancou e and also ound ha highe ma -
ke concen a ion leads o highe p ices. Kihm e al. (2016)
ound in he Ge man gasoline ma ke ha highe ma ke
concen a ion, in he o m o he He indahl-Hi schman in-
dex, leads o signi ican ly highe p ices wi hin a 5 kilome e
adius o he gas s a ion. Clemenz and Gugle (2006) an-
alyzed he impac o ma ke concen a ion on p ices a he
dis ic and municipali y le el in hei pape . No signi ican
ela ionship was ound a he dis ic le el, bu a he mu-
nicipali y le el. Be gan ino e al. (2018) ha e looked a he
municipali ies in Rome. They disco e ed ha he highe he
ma ke concen a ion in a gi en municipali y, he highe he
p ice o bo h gasoline and diesel a gas s a ions. Hosken e
al. (2008), on he o he hand, did no iden i y a signi ican
ela ionship. A la ge pa o he li e a u e, bo h in he e ail
gasoline ma ke and in o he indus ies, desc ibes a posi-
i e ela ionship be ween ma ke concen a ion and p ices.
Based on his, he hi d hypo hesis was de eloped:
Hypo hesis 3 - P ices end o be highe , as ma ke
concen a ion inc eases.
In gene al, i can be said ha hese mechanisms ha e
been li le s udied in his o m o he Ge man e ail gasoline
ma ke . In addi ion, many esea che s ha e only looked a
selec ed local ma ke s in he espec i e coun ies, and e y
a ely ha e he en i e coun y o all gas s a ions been in-
cluded in he analysis. My wo k aims o ill hese gaps in
he li e a u e.
3. The Ge man Re ail Gasoline Ma ke
Oil is s ill he mos impo an sou ce o ene gy in Ge -
many, accoun ing o 33% o o al ene gy supply in 2018.
Ge man oil demand has allen much mo e slowly han do-
mes ic oil p oduc ion o e he pas decade, so Ge many e-
mains hea ily dependen on oil impo s. Ge many has a high
dependency on oil impo s o abou 97%. Oil impo de-
pendency is calcula ed as domes ic oil p oduc ion di ided
N. Fiedle /Junio Managemen Science 10(2) (2025) 424-440428
by o al oil demand. Ge many’s op c ude oil supplie s in
2018 we e Russia (36%), No way (12%), Uni ed Kingdom
(8%), Kazakhs an (8%), Libya (8%), and Nige ia (6%). Fo
oil p oduc s, Ge many p oduces mos o i s demand in i s
domes ic e ine ies. This co e ed 88% o o al demand in
2018. Accoun ing o mo e han hal o o al oil usage, he
anspo a ion sec o is he la ges oil consume . Diesel is
he mos consumed oil p oduc , ollowed by gasoline (IEA,
2020). Ge many has one o he la ges e ining capaci ies in
Eu ope, wi h a o al o 13 e ine ies sp ead ac oss he en i e
coun y (Mine alölwi scha s e band, 2020). A map wi h
he oil in as uc u e o Ge many is shown in Figu e 1.
In his map a e shown he loca ions o he 13 e ine ies,
he oil pipelines and oil s o age si es o ge a ough idea o
how he oil in as uc u e looks like. In he e ail uel sec-
o , he e we e 14460 ac i e gas s a ions in Ge many in 2022
(S a is a, 2023a). Figu e 2shows a map o Ge many wi h
he indi idual gas s a ions o ge an imp ession o how he
gas s a ions a e dis ibu ed in Ge many.
In Ge many, i e companies oge he own 68% o he
ma ke sha e o o al uel sales in 2022. These i e compa-
nies a e A al (21%), Shell (20%), Je (10%), To al (9.5%)
and Esso (7%). The ma ke sha e o gas s a ion ope a o s by
numbe o gas s a ions shows a qui e simila si ua ion. These
i e companies ope a e abou hal o all gas s a ions in Ge -
many (S a is a, 2023b).
The yea 2022 was shaped by e y high uel p ices in Ge -
many. Fo d i e s in Ge many, i was he mos expensi e yea
e e o ill up hei uel anks. A li e o diesel cos s an a e -
age o 1.95 eu os, and a li e o Supe E10 1.86 eu os. Gaso-
line was hus a ound 27 cen s mo e expensi e han in 2012,
he p e ious eco d yea , and he p ice pe li e o diesel was
e en almos 47 cen s highe han in 2012 (P ack, 2023). Fig-
u e 3shows an o e iew o how he a e age daily uel p ices
ha e de eloped in 2022.
The cou se o uel p ices his yea has been shaped by a
ew s iking ac o s. An ex emely la ge inc ease in he p ices
o all ypes o uel is seen a he end o Feb ua y. On 23
Feb ua y, he p ice o diesel was s ill a 1.67 eu os pe li e
and has isen wi hin abou wo and a hal weeks by 38%
pe cen o 2.31 eu os pe li e . This p ice shock was igge ed
by he in asion o Russian oops in o Uk aine on Feb ua y
24, 2022. As al eady desc ibed abo e, Russia is Ge many’s
la ges c ude oil supplie .
The nex sha p change in he end ook place on he i s
o June. Wi hin one day, he E5 p ice has allen by abou
28 cen s o 1.93 eu o pe li e . This was igge ed by he
so-called “Tank aba ”. In English, his means a go e nmen -
unded uel discoun . Ene gy ax a es we e educed o a
limi ed pe iod o h ee mon hs om June 1, 2022, o some
ypes o uel o he ex en pe mi ed unde Eu opean law. Fo
gasoline i was educed by 29.55 cen s pe li e and o diesel
by 14.04 cen s pe li e . As a consump ion ax, he ene gy
ax is in ended o be passed on in ull o he end consume .
The pu pose o he “Tank aba ” was o elie e people who
depend on he ca , such as commu e s, amilies, as well as
business people, especially in he c a s and logis ics indus-
ies (Bundes egie ung, 2022). On Augus 31, his bene i
expi ed, which esul ed in a huge inc ease in p ices o all
ypes o uel on he i s o Sep embe . The p ice o diesel
has isen by abou 10 cen s, E5 by abou 24 cen s and E10 by
abou 23 cen s pe li e om one day o he nex .
The ques ion a ises how he uel p ices o one li e a e
composed. B (2023) has illus a ed his on he basis o one
li e o diesel. A e ail p ice o 1,762 pe li e o diesel is
used. O his amoun , 0.2813 eu o is VAT, esul ing in a ne
selling p ice o 1.4807 eu o. 0.4704 eu o is due o ene gy
ax, 0.0950 eu o is due o CO2p icing, 0.0030 eu o is due o
pe oleum s ockpiling le y and he alue o goods excluding
axes (p ice o explo a ion, c ude oil, p ocessing and ans-
po a ion) equals 0.9123 eu o. This esul s in a o al amoun
o s a u o y le ies o 0.8497 eu os pe li e . This shows ha
he s a u o y le ies o m a la ge pa o he uel p ice.
4. Me hodology
To es he h ee hypo heses, he ollowing da a and
me hodology is used. The me hodology is s ongly inspi ed
by he pape “Loca ional choice and p ice compe i ion: some
empi ical esul s o he aus ian e ail gasoline ma ke ” by
Clemenz and Gugle (2006). I ha e a da ase o he Ge man
gas s a ions wi h he ollowing in o ma ion. Fo each gas
s a ion a sepa a e ID, he company and in o ma ion abou
he loca ion, such as zip code, s ee , and coo dina es. In
addi ion, o each gas s a ion o he mos pa hou ly diesel
p ices o he comple e yea 2022. The da a o igina es om
he websi e Tanke könig and was e ie ed using an API.
Based on his, i s a daily and hen an annual a e age diesel
p ice was calcula ed o each gas s a ion. Addi ional loca ion
in o ma ion, such as he dis ic key and ede al s a e key,
was added o each gas s a ion based on he zip code using
he OpenPLZ API. In Ge many, each municipali y can be p e-
cisely iden i ied by means o he egional key. Acco ding o
he Fede al S a is ical O ice (2023c), he de ini ion o he
egional key is: “12-digi key o unambiguous iden i ica-
ion o a municipali y wi h he componen s: Fede al s a e
(2 digi s), adminis a i e dis ic (1 digi ), dis ic (2 digi s),
associa ion o municipali ies (4 digi s) and municipali y (3
digi s).” A e da a cleaning, i.e. emo ing inco ec in o -
ma ion, duplica es and missing alues, a da ase wi h 14097
gas s a ions emained. So, abou 97.5% o he Ge man gas
s a ions we e conside ed in he analysis. This is a e y high
pe cen age o a coun y’s o al gas s a ions included in he
analysis, compa ed o many o he s udies. Some o he s ud-
ies use only a po ion o a coun y’s o al gas s a ions such as
o example Lee (2007) and Sen (2003) o Ecke and Wes
(2005).
I is a ela i ely di icul hing o di ide Ge many in o
app op ia e local gasoline ma ke s. The ma ke de ini ion
should no be oo na ow and no oo b oad. Following
Clemenz and Gugle (2006), each dis ic was de ined as a
local ma ke . I he espec i e ma ke s we e measu ed inac-
cu a ely, i is possible ha he es ima es unde es ima e he
ue ela ionship be ween he dependen and independen
N. Fiedle /Junio Managemen Science 10(2) (2025) 424-440 429
Figu e 1: Map o Ge many’s oil in as uc u e (Sou ce: IEA (2020))
Figu e 2: Map o he indi idual gas s a ions in Ge many
a iables. Unless hese imp ecisions a e co ela ed wi h ou
ele an a iables, inc eased whi e noise a ec ing s a is ical
signi icance is he mos likely consequence. In hei pape ,
he wo au ho s conduc ed a obus ness es by de ining mu-
nicipali ies in Aus ia as local ma ke s. Thei esul s a e o-
bus o he change in ma ke de ini ion. Fo his eason, i
is assumed ha o Ge many, he dis ic s also ep esen ap-
p op ia e local gasoline ma ke s. Based on his, he da a was
g ouped by dis ic and he co esponding alues we e calcu-
la ed a he dis ic le el. The a e age diesel p ice (dieselk)
and he numbe o gas s a ions (Nk) o each dis ic o he
yea 2022 we e calcula ed. In addi ion, wo measu es o
N. Fiedle /Junio Managemen Science 10(2) (2025) 424-440430
Figu e 3: Daily a e age p ices o diesel, E5 and E10 in Ge many in 2022
ma ke concen a ion we e calcula ed o each dis ic . Fi s ,
he ma ke sha e o he la ges i m in he espec i e coun y
(C1k). This was calcula ed as ollows:
C1k=N1,k
Nk
N1,kis he numbe o gas s a ions o he la ges company
in dis ic k. The la ges company is called he company wi h
he mos gas s a ions in dis ic k. Nkis he o al numbe o
gas s a ions in dis ic k. In addi ion, he ma ke sha e o
he ou la ges companies combined (C4k) was calcula ed
as ollows:
C4k=P4
n=1Nn,k
Nk
Nn,kis he numbe o gas s a ions ope a ed by he n la ges
i ms in dis ic k. Nkis as abo e, he o al numbe o gas
s a ions in dis ic k.
The a ea (Ak)(Bundesam , 2023b), numbe o inhabi-
an s (Popk) (Bundesam , 2023a) and numbe o mo o e-
hicles (V Tk) (K a ah -Bundesam , 2023) we e also added
o each dis ic . Based on his da a, he gas s a ion densi y
(SDk)was calcula ed as ollows:
SDk=Nk
Ak
The popula ion densi y (PDk):
PDk=Popk
Ak
The mo o - ehicle densi y (V Dk):
V Dk=V Tk
Ak
In addi ion, he numbe o mo o ehicles pe head (Vk).
Acco ding o Clemenz and Gugle (2006), gas s a ion densi y
is a sui able (in e se) app oxima ion o he a e age dis ance
be ween gas s a ions. Bu only wi h he condi ion ha gas
s a ions do no clus e in ce ain a eas. Thei esea ch sug-
ges s ha a clus e ing o gas s a ions does no eally occu
o en on a e age. The e o e, i can be assumed ha gas s a-
ion densi y is a sui able dis ance measu e and can be used
as a a iable o spa ial compe i ion.
All he impo an a iables a e lis ed in he able 1 o
sake o be e o e iew.
4.1. Gas s a ion densi y
To es he i s hypo hesis, se e al o dina y leas squa es
eg essions we e pe o med. The supe o dina e equa ion, in
analogy o (Clemenz & Gugle , 2006), looks as ollows:
lnSDk=β0+β1DEMAN Dk+β2Ck+ϵk
k=1, ..., 396 ep esen s he dis ic s in Ge many, lnSDk
is he na u al loga i hm o he densi y o gas s a ions pe
squa e kilome e o he espec i e dis ic . DEMAN Dk=
{lnPDk,lnV Dk}, wi h he na u al loga i hm espec i ely o
popula ion densi y, mo o ehicle densi y, and mo o ehi-
cles pe head. The e is no sales da a o he like a ailable o
e lec demand o diesel. Bu gasoline demand is no eally
p ice sensi i e (B ons e al., 2008; Espey, 1998; Hanly e al.,
2002). Fo his eason, he a iable popula ion densi y, mo o
ehicle densi y and mo o ehicles pe capi a ep esen he
di e en demand o diesel in he indi idual dis ic s qui e
well.
N. Fiedle /Junio Managemen Science 10(2) (2025) 424-440 431
Table 1: Va iable De ini ion
Va iable De ini ion
PopkNumbe o inhabi an s in dis ic k
AkA ea o dis ic k in squa e kilome e s
NkNumbe o gasoline s a ions in dis ic k
V TkNumbe o mo o - ehicles in dis ic k
a g_dieselk
Re ail p ice cha ged o diesel pe li e a e aged o e he yea 2022 a e aged o e all
s a ions wi hin dis ic k in cen
SDk=Nk/AkDensi y o gasoline s a ions in dis ic k. (In e se) p oxy o dis ances be ween s a ions
PDk=Popk/AkPopula ion densi y in dis ic k
SPk=Popk/NkInhabi an s pe gasoline s a ion in dis ic k
C1k
Ma ke sha e o he la ges i m in dis ic k de ined as C1k=N1,k
Nk, whe e N1,kis he
numbe o gasoline s a ions ope a ed by he la ges (mos gas s a ions in dis ic ) i m in
dis ic k
C4k
Sum o ma ke sha es o he la ges ou i ms in dis ic k de ined as C4k=P4
n=1Nn,k
Nk,
whe e Nn,kis he numbe o gasoline s a ions ope a ed by he n la ges i m in dis ic k
Vk
Deg ee o mo o iza ion de ined as he numbe o mo o -ope a ed ehicles pe head in
dis ic k
V Dk=V Tk/AkMo o - ehicle densi y in dis ic k
The con ol a iable in his model is ma ke concen a-
ion. The na u al loga i hm o he ma ke sha e o he la ges
(lnC1k) o he la ges ou i ms (lnC4k) in dis ic k. ϵk
ep esen s he e o e m. He e is an o e iew o he exac
equa ions ha we e analyzed:
lnSDk=β0+β1lnPDk+ϵk(1)
This equa ion was used o measu e pu ely he e ec o
popula ion densi y on gas s a ion densi y, also i se es as
a basis o compa e he esul s o he equa ions wi h con ol
a iables.
lnSDk=β0+β1lnV Dk+ϵk(2)
In eq. 2, mo o ehicle densi y se es as a p oxy o de-
mand. This a iable could be a be e p oxy o demand han
popula ion densi y, since o en in ci ies, i.e. e y densely pop-
ula ed a eas, he e a e ela i ely ewe people who own a
mo o ehicle. This would make he esul ing demand less.
Fo easons o compa abili y wi h he esul s o Clemenz and
Gugle (2006), his a iable was no used in he o he equa-
ions.
lnSDk=β0+β1lnPDk+β2lnC1k+ϵk(3)
lnSDk=β0+β1lnPDk+β2lnC4k+ϵk(4)
In eq. 3and eq 4, ma ke concen a ion was included as
a con ol a iable. In he e ail gasoline ma ke , i canno be
assumed ha each company ope a es only one s a ion. The e
is ye no clea answe as o how ma ke concen a ion a ec s
he numbe o gas s a ions in a local ma ke . Howe e , i is
ai ly sa e o say ha he numbe o s o es a pu e monopolis
will se up in he absence o en y h ea is he lowe bound.
On he o he hand, he uppe limi on he numbe o s o es is
gi en by he numbe o loca ions a monopolis will es ablish
i he ma ke is ee o en e . Mo eo e , he e is no clea indi-
ca ion o he ela ionship be ween ma ke concen a ion and
gas s a ion densi y (Clemenz & Gugle , 2006). Fo his ea-
son, he wo ma ke concen a ion measu es we e included
in he analysis because hey may al e he e ec o popula ion
densi y.
lnSDk=β0+β1lnPDk+β2lnV k+ϵk(5)
In eali y, he choice o a company’s loca ion is in lu-
enced by many o he ac o s han he numbe o po en ial
cus ome s alone. The demand pe indi idual po en ial cus-
ome is also a ac o . Fo his eason, he numbe o ca s
pe capi a is included in equa ion 5 o ge a be e p oxy o
demand.
lnSDk=β0+β1lnPDk+ϵk(6)
Equa ion 6is, ob iously, he same as Equa ion 1, bu he
da a se ha was analyzed is di e en . Popula ion densi y
a ies conside ably be ween dis ic s, as u ban a eas a e also
conside ed. I could be ha en y decisions o companies
in ci ies a e in luenced by di e en ac o s han in u al a -
eas, o example, a ailabili y o space, highe se up cos s
o highe en s, e c. Fo his eason, he obus ness o he
esul s is es ed by emo ing he 20% o dis ic s wi h he
highes popula ion densi y om he o iginal da ase . This
N. Fiedle /Junio Managemen Science 10(2) (2025) 424-440438
ula ion densi y shows he ele ance o au omobili y. A highe
numbe o mo o ehicles pe squa e kilome e could po en-
ially lead o g ea e use o exis ing gas s a ions, he eby in-
c easing he demand o new gas s a ions in ha pa icula
a ea. This is also shown by he coe icien o he loga i hm
o mo o ehicles pe capi a in Equa ion 5. I mo o ehi-
cles pe capi a inc ease by en pe cen in a dis ic , hen gas
s a ion densi y inc eases by app oxima ely 0.275%. Includ-
ing a iables ela ed o mo o ehicles in he analysis makes
sense o possibly mi iga e he e ec o ci ies o some ex en .
In densely popula ed u ban a eas, public anspo is o en
be e de eloped, which could lead o ewe people owning
a ca . In u n, his could ha e an e ec on gas s a ion densi y.
When he 20% o dis ic s wi h he highes popula ion den-
si y, which a e mos ly ci ies, a e emo ed om he analysis,
he e ec o he popula ion densi y inc eases sligh ly.
Al hough ma ke concen a ion was only a con ol a i-
able in he es o he i s hypo hesis, i is in e es ing o
look a his e ec as well. Ma ke concen a ion has a neg-
a i e impac on gas s a ion densi y, so he highe he ma -
ke concen a ion, he lowe he a e age gas s a ion densi y.
This sugges s ha ma ke s wi h ew dominan playe s end
o ha e ewe gas s a ions. One eason could be ha he
compe i i e ba ie s ha new playe s ha e o o e come in
a highly concen a ed ma ke mean ha ewe new gas s a-
ions a e opened. Finding ou he exac causes, howe e , is
beyond he scope o his pape and could be he subjec o
u he esea ch. In a concen a ed ma ke , po en ial cus-
ome s could ace mo e limi ed op ions o choose om. On
he one hand, due o gene ally ewe gas s a ions, and on
he o he hand, he la ge companies ope a e se e al gas s a-
ions, which makes he numbe o gas s a ions ope a ed by
a ious o he companies smalle . The esul s sugges ha
ma ke concen a ion plays a ole in shaping he s uc u e
o he gas s a ion ma ke . In e ms o compe i ion and eg-
ula o y policy, ma ke concen a ion should be closely moni-
o ed u he o ensu e ha i does no a ec compe i i eness
and consume choice.
O he s udies ha e al eady ound a somewha simila e -
ec . As al eady shown in he esul s sec ion, he coe icien o
he one i m a io was -0.1698 (p <0.01), while o Clemenz
and Gugle (2006) i was -0.132. The coe icien s o he ou
i m a io a e also wi hin he same ange. I can be p edic ed
ha i he ma ke sha e o he la ges company in he coun y
inc eases by one pe cen age poin , hen he gas s a ion den-
si y dec eases by abou 0.15% on a e age. As you can see,
he e is a a he small, bu ne e heless signi ican e ec . The
esul s o Gö z and Gugle (2006) also gi e suppo o his e -
ec .
The esul s shown abo e suppo he second hypo hesis
ha unde spa ial compe i ion, p ices end o be lowe as he
densi y o selle loca ions inc eases. In conc e e e ms, i can
be obse ed ha a one pe cen age poin inc ease in he den-
si y o gas s a ions is associa ed, on a e age, wi h a dec ease
o abou 0.002% in he diesel p ice in a dis ic . One cause
may be he inc eased spa ial compe i ion ha comes wi h an
inc easing numbe o gas s a ions in a geog aphically de ined
ma ke . I he densi y o gas s a ions inc eases, hen logi-
cally he a e age dis ance be ween he gas s a ions becomes
smalle . As al eady explained in he heo y sec ion, a smalle
dis ance be ween compe ing gas s a ions leads o mo e in-
ense compe i ion. This mo e in ense compe i ion may lead
gas s a ion ope a o s o educe p ices in o de o a ac po-
en ial cus ome s and hus secu e o inc ease hei ma ke
sha e.
Ano he possible cause could be he supply and demand
ela ionship. A la ge numbe o gas s a ions in a ma ke
could inc ease he supply o uel. I he demand o uel in a
ma ke is ela i ely cons an , an inc eased supply could lead
o a dec ease in p ices.
In hei pape , Clemenz and Gugle (2006) looked a he
impac o gas s a ion densi y on he ma gin a gas s a ions
a he dis ic le el. Bu he e a e d awbacks o calcula ing
he ma gin in hei s udy. They make some gene alized as-
sump ions o calcula ing he cos s. Fo example, hey make
blanke assump ions o anspo a ion cos s o and wi hin
Aus ia and do no dis inguish be ween gas s a ions o dis-
ic s. They hus calcula e a single nume ical alue o he
cos s wi h which hey hen calcula e he ma gin o all gas
s a ions. I chose he diesel p ice a he han he ma gin as he
dependen a iable because his way he esul s a e no dis-
o ed by possible inaccu a e assump ions. E en i i makes
no sense o compa e he absolu e alues, i is a leas possible
o compa e he di ec ion o he e ec . The wo esea che s
also ound a signi ican nega i e e ec o gas s a ion densi y.
They also pe o med a obus ness es by excluding Vienna
dis ic s om he analysis. Thei esul s a e obus o he
change in he analyzed da a. I also pe o med a obus ness
es as desc ibed ea lie by excluding he 20% o dis ic s wi h
he highes popula ion densi y. The e ec is s ill nega i e bu
no longe signi ican (p =0.656). The educed da ase has an
a e age gas s a ion densi y o 0.045 gas s a ions pe squa e
kilome e , o in e sely 30.218 squa e kilome e s pe gas s a-
ion. The excluded da a, on he o he hand, has an a e age
gas s a ion densi y o 0.267 gas s a ions pe squa e kilome-
e o , in e sely 4.410 squa e kilome e s pe gas s a ion. You
can see ha he e is a e y big di e ence be ween he a e age
densi y o gas s a ions. This subs an ial di e ence could be
he eason why he esul s in he obus ness es a e no longe
signi ican . The ela i ely high densi y o gas s a ions in he
excluded dis ic s shows a high deg ee o spa ial compe i ion.
Lee (2007) was able o disco e ha gas s a ions compe e
mos when hey a e less han a mile apa and he in ensi y
o compe i ion con inues o dec ease wi h dis ance. Ca doso
e al. (2020) ound ha he close he compe i o ’s loca ion
is o exis ing gas s a ions, he mo e p ices all. In u al a eas,
he a e age dis ance be ween gas s a ions may be oo la ge
o compe i ion o signi ican ly a ec p ices. This could be
he eason why he e ec o gas s a ion densi y on he diesel
p ice is no longe signi ican in he educed da a se . I may
be ha he dis ic is chosen oo la ge as he local ma ke in
u al a eas. Ano he eason could be ha he ela ionship be-
ween gas s a ion densi y and diesel p ice in u ban a eas may
be in luenced by o he ac o s ha a e mo e p onounced in
N. Fiedle /Junio Managemen Science 10(2) (2025) 424-440 439
ci ies. These indings may open up new esea ch pe spec i es
o in es iga e in mo e de ail he speci ic in luence ac o s in
u ban and u al a eas and o unde s and he ela ionship be-
ween gas s a ion densi y and diesel p ice mo e ex ensi ely.
Kaldo (1935) has ound ha i ms compe e no only in a lo-
cal ma ke bu also wi h i ms in neighbo ing ma ke s. Com-
pe i ion is less in ense, bu i is s ill he e. This is a ac o
ha was no included in his pape , bu could be conside ed
in u he esea ch.
The esul s shown abo e suppo he hi d hypo hesis ha
p ices end o be highe , as ma ke concen a ion inc eases.
In conc e e e ms, i can be obse ed ha a one pe cen in-
c ease in he ou i m a ion is associa ed, on a e age, wi h
an inc ease o abou 0.0146% in he diesel p ice in a dis-
ic . This e ec could be due o a ious causes based on he
ma ke powe o he companies. As ma ke concen a ion in-
c eases, so does a company’s ma ke powe , since he e a e
ewe compe i o s who could o ce p ice cu s. In such a sce-
na io, he di e en gas s a ions o he same company could
ha e incen i es o keep p ices high. Gas s a ions o he same
company will mos likely no compe e wi h each o he . I
one o a ew companies ha e a la ge ma ke sha e, hen his
could lead o less p ice compe i ion o e all. Highe ma ke
concen a ion o ma ke sha e would no necessa ily ha e o
be due o ma ke powe , bu could also be due o he i ms’
e iciency.
Bu Muelle (1986) made clea in his wo k ha i high
concen a ion and high ma ke sha es we e due o highe e -
iciency, one would expec lowe p ices in concen a ed ma -
ke s, which he did no obse e. This is consis en wi h he
p esen ed esul s.
High ma ke concen a ion could also esul in a high en-
y ba ie o new ma ke pa icipan s. I may be mo e di -
icul o new compe i o s o es ablish hemsel es i a ew
companies domina e he majo i y o he ma ke . As al eady
shown in he esul s, ma ke concen a ion has a nega i e
impac on gas s a ion densi y. Fo his eason, he e a e, on
a e age, p opo ionally ewe gas s a ions in a concen a ed
ma ke ha could be in compe i ion wi h each o he . This
would also lead o less spa ial compe i ion.
Ano he eason o he highe p ices could be ac ical col-
lusion. Pepall e al. (2014) lis ed se e al ac o s in hei book
ha acili a e collusion, including high ma ke concen a ion,
signi ican en y ba ie s, and p oduc homogenei y. These
a e all poin s ha play a ole in he Ge man e ail gasoline
ma ke . Co e ill (1990) has shown ha aci collusion does
exis in concen a ed ma ke s. Taci collusion is also o en de-
sc ibed as p ice ollowship. Bal o (2001) has shown ha he
highe he ma ke concen a ion, he g ea e he isk ha a
u he inc ease will lead o highe p ices h ough collusion.
As desc ibed in he heo y sec ion, some esea che s ha e
looked a he ela ionship be ween ma ke concen a ion and
p ices in he e ail gasoline ma ke . Clemenz and Gugle
(2006) did no ind a signi ican ela ionship a he dis ic
le el. In con as , I was able o ind a signi ican ela ion-
ship. In he wo k o Be gan ino e al. (2018), he coe icien s
o b and ma ke sha e, h ee i m a io, and HHI we e also
all posi i e and highly signi ican in he eg essions o gaso-
line and diesel p ices. The di ec ion o he e ec o ma ke
concen a ion is also consis en wi h he indings o many
o he esea che s, such as Ecke and Wes (2004), Kihm e
al. (2016), and Sen (2003).
Due o he lack o sales and e enue da a, he ma ke con-
cen a ion was calcula ed on he basis o he numbe o gas
s a ions ope a ed by each company. I is impo an o no e
ha his me hod may no cap u e all he sub le nuances o
he ma ke s uc u e. Fu he esea ch could conside al e -
na i e me hods o measu ing ma ke concen a ion, i sales
da a om indi idual gas s a ions a e a ailable, o alida e
he esul s.
The e ec o ma ke concen a ion was obus o he e-
duc ion in he size o he da a analyzed. This sugges s ha
ma ke concen a ion also has an impac on p ices in less
densely popula ed a eas.
I should be no ed ha a po en ial endogenei y p oblem
could be in he in e ac ion be ween ma ke concen a ion
and diesel p ices. Singh and Zhu (2008) cla i y ha he
undamen al p oblem is ha “ma ke s uc u es a e no an-
domly assigned”. They s a e ha companies ake in o ac-
coun demand and cos condi ions as well as po en ial com-
pe i o s when making ma ke en y decisions. The ma ke
s uc u e hus eme ges as a consequence o hese s a egic
decisions. They u he poin ou he example ha ma ke s
wi h unobse ed high cos s a e likely o ha e highe p ices.
On he o he hand, hese ma ke s a e also likely o a ac
ewe ma ke pa icipan s. This may in u n lead o a highe
ma ke concen a ion. The p ice would hus be in luenced o
a ce ain ex en by he ela i ely highe cos s in ce ain e-
gions and no mainly by ma ke concen a ion. A opic o
u he esea ch would be o wha ex en he ma ke s uc-
u e o ma ke concen a ion in he Ge man e ail gasoline
ma ke is endogenously de e mined by p ices.
The indings o his pape no only con ibu e o he he-
o e ical unde s anding o he spa ial ela ionships be ween
popula ion densi y, gas s a ion densi y, ma ke concen a ion,
and uel p ices, bu also o e po en ial applica ions o bo h
academic esea ch and he co po a e wo ld. Fo u he e-
sea ch, some app oaches ha e al eady been ouched upon.
In summa y, he indings ob ained could se e as a basis o
s udy simila spa ial ela ionships in o he indus ies o ge-
og aphic egions. Applica ions in he co po a e wo ld a e
also eme ging om he indings o his s udy. Fo gas s a-
ion owne s, he esul s o e aluable insigh s in o loca ion
planning. The ela ionship be ween gas s a ion densi y, ma -
ke concen a ion and p ices could lead companies o adjus
hei compe i i e s a egies. In ma ke s wi h high concen a-
ion, di e en ia ion s a egies o coope a ions could be mo e
ele an , while in highly compe i i e ma ke s, p ice leade -
ship could play a ole. Companies could also make s a egic
decisions such as acquisi ions o gas s a ion loca ions based
on he indings on he ela ionship be ween ma ke concen-
a ion and p ices. As a conclusion o companies, one could
d aw ha on a e age i makes mo e sense o ake o e gas
s a ions om o he companies han o build new si es, espe-
N. Fiedle /Junio Managemen Science 10(2) (2025) 424-440440
cially in densely popula ed a eas. This way, he densi y o
gas s a ions does no inc ease and does no ha e a nega i e
e ec on he p ice, bu he ma ke concen a ion becomes
highe and has a posi i e e ec on he p ice.
Re e ences
Asplund, M., & Sandin, R. (1999). Compe i ion in in e ela ed ma ke s: An
empi ical s udy. In e na ional Jou nal o Indus ial O ganiza ion,
17(3), 353–369.
Bal o, D. A. (2001). Supe ma ke me ge en o cemen . Jou nal o Public Pol-
icy & Ma ke ing,20(1), 38–50.
Ba on, J. M., Taylo , B. A., & Umbeck, J. R. (2004). Numbe o selle s, a -
e age p ices, and p ice dispe sion. In e na ional Jou nal o Indus-
ial O ganiza ion,22(8), 1041–1066. h ps://doi.o g/10.1016
/j.ijindo g.2004.05.001
Belle lamme, P., & Pei z, M. (2015). Indus ial o ganiza ion: ma ke s and
s a egies. Camb idge Uni e si y P ess.
Be gan ino, A. S., Capozza, C., & In ini, M. (2018). Empi ical in es iga ion
o e ail gasoline p ices.
Be gan ino, A. S., Capozza, C., & In ini, M. (2020). Empi ical in es iga ion o
e ail uel p icing: The impac o spa ial in e ac ion, compe i ion
and e i o ial ac o s. Ene gy Economics,90, 104876. h ps://do
i.o g/10.1016/j.eneco.2020.104876
B . (2023). Zusammense zung des Benzin-/Dieselp eises. Re ie ed Augus
21, 2023, om h ps://www.b .de/da en-und- ak en/benzinp
eis-zusammense zung
B ons, M., Nijkamp, P., Pels, E., & Rie eld, P. (2008). A me a-analysis o
he p ice elas ici y o gasoline demand. A SUR app oach. Ene gy
Economics,30(5), 2105–2122. h ps://doi.o g/10.1016/j.eneco
.2007.08.004
Bundesam . (2023a). Fo sch eibung des Be ölke ungss andes. Re ie ed
Augus 21, 2023, om h ps://www- genesis.des a is.de/gen
esis/online?ope a ion=s a is ic%5C&le elindex=0%5C&le elid
=1682174311638%5C&code=12411%5C#ab eadc umb
Bundesam . (2023b). Gebie s läche: K eise, S ich ag. Re ie ed Augus 21,
2023, om h ps://www-genesis.des a is.de/genesis/online%5
C#as uc u e
Bundesam . (2023c). Regionales - Regionalschlüssel (RS). Re ie ed Augus
12, 2023, om h ps://www.des a is.de/DE/Themen/Laende -
Regionen/Regionales/Gemeinde e zeichnis/Glossa / egionalsc
hluessel.h ml
Bundes egie ung. (2022). F agen und An wo en zum „Tank aba “. Re-
ie ed Augus 21, 2023, om h ps://www.bundes egie ung.d
e/b eg-de/ak uelles/ aq-ene gies eue senkung-2049702
Ca doso, L. C. B., Uchôa, F., Huamani, W., & Gomez, R. V. (2020). P ice e ec s
o spa ial compe i ion in B azilian gas s a ions.
Clemenz, G., & Gugle , K. (2006). Loca ional choice and p ice compe i ion:
some empi ical esul s o he aus ian e ail gasoline ma ke . Em-
pi ical Economics,31(2), 291–312. h ps://doi.o g/10.1007/s00
181-005-0016-7
Conno , J. M. (1990). Empi ical challenges in analyzing ma ke pe o -
mance in he US ood sys em. Ame ican Jou nal o Ag icul u al
Economics,72(5), 1219–1226.
Co e ill, R. W. (1986). Ma ke powe in he e ail ood indus y: E idence
om Ve mon . The Re iew o Economics and S a is ics, 379–386.
Co e ill, R. W. (1990). Food me ge s: Implica ions o pe o mance and pol-
icy. Re iew o Indus ial O ganiza ion, 189–202.
Ecke , A., & Wes , D. (2004). A Tale o Two Ci ies: P ice Uni o mi y and P ice
Vola ili y in Gasoline Re ailing. The Annals o Regional Science,38,
25–46. h ps://doi.o g/10.1007/s00168-003-0144-y
Ecke , A., & Wes , D. S. (2005). P ice uni o mi y and compe i ion in a e ail
gasoline ma ke . Jou nal o Economic Beha io & O ganiza ion,
56(2), 219–237. h ps://doi.o g/10.1016/j.jebo.2003.09.006
Espey, M. (1998). Gasoline demand e isi ed: An in e na ional me a-
analysis o elas ici ies. Ene gy Economics,20(3), 273–295.
Gö z, G., & Gugle , K. (2006). Ma ke concen a ion and p oduc a ie y
unde spa ial compe i ion: E idence om e ail gasoline. Jou nal
o Indus y, Compe i ion and T ade,6, 225–234.
Hanly, M., Da gay, J., & Goodwin, P. (2002). Re iew o income and p ice
elas ici ies in he demand o oad a ic. Depa men o T ans-
po , London.
Hosken, D. S., McMillan, R. S., & Taylo , C. T. (2008). Re ail gasoline p icing:
Wha do we know? In e na ional Jou nal o Indus ial O ganiza-
ion,26(6), 1425–1436.
IEA. (2020). Ge many 2020 - Ene gy Policy Re iew. h ps://www.iea.o g/
epo s/ge many-2020
Kaldo , N. (1935). Ma ke impe ec ion and excess capaci y. Economica,
2(5), 33–50.
Keele , E. B., Melnick, G., & Zwanzige , J. (1999). The changing e ec s o
compe i ion on non-p o i and o -p o i hospi al p icing beha io .
Jou nal o Heal h Economics,18(1), 69–86.
Kihm, A., Ri e , N., & Vance, C. (2016). Is he Ge man e ail gasoline ma ke
compe i i e? A spa ial- empo al analysis using quan ile eg es-
sion. Land Economics,92(4), 718–736.
Kolle , R. H., & Weiss, L. W. (1989). P ice le els and selle concen a ion:
The case o Po land Cemen . Concen a ion and P ice, 17–40.
K a ah -Bundesam . (2023). Bes and nach Zulassungsbezi ken (FZ 1). Re-
ie ed Augus 12, 2023, om h ps://www.kba.de/DE/S a is i
k/P oduk ka alog/p oduk e/Fah zeuge/ z1%5C_b%5C_uebe si
ch .h ml
Lee, S.-y. (2007). Spa ial Compe i ion in he Re ail Gasoline Ma ke : An
Equilib ium App oach Using SAR Models.
Mee beeck, W. (2003). Compe i ion and Local Ma ke Condi ions on he Bel-
gian Re ail Gasoline Ma ke . De Economis ,151, 369–388. h ps:
//doi.o g/10.1023/B:ECOT.0000006590.66223.9a
Mine alölwi scha s e band. (2020). MWV-Jah esbe ich 2020. h ps://en
2x.de/se ice/publika ionen/
Muelle , D. C. (1986). P o i s in he long un. Camb idge Uni e si y P ess.
Newma k, C. M. (2004). P ice-concen a ion s udies: The e you go again.
An i us Policy Issues, 9–42.
Pepall, L., Richa ds, D., & No man, G. (2014). Indus ial o ganiza ion: Con-
empo a y heo y and empi ical applica ions. John Wiley & Sons.
P ack, N. (2023). Sp i p eis-En wicklung: Benzin- und Dieselp eise sei
1950. Re ie ed Augus 21, 2023, om h ps://www.adac.de/
e keh / anken-k a s o -an ieb/deu schland/k a s o p eisen
wicklung/
Salop, S. (1979). Monopolis ic Compe i ion Wi h Ou side Goods. Bell Jou -
nal o Economics,10, 141–156. h ps://doi.o g/10.2307/30033
23
Sen, A. (2003). Highe p ices a Canadian gas pumps: In e na ional c ude oil
p ices o local ma ke concen a ion? An empi ical in es iga ion.
Ene gy Economics,25(3), 269–288.
Singh, V., & Zhu, T. (2008). P icing and ma ke concen a ion in oligopoly
ma ke s. Ma ke ing Science,27(6), 1020–1035.
S a is a. (2023a). Anzahl de Tanks ellen in Deu schland nach Tanks ellen-
yp on 1999 bis 2022. Re ie ed Augus 12, 2023, om h ps:
//de.s a is a.com/s a is ik/da en/s udie/72262/um age/anza
hl-de - anks ellen-in-deu schland-nach- anks ellen yp-zei eihe
/%5C#:~: ex =Im%5C%20Jah %5C%202022%5C%20gab%5
C%20es,im%5C%20Ve gleich%5C%20zum%5C%20Vo jah %5
C%20nich .
S a is a. (2023b). S a is ik-Repo zum deu schen Tanks ellenma k . h ps:
//de.s a is a.com/s a is ik/s udie/id/26070/dokumen / anks el
lenma k -s a is a-dossie /
Tanke könig. (2022). Sp i p eise in Ech zei . h ps:// anke koenig.de/inde
x.php
Ti ole, J. (1988). The heo y o indus ial o ganiza ion (Vol. 14). MIT P ess.
