Janßen, Thiemo
A icle
F om pic u es o pe cep ions: Explo ing he s a egic use
o isuals in CSR epo s and he impac o egula o y
manda es
Junio Managemen Science (JUMS)
P o ided in Coope a ion wi h:
Junio Managemen Science e. V.
Sugges ed Ci a ion: Janßen, Thiemo (2025) : F om pic u es o pe cep ions: Explo ing he s a egic
use o isuals in CSR epo s and he impac o egula o y manda es, Junio Managemen Science
(JUMS), ISSN 2942-1861, Junio Managemen Science e. V., Planegg, Vol. 10, Iss. 4, pp. 1009-1027,
h ps://doi.o g/10.5282/jums/ 10i4pp1009-1027
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/334186
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(4) (2025) 1009-1027
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 4, Decembe 2025
JUNIOR
MANAGEMENT
SCIENCE
Ma ie-Clai e Joyeaux,Wo k Less, Li e Mo e? The Impac o an
In oduc ion o he Fou -Day Wo king Week on
Happiness in he Con ex o he Icelandic Fou -Day
Wo king Week Expe imen
Niklas Janßen, In eg a ing Sus ainabili y in Risk Managemen
and In e nal Con ol Sys ems: An Empi ical
Assessmen o ESG Repo ing o Ge man DAX40 Fi ms
Tobias Kese ü,A New Dimension o T anspa ency: ESG
Disclosu e and I s E ec on Sha eholde Beha io
An onia Engel, ESG Regula ion Ac oss he Globe: Does ESG
Regula ion Pay O ?
Finn Ma hes Gooßen, The Impac o Female Boa d Membe s
on ESG Pe o mance: An Empi ical Analysis
Louisa Felici as Bläße , Implici Measu emen o he Mo al
Sel -Image Using he Go/No-Go Associa ion Task
(GNAT) - An Empi ical In es iga ion o he Con e gen
Validi y Be ween Explici and Implici Measu es
Lau a Wi edu, Who Bea s he Cos s o he UK So D ink Tax?
An Empi ical S udy o Medium-Te m E ec s
Thiemo Janßen, F om Pic u es o Pe cep ions: Explo ing he
S a egic Use o Visuals in CSR Repo s and he Impac
o Regula o y Manda es
Vanessa Jeske, De e minan s o Co po a e Bond Mu ual Fund
Flows
Felix Yumuşak, The In luence o Leade ship S yle on he
Accep ance o Gene a i e AI in he Wo kplace -The
Role o O ganiza ional Commi men , Job Insecu i y
and In e ac ion F equency
831
858
876
904
940
966
985
1009
1028
1053
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
F om Pic u es o Pe cep ions: Explo ing he S a egic Use o Visuals in CSR Repo s
and he Impac o Regula o y Manda es
Thiemo Janßen
Technical Uni e si y o Munich
Abs ac
This hesis explo es he s a egic use o isuals in CSR epo s, compa ing companies in he EU and Swi ze land. Using
au oma ed image classi ica ion and clus e ing, 11,455 images om sus ainabili y epo s we e analyzed. The s udy inds ha
i ms in en i onmen ally o socially sensi i e indus ies, especially hose wi h highe CO2emissions pe e enue, end o use
mo e images—o en aligned wi h hei speci ic challenges. This sugges s ha isuals a e no me ely illus a i e bu se e o
shape co po a e pe cep ion, po en ially di e ing a en ion om nega i e impac s. The in oduc ion o he EU’s Non-Financial
Repo ing Di ec i e (NFRD) appea s o educe image eliance, indica ing a posi i e egula o y e ec on anspa ency. O e all,
he indings highligh ha images can be used no jus o complemen ex , bu o cons uc a mo e a o able co po a e
na a i e. The s udy unde sco es he impo ance o c i ically assessing isual elemen s in CSR disclosu es, as hey may sub ly
in luence s akeholde s’ pe cep ion beyond wha is e bally communica ed.
Keywo ds: CSR; image analysis; NFRD; g eenwashing; isual communica ion
1. In oduc ion
One o he mos dominan opics in oday’s wo ld, bo h
o ins i u ions and o socie y in gene al, is he ad ance o
global wa ming and i s consequences (Ca le ion & Hsiang,
2016). Looking a he mos signi ican con ibu o s o his
de elopmen , we ind ha companies and hei ac ions emi
a as majo i y o he esponsible global g eenhouse gas emis-
sions. The indus y sec o alone is esponsible o 37 pe cen
o hese emissions (Wo ell e al., 2009). This eali y has
led o a heigh ened ocus on he en i onmen al ac i i ies o
companies and, in he shadow o his de elopmen , also on
he esponsible beha io o companies in gene al, o en e-
e ed o as CSR (I use he e m CSR in e changeably wi h
ESG h oughou his pape . When I men ion sus ainabili y, I
also mean he ealm o CSR1). This led o s akeholde s calling
1I is impo an o no e, hough, ha CSR gene ally e e s o a company’s
olun a y ac ions o con ibu e posi i ely o socie y and he en i onmen ,
emphasizing e hical beha io and social impac . ESG, on he o he hand,
is a amewo k o assessing a company’s impac in speci ic a eas o en-
o g ea e anspa ency and accoun abili y h ough comp e-
hensi e epo ing (Huang & Kung, 2010)
Due o his inc easing p essu e on companies, we ha e
seen a sha p inc ease in he numbe o companies issuing sus-
ainabili y epo s (Se a eim & Amel-Zadeh, 2017; S olowy &
Paugam, 2018). The quali a i e na u e o CSR (compa ed o
he mo e quan i a i e na u e o inancial epo s) and he ac
ha di e en indus ies a e a ec ed di e en ly pose chal-
lenges o he eade s in ex ac ing he meaning ulness o
hese epo s and compa ing hem o o he companies (Ch is-
ensen e al., 2021; Se a eim & Amel-Zadeh, 2017). Some
coun ies espond e en u he o his call o anspa ency
by in oducing epo ing manda es, such as he EU’s Non-
Financial Repo ing Di ec i e (NFRD). The goal is o inc ease
he compa abili y and anspa ency o hese epo s (Di ec-
i e 2014/95, eci al 1) and c ea e an incen i e o d i e sus-
ainabili y e o s (Di ec i e 2014/95, eci al 3). While e-
i onmen , social in ol emen , and go e nance, o en linked wi h how
hese a e in eg a ed in o he company’s ope a ional and s a egic man-
agemen .”
DOI: h ps://doi.o g/10.5282/jums/ 10i4pp1009-1027
© 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.
T. Janßen /Junio Managemen Science 10(4) (2025) 1009-10271010
po ing can inc ease anspa ency, companies also use e-
po s o hei own in e es , which can manipula e eade s’
pe cep ions (Siano e al., 2017). A heo y applied o explain
his is Legi imacy Theo y, which a gues ha companies use
epo s o jus i y hei ac ions and con ey ha hey ac ac-
co ding o he g ea e good o socie y (Boi al & Hen i, 2017;
Cho e al., 2009; Suchman, 1995). Following his, i can be
easoned ha he use o , e.g., boile pla e language o he de-
libe a e aming o epo s can be used o c ea e a posi i e
image o he company while i does no change i s beha io
(Boi al, 2013; Ch is ensen e al., 2021). The quali a i e na-
u e o CSR epo s men ioned be o e makes alsi ying claims
and dis inguishing be ween ac ual in o ma ion e en ha de .
While i could be a gued ha a manda e would sol e his
p oblem by es ablishing s anda ds, i has been shown in he
pas ha epo ing manda es o en inc ease he use o boil-
e pla e language (Dye e al., 2017).
Besides boile pla e language, he use o images in CSR
epo s is pa icula ly in e es ing in his con ex . Visuals can
be po en ools o in luencing pe cep ion, as hey o en shi
he ocus owa ds hem (T e sky, 1974), subconsciously a -
ec one’s pe cep ion (Posne e al., 1976), and do no o e
quan i iable da a ha can be c i ically assessed.
This aises he ques ion o how and why companies use
images in hei CSR epo s. To add ess his, I i s in es iga e
he implemen a ion o CSR epo ing, i s his o ical con ex ,
and he implica ions o in oducing manda es like he NFRD.
This is ollowed by an examina ion o ends in CSR epo ing
beha io s in gene al and he ole o isual elemen s, oge he
wi h hei in luence on s akeholde pe cep ion.
Building on his ounda ion, he hesis will explo e which
speci ic cha ac e is ics o companies in luence hei use o im-
ages in CSR epo s and whe he a di e ence eme ges in he
p esence o epo ing manda es. By compa ing he Eu opean
Union, wi h i s NFRD manda e, and Swi ze land, which lacks
such a manda e, we will gain insigh s in o he impac o man-
da es on he use o isuals wi hin hese epo s. The ea-
son behind compa ing he EU wi h Swi ze land is ha hose
coun ies, apa om he manda e, ace simila socie al p es-
su es and ope a e in a simila en i onmen .
Fo s akeholde s o e ec i ely e alua e CSR Repo s, i
is essen ial o ha e a comp ehensi e unde s anding o he
unde lying mo i a ions and con ex s ha d i e he use o i-
suals wi hin hese documen s. Images possess he abili y o
in luence iewe s, o en on a subconscious le el. They can
ei he highligh achie emen s o di e a en ion away om
un a o able elemen s. I is essen ial o be able o iden i y in-
s ances whe e images a e employed no o cla i y bu a he
o manipula e pe cep ion.
2. CSR Repo ing: Why i Ma e s, His o y and he NFRD
2.1. Inc easing ocus on CSR
In he co po a e landscape, s akeholde s play a c ucial
ole in shaping he ac i i ies o companies, including hose e-
la ed o CSR. Companies a e inc easingly expec ed o ope a e
sus ainably and e hically. An example o his de elopmen is
he Business Round able’s 2019 s a emen on he pu pose o
business. Top US execu i es decided ha a company’s p i-
ma y pu pose is no only o se e i s sha eholde s bu also
i s cus ome s, employees, supplie s, and communi ies (Busi-
ness Round able, 2019). “These execu i es a e esponding o
moun ing p essu e ha a company needs o do ’good’ while
doing business, whe he ha means keeping ca bon emis-
sions low, wa e ways clea , o wo ke s heal hy” (Ch is ensen
e al., 2021, p. 1177). F om a mo e gene al pe spec i e, he
17 Sus ainable De elopmen Goals (SDGs), a co e pa o he
Uni ed Na ions (UN) Agenda 30, also indica e a shi in wha
is expec ed om companies by go e nmen s and socie y, de-
ailing he opics a CSR ini ia i e can o should be cen e ed
a ound. E en hough hese a e no explici ly published o
guideline coope a ion’s hey sugges an a i ude change and
an enhanced awa eness o sus ainabili y opics om socie y
and consequen ly he consume s.
Tha consume s a e an in eg al pa o a company’s busi-
ness is sel -e iden . Howe e , in he ealm o CSR, consume s
also play he mos signi ican ole in shaping co po a e ac-
ions owa ds CSR ac i i ies. A s udy om Accen u e e-
ealed ha 68% o CEOs acknowledge consume s as he mos
in luen ial s akeholde g oup conce ning hei sus ainabili y
agenda ( ollowed by go e nmen s and in es o s) (Accen u e,
2022).
The in es o ’s ocus on CSR can be seen in a s udy con-
duc ed by BlackRock (2020). They asce ain ha he e has
been an inc ease in Asse s unde Managemen ela ed o
CSR-con o ming p oduc s om 895 billion in 2015 o 1833
billion in 2020 (only up o Sep embe ). They u he s a e
ha o 78% o all in es o s su eyed, CSR is a cen al com-
ponen o hei s a egy. While he mos p ominen eason
In es o s adop sus ainabili y in es ing is ha i is he “ igh
hing o do”, he second and hi d mos men ioned easons
a e isk-based (BlackRock, 2020). F om his, you can a gue
ha in es o s ecognize he p essu e om socie y and con-
sume s on companies, as misbeha io could nega i ely a ec
a company’s business.
2.2. CSR epo ing as a communica ions ool
To mi iga e hese p essu es, companies use CSR epo -
ing o con ey legi imacy o hei s akeholde s (Cho e al.,
2009). Legi imacy in his sense is de ined as “a gene alized
pe cep ion o assump ion ha an en i ies ac ions a e desi -
able, p ope , o app op ia e wi hin some socially cons uc ed
sys em o no ms, alues, belie s, and de ini ions” (Suchman,
1995, p. 574), and hus se ing as a c i ical pilla in co po a e
communica ion s a egies and he main enance o a posi i e
co po a e image. This concep can also be obse ed in adi-
ional inancial epo ing. Da ison & Wa en highligh ha
“Annual epo s a e almos uni e sally used as a means o
molding co po a e iden i y and epu a ion” (Da ison & Wa -
en, 2009, p. 846).
Fu he mo e, E ickson e al. (2011) examine how com-
panies use epo ing o s a egically manage in o ma ion, es-
pecially du ing imes o c isis o when acing ad e se e en s.
T. Janßen /Junio Managemen Science 10(4) (2025) 1009-1027 1011
They ound ha epo s can be used o shape s akeholde
pe cep ion and educe po en ial ad e se eac ions, he eby
p ese ing legi imacy and a posi i e co po a e image. Volun-
a y epo ing can also se e as a means o dec ease ex e nal
cos s o p essu e om s akeholde s, allowing companies o
p oac i ely de end and es o e hei image in he case o mis-
beha io o w ongdoing (Ta e e al., 2009). In essence, CSR
epo ing is a s a egic ool o enhance legi imacy and co po-
a e image by demons a ing i s commi men o sus ainabil-
i y, add essing expec a ions, and po en ially al e ing pe cep-
ions o i s pe o mance and e hical s ance (Boi al & Hen i,
2017). F om his pe spec i e, i can be explained why com-
panies ha don’ pe o m well s ill ind i essen ial o epo
on CSR. By epo ing, hey seek o main ain us and c edi-
bili y wi h hei s akeholde s (Nielsen & Thomsen, 2018).
As men ioned in he p eceding chap e , in es o s’ g ow-
ing ocus on CSR can be seen as ano he signi ican ac o o
companies o issue CSR epo s. As mo e capi al is in es ed
in ESG-con o ming asse s, hese in es o s consequen ly de-
mand in o ma ion abou his ealm. No being anspa en
by no disclosing CSR in o ma ion would lead in es o s o
neglec he company when sea ching o ESG-con o ming in-
es men s. Hence, CSR epo ing also becomes mo e c i ical
om a classical sha eholde pe spec i e.
The e o e, he inc ease in public awa eness and p essu e
is also e lec ed in he numbe o companies ha publish
CSR epo s seeking o legi imize hei ac i i ies. The sha e
o companies disclosing CSR in o ma ion ose be ween 2002
and 2015 om 10% o 80% in he EU ( ega ding companies
om he Eu oS oxx600) and om 10% o 60% in he US
( ega ding companies in he S&P 500) (S olowy & Paugam,
2018). Ano he s udy ound ha be ween he ea ly 1990s
and 2016, he numbe o companies issuing some o m o
sus ainabili y epo ing inc eased om a ound 20 o almos
900 (Se a eim & Amel-Zadeh, 2017).
2.3. Di e ences om adi ional epo ing
Despi e hei simila i y in unc ion, adi ional ( inancial)
epo ing and CSR epo ing di e signi ican ly in many as-
pec s. In he ollowing sec ion, we will look a some key di -
e ences.
Fi s and mos appa en , ex e nali ies play a cen al ole
in CSR epo ing. The in o ma ion disclosed is no only abou
he company i sel bu also i s en i onmen al and socie al
impac . I could also ex end ou side adi ional bounda ies,
o example, “when a i m imposes child labo es ic ions on
i s supply chain” (Ch is ensen e al., 2021, p. 1186).
Fu he mo e, he eade ship o CSR epo s is po en ially
much la ge han ha o inancial epo ing. As discussed in
he p e ious chap e , he e a e mo e s akeholde s engaged in
he CSR dimension o a company, making i mo e challeng-
ing o add ess all he in e es s o hese di e se s akeholde s.
Mo eo e , some s akeholde s, e.g., cus ome s, a e less expe-
ienced in eading co po a e epo s and use he in o ma ion
o di e en pu poses han adi ional in es o s, mainly e al-
ua ing he company’s con ibu ion and adhe ence o hei
no ms and alues (Ch is ensen e al., 2021).
The di e si y o opics add essed by he epo s is ano he
di e en ia ing ac o . As he concep s o CSR and sus ain-
abili y a e no clea ly de ined, hey include a wide ange
o ESG opics, ac ions, and policies. They also di e sub-
s an ially be ween di e en indus ies, na ions, and compa-
nies (Ch is ensen e al., 2021). Each company migh im-
plemen di e en solu ions o app oaches o add ess CSR-
ela ed opics. While many o hese ac i i ies migh esul
in echnical o measu able ou pu s (like he o al amoun
o CO2 Emissions), hey canno be measu ed in adi ional
mone a y e ms. Also, he e a e s ill o he ac ions o ac i i-
ies ha canno be quan i ied a all. CSR is o en seen as a
s a egic endea o ha p io i izes long- e m ad an ages o a
company o e immedia e inancial gains (Benabou & Ti ole,
2010); hese long- e m ac i i ies a e o en quali a i e (Ch is-
ensen e al., 2021). This si ua ion makes i di icul o apply
s anda d epo ing p ac ices like double-en y bookkeeping,
esul ing in a a ie y o di e en o ma s, making compa -
isons and s anda diza ion di icul (Ch is ensen e al., 2021;
Ki zmuelle & Shimshack, 2012).
Due o he di e si y o use s and opics, a b oad spec-
um o in e es s and p e e ences o igina es in e nally and
ex e nally om he company. As a esul , CSR epo s se e
a a ie y o unc ions ha migh di e o e ime, o exam-
ple, when social ac i is s a ge he i m o when exogenous
shocks, such as a na u al ca as ophe, occu (Ba on, 2001;
Bone i e al., 2023; Ch is ensen e al., 2021). When a -
ge ed by a g oup (o when concei ed o be a ge ed), he
company migh s a egically use he epo o (p oac i ely)
de end i sel agains (po en ial) claims.
S emming om hese di e ences, he key issues a e he
he e ogenei y o he published epo s and he ha dly quan-
i iable me ics. This he e ogenei y leads o di icul ies in
compa ing companies agains each o he (Ch is ensen e al.,
2021), which is one o he cen al goals ha eade s, es-
pecially in es o s, y o accomplish when e alua ing CSR
epo s. This is highligh ed by a BlackRock s udy whe e In-
es o s s a ed ha he bigges challenge o hem o adop
sus ainable in es ing is he “Poo quali y o a ailabili y o
ESG da a and analy ics” (53% o esponden s) and “Poo
quali y o sus ainabili y in es men epo ing”(33% o e-
sponden s) (BlackRock, 2020). Se a eim and Amel-Zadeh
(2017) also suppo his na a i e; hey poin ou ha he
bigges challenge o using ESG in o ma ion is he lack o
compa abili y. Fu he , hey also sugges ha he quan i i-
ca ion and eliabili y o he epo ed in o ma ion a e majo
challenges.
Se e al de elopmen s ha e ied o s anda dize epo -
ing p ac ices and es ablish g ea e compa abili y o add ess
his issue, mainly h ough he de elopmen o guidelines and
epo ing manda es. These a e expec ed o imp o e homo-
genei y, a leas wi hin an indus y (Ch is ensen e al., 2021).
2.4. Homogeniza ion app oaches
As he need o de ailed and homogenous CSR disclo-
su es inc eases, se e al o ganiza ions ha e ini ia ed e o s o
T. Janßen /Junio Managemen Science 10(4) (2025) 1009-10271012
es ablish olun a y epo ing s anda ds o enhance he con-
sis ency o epo s published. Fo example, he Sus ainabil-
i y Accoun ing S anda ds Boa d (SASB) de eloped “indus y-
speci ic disclosu e s anda ds ac oss inancially ma e ial en i-
onmen al, social, and go e nance opics” (Ch is ensen e al.,
2021, p. 1177). This app oach ocuses mo e on he in eg a-
ion o ESG epo ing in o inancial epo ing. This s anda d
aims o guide companies in 77 di e en indus ies, allowing
hem o iden i y, manage, and communica e sus ainabili y in-
o ma ion o in es o s. Ini ially de eloped o SEC2 ilings
and ailo ed o hese speci ic equi emen s, he guidelines a e
also used globally.
Simila ly, he Global Repo ing Ini ia i e (GRI) aids com-
panies in communica ing hei impac on c i ical sus ainabil-
i y issues by de eloping global s anda ds o sus ainabili y
epo ing. The di e ence he e is ha he ange o opics
disclosed is mo e ex ensi e, as hese guidelines a e no e-
s ic ed o only inancially ma e ial in o ma ion bu also en-
compass ex e nali ies and ake a mo e gene al app oach o
he sus ainabili y impac o companies. I is also designed
wi h a globally applicable usage in mind. These e o s e-
lec a mo emen owa ds g ea e uni o mi y wi hin indus-
ies, as hey p o ide amewo ks ha companies can ollow
(Ch is ensen e al., 2021).
Some coun ies and egions ha e decided o ake CSR
epo ing p ac ices and hei impo ance e en u he . The
EU also did so wi h i s Non-Financial Repo ing Di ec i e
(NFRD), which he Eu opean Pa liamen passed on 15 Ap il
2014 (Di ec i e 2014/95). This di ec i e manda es publicly
lis ed companies wi h o e 500 employees and ei he asse s
exceeding 20 million EUR o sales o e 40 million EUR pe
yea o p oduce non- inancial (CSR) epo s, s a ing wi h he
iscal yea 2017. Hence, he i s epo s unde his di ec i e
we e published in 2018 (Fiech e e al., 2022). The NFRD
op s o a double ma e iali y pe spec i e, simila o he GRI,
no only conside ing inancially ma e ial in o ma ion bu also
ex e nali ies and how hei ac i i ies a ec socie y and he
en i onmen (Ch is ensen e al., 2021). This encompasses
policies, main isks, and ou comes conce ning en i onmen-
al ma e s, social and employee aspec s, human igh s, an i-
co up ion, and di e si y on he boa d o di ec o s (Fiech e
e al., 2022). The goal o he di ec i e is o inc ease he ans-
pa ency o he epo ed in o ma ion ac oss all membe s a es
and sec o s (Di ec i e 2014/95, eci al 1), as well as se ing
an incen i e o companies o engage in ESG- ela ed ac i i-
ies (Di ec i e 2014/95, eci al 3) as neglec ing he achie e-
men o ESG ela ed goals would be ecognized and “pun-
ished” by in es o s and socie y. Addi ionally, he “disclosu e
o non- inancial in o ma ion helps he measu ing, moni o -
ing, and managing o unde akings’ pe o mance and hei
impac on socie y” (Di ec i e 2014/95, eci al 3).
Despi e hese ad ancemen s, bo h guidelines and man-
da es lea e conside able oom o indi idual in e p e a ion
and aming. Volun a y guidelines allow companies o selec-
2Uni ed S a es Secu i ies and Exchange Commission
i ely disclose ce ain aspec s, ailo ing hei epo s o e-
lec a o able cha ac e is ics while po en ially omi ing less
a o able ones. Simila ly, e en manda o y di ec i es can only
pa ially s anda dize epo ing due o he inhe en ly quali a-
i e na u e o some da a and he a ied quan i a i e assess-
men s ha di e en companies migh use. Also, i has been
shown ha manda es, in gene al, inc ease he use o boile -
pla e language and, hence, loss o epo eadabili y (Dye
e al., 2017). This migh also be ue o CSR epo ing and
pe haps be e en mo e p e alen as he con en is, by de aul ,
mo e quali a i e, which could blu ac s and alk e en mo e.
The e is also a lack o de ailed EU documen a ion o guid-
ance on en o cemen a he coun y le el, which can lead o
inconsis encies in how hese egula ions a e applied (Fiech e
e al., 2022).
This chap e highligh ed why companies epo on CSR-
ela ed opics, he complexi y o CSR epo ing, and ap-
p oaches aken o ha monize epo ing p ac ices. While
guidelines and manda es push companies owa ds g ea e
anspa ency and accoun abili y, hey also equi e ca e ul
conside a ion o ensu e ha epo s genuinely e lec he
company’s impac a he han me ely con o ming o he le -
e o he law. In he ollowing chap e , we will look a ends
and de e minan s conside ing he ex en and con en o CSR
epo s and he possible e ec o he NFRD be o e we ake a
close look a he ole o isuals, in pa icula pic u es, in CSR
epo s. This will lay he ounda ion o my (pic u e) con en
analysis o ESG epo s, ocusing on he EU and Swi ze land.
3. T ends in CSR Repo ing
Building on he ounda ion se in he p e ious chap e ,
we will di e deepe in o he p ac ical implemen a ion o CSR
epo ing and he ac o s ha shape epo ing ou comes. As
we ha e seen, he cen al cha ac e is ic o hese epo s is
he inhe en he e ogenei y in wha companies epo olun-
a ily. Repo ing s anda ds in end o educe he p oblem,
bu as Ch is ensen e al. (2021) highligh , companies’ ad-
he ence o hese guidelines a ies signi ican ly. This is be-
cause, apa om s anda ds, se e al ac o s in luence epo -
ing p ac ices and ou comes. These include manage ial in-
cen i es and o he ins i u ional a angemen s. This implies
ha epo ing also ollows i s own agenda and aligns wi h
he p e ious heo e ical discussions. Also, wi h he in o-
duc ion o manda es aimed a enhancing uni o mi y, he e
a e s ill challenges in aligning s akeholde expec a ions o
anspa ency wi h ac ual business p ac ices. These manda es
o en inc ease boile pla e language, as obse ed in a ious
egula o y con ex s like he K-10 disclosu es in he US. This
inc ease in gene ic con en can educe he speci ici y and
eadabili y o epo s, po en ially dilu ing he e ec i eness o
hese disclosu es in con eying clea and ac ionable in o ma-
ion and obscu ing speci ic da a behind gene ic desc ip ions
(Ch is ensen e al., 2021; Dye e al., 2017).
Fu he mo e, i ms end o ocus p edominan ly on pos-
i i e achie emen s in hei epo s while ailing o c i ically
engage wi h nega i e aspec s, which can mislead he eade
T. Janßen /Junio Managemen Science 10(4) (2025) 1009-1027 1013
abou he ue s a e o CSR pe o mance (Boi al & Hen i,
2017). O he esea che s made simila indings, poin ing
ou ha CSR communica ion can manipula e he eade , be-
ing decoupled om ac ions and a emp ing o bene i om
a sus ainable image wi hou changing he company’s ac ual
p ac ices (Siano e al., 2017). This aligns wi h my heo e -
ical ounda ion ha Legi imacy Theo y and ying o c ea e
a posi i e pic u e o he company a e key d i e s in CSR e-
po ing.
3.1. De e minan s and T ends o CSR Repo ing
When looking a he a iabili y in CSR epo ing and ex-
plo ing i s pa e ns, se e al de e minan s can be ound ha
in luence epo ing beha io s. These include company size,
indus y cha ac e is ics, co po a e pe o mance, and li iga-
ion isk. These ac o s shape companies’ s a egic epo -
ing beha io s and signi ican ly in luence how i ms app oach
hei CSR disclosu es.
Company Pe o mance: Companies showing a s ong
ESG pe o mance a e o en mo e anspa en and end o dis-
close mo e nume ical da a and ac ual in o ma ion ha p o-
ides a clea iew o hei sus ainabili y ini ia i es. The eby
p o iding a posi i e bu accu a e image o hei pe o mance.
Simila ly, a s udy o he disclosu e o en i onmen ally sen-
si i e companies in China ound ha companies wi h be e
en i onmen al pe o mance epo mo e on en i onmen al
ma e s and disclose mo e quan i a i e in o ma ion (He &
Lo us, 2014). In con as , companies wi h ela i ely weake
pe o mance may ely mo e on quali a i e s a emen s and
use he epo s o s a egically p ojec a posi i e image,
he eby obscu ing he eade (Ga cía-Sánchez & A aújo-
Be na do, 2020).
In luence o Company Size: The size o a company has
a signi ican e ec on epo ing p ac ices. I is, in ac , one o
he de e minan s ha ha e been consis en ly ound o ha e
a posi i e in luence on sus ainabili y epo ing. La ge i ms
end o ha e highe isibili y and a e assessed mo e ho -
oughly by s akeholde s; hus, hey a e incen i ized o ex en-
si ely co e CSR opics. Ch is ensen e al. (2021) highligh
ha he b ead h o he issues add essed in CSR epo s di-
ec ly co ela es wi h company size, wi h la ge i ms likely
acing mo e media exposu e and s akeholde p essu e. O he
pape s suppo his inding by poin ing ou ha he size o
a company is a c i ical in e nal de e minan ha posi i ely
in luences sus ainabili y epo ing, p ima ily because la ge
companies a e mo e isible and, hence, mo e ho oughly in-
es iga ed (Hahn & Kühnen, 2013).
Financial pe o mance: Some esea che s also ound a
posi i e co ela ion be ween a company’s p o i abili y and
en i onmen al disclosu es, speci ically sugges ing ha mo e
p o i able companies may ha e mo e esou ces o alloca e o-
wa d disclosing en i onmen al in o ma ion. (Game schlag
e al., 2011)
Indus y Cha ac e is ics and Li iga ion Risk: The in-
dus y in which a company ope a es and he associa ed isks
o ha indus y also in luence epo ing beha io s. High-
pollu ing indus ies (en i onmen ally sensi i e indus ies)
epo mo e ex ensi ely on CSR ac i i ies and in o ma ion
(Game schlag e al., 2011), possibly o mi iga e po en ial
backlash o egula o y con ols. Simila ly, G ougiou e al.
(2016) ound ha companies in socially o en i onmen ally
sensi i e indus ies – o en e e ed o as “sin” indus ies –
end o epo mo e on indica o s ha show socially esponsi-
ble business p ac ices. They a gue ha i ms wi h signi ican
li iga ion isks a e mo e likely o use CSR epo ing o build a
esponsible co po a e image, po en ially o mi iga e backlash
and minimize he di ec o indi ec cos s o legal challenges.
Ano he s udy ound ha companies in con o e sial sec o s
(he e e e ing o socially sensi i e indus ies) ocus hei
epo s mo e on communi y and people’s achie emen s o
enhance hei legi imacy despi e he nega i e impac s o
hei p oduc s o se ices (By d e al., 2017).
Selec i e epo ing and Manipula ion: As indica ed a
he beginning o his chap e , Guidelines, like he GRI, de-
eloped o inc ease anspa ency in CSR epo ing, s ill mus
be c i ically assessed when i comes o hei applica ion. A
S udy ocusing on companies ha ecei ed he highes a -
ing om he GRI ini ia i e (A o A+ a ing)3disco e ed ha
he e was s ill a signi ican unde epo ing o ad e se e en s,
iola ing GRI p inciples o balance, comple eness, and ans-
pa ency. This unde epo ing and he endency o emphasize
posi i e achie emen s o e nega i e ones sugges ha some
i ms a emp o main ain a posi i e co po a e image a he
han p o iding genuine anspa ency. Due o he inconsis en-
cies in epo ed da a and he quali a i e na u e o many GRI
indica o s, he challenges in he compa abili y o CSR epo s
ha e o be ecognized e en in he ace o Guidelines (Boi al,
2013).
In his subchap e , we ha e examined he de e minan s
in luencing CSR epo ing p ac ices. Mo ing o wa d, we ex-
plo e he speci ic impac s o epo ing manda es using he
NFRD as an example, explo ing how hey in luence co po-
a e beha io and he in eg i y o CSR disclosu es.
3.2. E ec s o a epo ing manda e (NFRD)
The implemen a ion o he NFRD has ma ked a signi -
ican shi in ESG epo ing s anda ds and p ac ices in he
EU, wi h he aim o enhance anspa ency and compa abil-
i y ac oss i ms. The e a e se e al indings ha indica e i s
(pa ly) success. A s udy among Spanish i ms om 2013-
2018, o example, ound ha he e has been a signi ican
inc ease in companies adhe ing o he GRI s anda ds, indi-
ca ing an imp o emen in anspa ency di ec ly in luenced
by he di ec i e. (Ga cía-Sánchez & A aújo-Be na do, 2020)
A mo e de ailed analysis ac oss a b oade spec um o EU
companies by Fiech e e al. (2022), s udying he e ec s o
he NFRD, no ed an inc ease in epo ing anspa ency begin-
ning in 2014, he yea he esolu ion was passed. Wi h hei
3Le els A+and A a e he mos demanding and comple e le els o use o
his epo ing sys em. Fo ins ance, Le el A assumes ha he sus ainable
de elopmen epo conside s each co e and sec o supplemen indica o .
Le el A+assumes, in addi ion, ha he epo is audi ed by a hi d pa y
(GRI, 2006; Boi al and Hen i, 2017, p. 5)
T. Janßen /Junio Managemen Science 10(4) (2025) 1009-10271014
esea ch ocusing on he yea s 2011 o 2018, hey highligh ed
ha no signi ican inc ease in anspa ency was obse able
be o e 2013, sugges ing ha his e ec is also a ibu able o
he NFRD ca alyzing hese imp o emen s.
The dis ibu ion o he manda e’s impac was no con-
sis en ac oss all dimensions o CSR. Al hough he e was a
no able ise in CSR ac i i y, as measu ed by he combined
social and en i onmen al sco es, he e ec was only s a is-
ically signi ican o he social sco e when analyzed inde-
penden ly (Fiech e e al., 2022). The limi ed impac on
he en i onmen al sco e could be a ibu ed o longe imple-
men a ion imes and he long- ange na u e o en i onmen al
ini ia i es. Addi ionally, he mos subs an ial enhancemen s
we e obse ed in companies ca ego ized as “high exposu e”
– hose wi h lowe le els o anspa ency and ac i i y p io
o he manda e (lowe han he median o obse ed coun-
ies be o e 2014). These i ms also showed imp o emen s
in hei en i onmen al ac i i ies by 2018, indica ing ha he
NFRD has been pa icula ly e ec i e in enhancing epo ing
and CSR p ac ices among i ms ha we e ini ially less com-
plian (Fiech e e al., 2022).
Addi ionally, he di ec i e seems o posi i ely a ec in-
es men s in o CSR in as uc u e. Companies wi h a highe
exposu e p io o he NFRD engaged mo e in es ablishing an
in e nal CSR Commi ee and launching social ini ia i es, o -
en ollowed by an inc ease in he ESG Sco e (Fiech e e al.,
2022).
Despi e hese posi i e changes, hey acknowledge ha
he e is no de ailed EU documen a ion o guidance on he
en o cemen o hese s anda ds a he coun y le el, which
could impac he consis ency and e ec i eness o he NFRD’s
implemen a ion ac oss membe s a es (Fiech e e al., 2022).
3.3. P elimina y Re lec ions on he Role o he Visuals in ESG
Repo s
While he NFRD has imp o ed anspa ency in CSR e-
po ing, he usage o pic u es was no included in he analy-
sis. Howe e , he use o images is in e es ing. S udies high-
ligh ed ha pic u es in he con ex o CSR epo s could be
used o g eenwashing (Chong e al., 2019) by o e shadow-
ing less a o able en i onmen al o social impac s, po en-
ially leading s akeholde s o misin e p e a company’s ac ual
p ac ices and pe o mances.
Mo eo e , o he s udies discuss how boile pla e lan-
guage, which inc eases in he ace o manda es, can be used
o g eenwashing in CSR epo s (Ch is ensen e al., 2021).
While hey do no speci ically men ion isuals, images can
ampli y his e ec , and i nonin o ma i e, could be seen as
a ype o boile pla e i sel . Visuals can emphasize posi i e
aspec s o downplay nega i e ones, sub ly shaping s ake-
holde pe cep ions and po en ially enhancing he impac o
boile pla e con en .
In addi ion, when i comes o g eenwashing, i has been
shown ha passi e g eenwashing, in con as o ac i e g een-
washing4, is pe cei ed as less se e e and induces smalle neg-
4Ac i e g eenwashing e e s o companies in en ionally c ea ing alse in-
a i e e ec s. In addi ion, communica ions compa ed o ac-
ion g eenwashing has been iden i ied o ha e a smalle neg-
a i e e ec in he scope o passi e g eenwashing. Thus, pas-
si e g eenwashing h ough communica ions has been ound
o be he leas se e e o m o g eenwashing (Ga i e al.,
2021). The usage o pic u es is mos likely o be a ibu ed o
ha ca ego y and is pe haps e en mo e sub le, making i im-
po an o unde s and he e ec s and in en ions wi h which
pic u es a e used.
We p elimina ily highligh ed he ole isuals migh play
in CSR epo ing, indica ing ha while hey may enhance he
cla i y o epo s, hey could also be used o c ea e a a o able
( alse) image. The nex chap e will discuss he psychological
and communica i e e ec s o isuals, examining how hey
in luence pe cep ion. Following ha , we will e iew exis ing
esea ch on he uses o isuals in CSR epo s o es ablish a
ounda ion o my analysis.
4. The Signi icance o Visuals in Pe cep ion
4.1. Impac o Visuals on Pe cep ion and Cogni ion
The human b ain p ocesses isuals in e olu iona ily olde
a eas o he b ain (Ha pe , 2002), indica ing hei impo -
ance in shaping ou pe cep ion and cogni ion. We also pay
mo e a en ion o isuals han we do o ex (T e sky, 1974)
- o en a a subconscious le el whe e hei sub le in luences
bias ou pe cep ion wi hou explici awa eness (Posne e al.,
1976). In gene al, isuals help in memo izing and ha e a
mo e powe ul place in memo y (Vasue & Howe, 1989), bu
when we p ocess and ocus on oo many isuals, i can po en-
ially lead o cogni i e “o e loading” and, hus, dis ac om
key in o ma ion (T e sky, 1974).
Visuals can play a c ucial ole in o ming aes he ic ap-
peal, which in u n in luences how indi iduals pe cei e in o -
ma ion. Resea ch has demons a ed ha aes he ics ha e he
powe o gene a e a posi i e imp ession, making hem highly
aluable in he ealm o co po a e communica ion (Legen-
d e e al., 2020). A s udy conduc ed by Townsend and Shu
(2010), p o ides e idence o his phenomenon. In he s udy,
in es o s we e asked wi h e alua ing he alue o a ic i-
ious company based on i s annual epo , wi h he le el o
aes he ic elemen s (such as he quan i y, size, and colo o
pic u es/images) a ying ac oss di e en g oups. They ob-
se ed ha aes he ics impac ed he judgmen o in es o s
abou he company o he same ex en as inancial me ics,
e ealing ha e en expe ienced s akeholde s can be in lu-
enced by he design choices o annual epo s. Ano he in-
e es ing obse a ion is ha when in es o s ha e been made
awa e ha aes he ics could in luence hei alua ion5, he e -
ec diminishes (Townsend & Shu, 2010), s eng hening he
a gumen ha he in luence o isuals is o en subconscious.
o ma ion, while passi e g eenwashing e e s o co e ing up in o ma ion
(Ga i e al., 2021)
5By asking he in es o s be o ehand how much hey hink ha as he ics
in luence hei decsison oge he wi h o he ac o s like p o i abili y e c.
T. Janßen /Junio Managemen Science 10(4) (2025) 1009-1027 1015
These indings a e c ucial o unde s anding how compa-
nies migh le e age pho og aphs o enhance hei co po a e
image.
4.2. S a egic Role o Visuals in Co po a e Na a i e Con-
s uc ion
The e a e wo p ima y pu poses ha isuals can se e in
epo ing. On he one hand, hey can enhance w i en in-
o ma ion by isually ep esen ing “ eali y” (desc ibe). On
he o he hand, hey ha e he powe o c ea e and con ey
hei own message (cons uc ) (Da ison, 2015). Visuals can
cons uc complex na a i es, which a e i al in shaping an
o ganiza ion’s iden i y and in luencing how s akeholde s pe -
cei e i (P es on e al., 1996). Howe e , because isuals ha e
he capaci y o c ea e and manipula e eali y, i is impo an
o app oach hem c i ically.
Da ison (2010) discusses an example o he complex mes-
sages anspo ed h ough images in he con ex o Business
po ai s. They se e o p o ide a human ace o he o ga-
niza ion, he eby es ablishing a connec ion wi h he com-
pany and inc easing in es o con idence and cus ome loy-
al y. They a e also used o ”b and” he op execu i es and,
he e o e, he company as eliable, compe en , and e hical
(Da ison, 2010). Repea ing a heme h oughou a documen
enhances he abili y o isuals o cons uc . As hey become
mo e amilia , ecogni ion and e en ion a e inc eased, a-
o ing he pe cep ion o he company (and hei p ac ices,
depending on he heme) (Da ison, 2008).
Due o hei po en abili y, he use o isuals is no wi h-
ou conce ns. Cho e al. (2009) exp ess ha isuals could
po en ially mislead s akeholde s by o e shadowing o dis-
ac ing om he ac ual con en . Thei s udy showed ha
adding pic u es o a ex -only ep esen a ion on co po a e
websi es leads o a signi ican inc ease in pe cei ed social
esponsibili y wi hou al e ing he ac ual in o ma ion being
ep esen ed. This change in pe cep ion, achie ed solely by
adding pic u es, could gi e ise o po en ial g eenwashing.
This e ec o pic u es can likely also be ex ended o com-
munica ion ia CSR epo s. Wi h his, we now con inue by
looking a wha has al eady been ound ou abou he usage
o pic u es in ESG epo s.
5. Insigh s in o he Use o Visuals in CSR Repo s
The p e ious chap e s gi e easons o belie e ha isu-
als a e no only in eg a ed in o CSR epo s o aes he ic o
explana o y easons bu also se e a c ucial ole in he p epa-
a ion o epo s (In e nizzi e al., 2022). O e he las ew
yea s, he e has been an inc ease in he numbe o pic u es
used in CSR epo s (Chong e al., 2019). Despi e hese de-
elopmen s, his a ea has only been ouched upon by a ew
s udies (Da ison, 2015).
5.1. S a egic Use o Visual Con en in CSR Repo s
F om he iewpoin o legi imacy heo y, o ganiza ions
use CSR epo s o s eng hen hei public image and ma -
ke posi ion. Images seem o be a cen al pa o his s a -
egy. Rämö (2011) highligh s ha pho og aphs in CSR epo s
se e o suppo he na a i e o esponsibili y and e hical be-
ha io , showcasing he commi men o socie al alues and in-
eg i y. Simila ly, In e nizzi e al. (2022) conclude ha hese
images pain a pic u e o compe ence and achie emen in he
economic and CSR ealm.
This a emp o po ay compe ence and achie emen
leads o a signi ican bias owa ds p edominan ly using pos-
i i e images and c ea ing an idealized iew o he company,
downplaying nega i e e ec s (Boi al, 2013) and, as indi-
ca ed in he p e ious chap e s, could mislead he eade
abou he ac ual company’s pe o mance. Pa icula ly s ik-
ing is ha he images almos occupy he same amoun o
space as he ex ual con en in CSR disclosu es (Boi al,
2013). This bias o isuals and he la ge space hey occupy
can dis ance he epo om ac ual sus ainabili y pe o -
mance, po en ially enhancing he company’s image a he
expense o anspa ency. Fu he mo e, i has been obse ed
ha less sus ainabili y-d i en companies addi ionally use
gene ic, non-speci ic images, symbolically aligning hem-
sel es wi h sus ainabili y hemes and enhancing hei image.
This app oach can be seen as a o m o g eenwashing, as he
po ayed and ac ual sus ainabili y pe o mance and a i ude
don’ align (H asky, 2012).
5.2. Analysis o Visuals in CSR Repo s
Rega ding he con en o he pic u es being used, he so-
cial dimension is he mos equen ly depic ed, ollowed by
he en i onmen al. No ably, In e nizzi e al. (2022) ound
ha on a e age, 2.38 pho os pe epo depic an en i on-
men al heme and 7.47 a social one. Ano he inding om
hei s udy is ha ega ding eadabili y and pe cei ed legi i-
macy, he op imal numbe o pic u es pe page is abou one
pe page.
Ano he s udy ocuses on he size and colo o he pic-
u es being used while di e en ia ing be ween s anda dized
( epo on all GRI indica o s) and non-s anda dized (do no
epo on all GRI indica o s) epo s. Less s anda dized e-
po s end o use la ge , mo e colo ul, and o e all, mo e
images, po en ially compensa ing o he lack o subs an i e
in o ma ion (Ga cía-Sánchez & A aújo-Be na do, 2020), u -
he complemen ing he iewpoin ha companies use im-
ages s a egically o o m imp essions.
Following an indus y-speci ic ocus, Chong e al. (2019)
ind ha companies om “sensi i e” indus ies in New
Zealand, such as ene gy, mining and cons uc ion, p ope y,
ag icul u e, ishing, and anspo , use mo e pic u es de-
pic ing an en i onmen al heme, again possibly o enhance
he pe cep ion o hei en i onmen al pe o mance and o
add ess s akeholde expec a ions.
Toge he wi h he p eceding chap e s, hese insigh s high-
ligh he impo ance o unde s anding wha ac o s de e -
mine companies’ use o images and he a ionale behind his
beha io .
In he ollowing, I will conc e ize he esea ch ques ion
and look a he da a used o he analysis, he me hodology
T. Janßen /Junio Managemen Science 10(4) (2025) 1009-10271016
applied, and he esea ch design be o e e alua ing he e-
sul s.
6. Me hodology and Da a
6.1. Da a Collec ion
Fo he analysis, I ga he ed wo ypes o da a. One is he
in o ma ion abou he pic u es in CSR epo s, which I will
call “p ima y da a” in he ollowing, and he o he is supple-
men a y da a encompassing company-speci ic me ics on a
yea ly basis (e,g., abou he indus y, inancial me ics, ESG-
ela ed in o ma ion, and company cha ac e is ics), his will
be called “seconda y da a”.
The companies I ha e ini ially conside ed include all com-
panies om he STOXX Eu ope 600 (EU companies) ha
we e pa o he index in all yea s be ween 2011 and 2020,
and all companies om he index SMI Expanded (Swiss com-
panies)
To ga he he p ima y da a in a o ma sui ed o analy-
sis, I used he Google Cloud Vision AI (Google, 2024) o i s
de ec each pic u e om he epo s and hen au oma ically
c ea e a lis o labels (in mos cases consis ing o 10 unique
labels) ha desc ibe i s con en s. The in o ma ion ga he ed
was hen s o ed on a “company-yea -pic u e le el”. Wha is
o be no ed is ha I only conside ed sus ainabili y epo s
published sepa a ely om he annual inancial epo o be
able o dis inguish be ween inancial and ESG in o ma ion
clea ly and au oma ically.
The seconda y da a was e ie ed om he LSEG Wo kspace
( o me Re ini i ), comp ising a b oad ange o me ics o
my desi ed inancial and ESG dimensions. I u he en iched
his (seconda y) da a by adding dummy a iables indica -
ing whe he a company ope a es in an en i onmen ally o
socially sensi i e indus y and is ope a ing close o i s cus-
ome s. This classi ica ion was done a he indus y le el
(ICB Codes). Indus ies conside ed en i onmen ally sensi-
i e we e chosen based on he classi ica ion o B anco and
Rod igues (2008) bu adjus ed o he ICB Codes and adjus ed
o include ai lines. They encompass he ollowing: Ai lines,
Cons uc ion and Ma e ial, Oil, Gas and Coal, Elec ici y,
Gas, Wa e , and Mul i u ili ies, and Basic Ma e ials. Simila
socially sensi i e indus ies a e based on he classi ica ion
om B amme and Milling on (2005), again adjus ed o he
ICB Codes, and encompass Pha maceu icals, Tobacco, Alco-
hol and So d inks, De ense, and Basic Ma e ials. Indus ies
ope a ing close o hei cus ome s a e again based on he
classi ica ion used by B anco (B anco & Rod igues, 2008),
adjus ed o he ICB Code classi ica ion, and addi ionally con-
side ing media, hey encompass Telecommunica ions, Con-
sume P oduc s and se ices, Media, Re aile s, Consume
S aples, Elec ici y, Gas, Wa e , and Mul i u ili ies.
A comple e o e iew o he conside ed me ics, hei ex-
plana ion, and hei a ailabili y can be ound in Appendix A.
6.2. Pic u e Classi ica ion
I p opose a unique app oach ha is di e en om man-
ual con en analysis, as I i s le e age he Cloud Vision AI
(Google, 2024) o au oma ically de ec and desc ibe he
images and hen use a Na u al Language P ocessing (NLP)
model o embed hese desc ip ions in a ec o space sui -
able o clus e ing me hods, which a e hen applied. This
app oach enabled me o analyze a o al o 11.455 pic u es.
The lis o labels desc ibing each pic u e o e e y com-
pany e e y yea , c ea ed by he Cloud Vision AI, cons i u es
he basis o my pic u e classi ica ion app oach. I ea ed his
lis as a sen ence o he NLP o be able o cap u e he seman-
ic meaning o each label in he con ex in which i is being
used. The model ha was op ed o is MPne , which is buil
upon he a chi ec u e o BERT and published in coope a ion
wi h Mic oso Resea ch. I chose his model as i e ec i ely
add esses BERT’s limi a ions in handling oken dependencies
and posi ion in o ma ion, leading o signi ican ly imp o ed
pe o mance on a a ie y o language unde s anding asks
(Song e al., 2020). Using MPne , I c ea ed wo d embed-
dings o each Label Lis ; a e some you s, I iden i ied ha
a educ ion in he dimensionali y was necessa y o be e cap-
u e he seman ic simila i ies be ween he embedded Lis s.
He e, I iden i ied ha a P incipal Componen Analysis (PCA)
educ ion (Wang, 2019) o 0,7 deli e ed he bes esul s in
educing he high dimensionali y while keeping enough in-
o ma ion o p ope ly dis inguish hei seman ic meanings.
To c ea e clus e s based on hese embeddings, I used k-
means clus e ing. To ind he igh numbe o clus e s, I ol-
lowed wo s eps. In he i s s ep, I used he Elbow Me hod
(Humai a & Rasyidah, 2018) o de ine he ange whe e he
op imal numbe o clus e s lies. The analysis showed ha he
in lec ion poin is be ween 5 and 30. In he nex s ep, I only
conside ed his ange, so I looked a he Silhoue e Sco e and
Da ies Bouldin Sco e (Yilmaz e al., 2007) o each k in ha
speci ied ange as a measu emen o clus e quali y. I iden-
i ied he h ee clus e s wi h he bes esul s based on hese
measu emen sco es and u he assessed hem h ough sam-
ple es ing. A e his assessmen , I op ed o 15 ini ial clus-
e s6. The sample es ing ga e insigh s in o u he manual
adjus men s ha could be made ia he consolida ion o clus-
e s and ealloca ion h ough keywo ds. The consolida ion
ook place a e he ealloca ion and can be seen in Figu e 1,
oge he wi h he ini ial and inal clus e s and hei espec i e
names. The ealloca ion was done in h ee s eps.
Fi s , I c ea ed a new Clus e con aining images showing
enewable ene gy gene a ion. I iden i ied ha i a label in
he label lis con ains he wo ds “wind” o “sola ” while no
being in he de ined exclusion lis 7, i can be alloca ed o his
new ca ego y.
In he nex s ep, I iden i ied 4 clus e s (2, 8, 9, 13) con-
aining images depic ing modes o anspo a ion ha belong
o Clus e 9. He e, I also iden i ied a lis o keywo ds8and an
exclusion lis 9whe e i a label con ains a wo d in he keywo d
6Silhoue e Sco e: 0,162; Da is Bouldin Sco e: 1,865
7Exclusion Lis o Renewable Ene gies: [‘window’, ‘windshield’, ‘wind-
b eake ’, ‘windsc een’, ‘wind ins umen ’, ‘wind wa e’ ,‘wind’]
8Keywo ds T anspo a ion: [‘boa ’, ‘ anspo ’, ‘ai ’, ‘ship’, ‘ ail’, ‘ olling
s ock’, ‘na al’]
9Exclusion Lis T anspo a ion: [‘boa s and boa ing–equipmen and sup-
T. Janßen /Junio Managemen Science 10(4) (2025) 1009-1027 1023
Table 6: Di e ence in Di e ence Reg ession on Pho og aph by Page Coun
Va iable Model 1 Model 2
Coe icien S d. P-Value†Coe icien S d. P-Value†
In e cep 0.294 (0.049) [0.000]0.328 (0.053) [0.000]
T ea men 0.272∗∗∗ (0.054) [0.000]0.226∗∗∗ (0.060) [0.000]
Pos 0.192∗∗ (0.068) [0.011]0.151∗(0.076) [0.062]
T ea _X_Pos -0.185∗(0.090) [0.056]-0.106 (0.098) [0.299]
Any_Sensi i e -0.193 (0.105) [0.002]
T ea _X_Sensi i e 0.222 (0.139) [0.116]
Pos _X_Sensi i e 0.188 (0.168) [0.426]
T ea _X_Pos _X_Sensi i e -0.563∗(0.260) [0.064]
Adjus ed R20.070 0.077
∗∗∗ p<.001, ∗∗ p<.01, ∗p<.05 (when using non obus s anda d e o s)
†P- alues when using obus s anda d e o s
No es: N =407 company-yea obse a ions, S d =S anda d de ia ion o In e cep s, all con inuous a iables ha e been s anda dized be o e he eg ession
his e ec is s a is ically signi ican a con en ional P- alue
h esholds (0.039), i becomes ma ginally non-signi ican
(P- alue: 0.056) when adjus ed o obus s anda d e o s. I
addi ionally used obus s anda d e o s as es ing he model
indica ed po en ial he e oscedas ici y. Consequen ly, I can-
no de ini i ely dismiss he null hypo hesis o no ea men
e ec , hough he e idence s ongly hin s a a po en ial e -
ec . In Model 2, “T ea _X_Pos ” ep esen s he impac on
non-sensi i e EU companies and shows no signi ican e ec
unde any P- alue condi ions. Howe e , he iple in e -
ac ion e m, “T ea _X_Pos _Sensi i e,” which measu es he
impac on sensi i e indus ies wi hin he EU compa ed o
Swi ze land, ini ially appea s signi ican a s anda d P- alue
h esholds (0.031) bu , again, is ma ginally non-signi ican
(P- alue: 0.064) wi h obus s anda d e o s. Despi e be-
ing unable o ejec he null hypo heses in bo h models, I
obse e s ong e idence o a ea men e ec , pa icula ly
concen a ed among sensi i e companies.
Table 7: Di e ence o he numbe o pic u es pe epo page
be ween 2018 and 2020
Sensi i e
Companies
Non-
Sensi i e
Companies
To al
EU - 0.424 - 0.033 - 0.069
Swiss +0.265 +0.135 +0.157
To quan i y his e ec in desc ip i e e ms, Table 7shows
he a e age “Pho og aph by Page Coun ” di e ence be ween
2018 and 2020 in he EU and Swi ze land, u he di e en-
ia ed by sensi i i y. He e, we see he la ges di e ence be-
ween he EU and Swi ze land in Sensi i e companies. While
companies in he EU, on a e age, educed he numbe o
pic u es pe epo by 0.42, he companies in Swi ze land
inc eased he numbe o pic u es used by 0.27, esul ing in
a change in hei di e ence o 0.69. Compa ing his o he
a e ages o e he whole pe iod o each egion depic ed in
Table 2(0.576 o he EU and 0.400 o Swi ze land), his
desc ip i e app oach indica es ha he magni ude o he po-
en ial T ea men e ec is conside able on an absolu e le el.
8. Discussion and Limi a ions
8.1. Discussion
In his sec ion, I ela e my indings o he heo e ical
amewo k discussed ea lie and o e po en ial explana ions
o esul s ha a e no di ec ly suppo ed o con adic ed by
exis ing heo y. To ecap, I expec ed o see companies in sen-
si i e indus ies o wi h compa ably lowe ESG pe o mance
using pic u es mo e equen ly, possibly c ea ing a mo e a-
o able image by edi ec ing a en ion away om less posi-
i e in o ma ion. Fu he , I an icipa ed ha companies would
use a highe a io o pic u es depic ing a heme om he a ea
whe e hey ace g ea e sc u iny o ha e poo sus ainabili y
pe o mance. Finally, knowing ha he in oduc ion o he
NFRD inc eased epo ing anspa ency, I expec ed o see a
co esponding dec ease in he use o pic u es, in e p e ing
inc eased anspa ency as educed eliance on images.
8.1.1. F equency o Pic u e Usage in CSR Repo s
My analysis indica es h ee a iables ha in luence he
equency o image usage. While b oade cha ac e is ics such
as ESG sco es o indus y sensi i i y (like ini ially expec ed)
did no di ec ly co ela e wi h he equency o images being
used, signi ican e ec s we e ound o CO2 Emissions pe
e enue. This migh sugges ha conc e e and measu able
pe o mance me ics in luence image use a he han gene al
sensi i i y o ESG pe o mance, possibly due o hei di ec
T. Janßen /Junio Managemen Science 10(4) (2025) 1009-10271024
disclosu e and isibili y in CSR epo s. CO2 emissions a e
cen al in CSR epo ing and a e easily unde s andable e en
by less expe ienced eade s. Fi ms wi h highe CO2 in ensi y
han hei pee s seem o employ mo e images, po en ially
c ea ing a posi i e public image and di e ing a en ion om
less a o able disclosu e.
Mo eo e , companies disclosing mo e ex ensi ely on CSR
opics (indica ed by a highe GRI Repo ing Sco e) use ewe
images. This sugges s ha mo e anspa en i ms ha p o-
ide ex ensi e, compa able (quan i a i e) in o ma ion migh
ely less on isuals o communica e hei na a i e. Con-
e sely, i ms wi h less subs an i e epo ing migh employ
isuals mo e s a egically o enhance pe cei ed anspa ency,
compensa ing o a lack o ac ual in o ma ion being dis-
closed, simila o he indings o Ga cía-Sánchez and A aújo-
Be na do (2020). These obse a ions align wi h my dis-
cussion abou he a ionale o CSR epo ing (Legi imacy
Theo y) and ha companies migh use aming and imp es-
sion managemen o enhance pe cei ed legi imacy and ha
isuals migh be seen as a po en ool o o e shadow less a-
o able aspec s (e.g., (Boi al, 2013; Cho e al., 2009; Posne
e al., 1976; Siano e al., 2017; Suchman, 1995)).
I obse ed he egion o be ano he ac o de e mining
he equency o pic u e usage. While companies in he EU
end o use mo e images o e he whole pe iod ( o easons
I canno di ec ly de i e om my discussions), I showed ha
he egula o y en i onmen a company ope a es in migh sig-
ni ican ly a ec and change companies’ decisions on pic u e
usage. My Di e ence in Di e ence (DiD) analysis, while
sligh ly insigni ican using obus s anda d e o s (P- alues:
0.056 /0.064), p o ides a easonable basis o belie e ha
he Non-Financial Repo ing Di ec i e (NFRD) has a nega i e
impac on he numbe o pic u es used in CSR epo s12. This
inding could be explained by p e ious li e a u e and my e-
sul s. Fiech e e al. (2022) showed ha he NFRD inc eases
epo ing anspa ency, and my analysis indica ed a dec ease
in pic u e usage h ough inc eased epo ing anspa ency
( ep esen ed by adhe ence o GRI guidelines -also a c ucial
ac o o anspa ency used in Fiech e ’s esea ch). Toge he ,
his migh explain, in pa , he obse ed e ec . The impac o
he NFRD on pic u e usage was mo e p onounced in sensi i e
indus ies. This g ea e e ec can po en ially be a ibu ed o
he inc eased sc u iny hese companies ace (Game schlag e
al., 2011), which likely in ensi ied wi h he NFRD, pushing
hese companies owa ds mo e p ecise epo ing.
8.1.2. De e minan s o Pic u e Con en
In my analysis, I obse e ha he social dimension dom-
ina es he isual con en in CSR epo s, wi h almos hal o
all pic u es depic ing his heme being in line wi h u he
esea ch (In e nizzi e al., 2022). This migh be because im-
ages depic ing people c ea e a sense o closeness. In gene al,
12 While he ea men e ec would be expec ed ei he a e 2014 ( he yea
he NFRD was passed) o 2016 ( he epo s o 2017 we e he i s o
all unde he di ec i e) I ound he e ec a e 2018. While he e is no
de ini e explana ion o his, I assume ha his migh be due o g adual
adap ion e ec s and eedback adjus men s.
ela edness and connec ions owa ds he company a e essen-
ial o keeping s akeholde us , e lec ing companies’ in en-
ions o align hemsel es wi h social esponsibili ies. (Da i-
son, 2010)
My indings indica e se e al ac o s ha could in luence
pic u e ca ego y dis ibu ion ac oss epo s. No ably, he so-
cial dimension appea s qui e s able, wi h only he closeness
o ope a ions o cus ome s (“Cus ome _P oximi y”) and ESG
sco e showing a ela ionship wi h he use o social images,
bo h being posi i e. I seems ha companies close o hei
cus ome s migh use isuals o emphasize hei communi y
in ol emen and social con ibu ions. The posi i e associa-
ion wi h ESG sco e is no s aigh o wa d. S ill, i migh sug-
ges ha companies wi h be e ESG pe o mance could be
using mo e social images, possibly due o he a he gene ic
con en o hese images and a educed need o dis inguish
hemsel es h ough specialized isuals.
In e es ingly, companies iden i ied as socially sensi i e do
no signi ican ly in luence he use o social- hemed pic u es
as ini ially expec ed. Howe e , hese companies show a e-
duced use o en i onmen al pic u es, which indica es a shi
in isual ocus. Al hough I can’ de ini i ely conclude ha
hese companies a e inc easing hei use o social images, he
coe icien s o Co po a e Excellence and Social a e bo h pos-
i i e. Social is signi ican ly highe , hin ing ha hese compa-
nies could a o social image y mo e.
En i onmen al images a e mo e equen ly used by com-
panies ac i ely engaged in en i onmen al ini ia i es. This
inding aligns wi h he dual capaci y idea o isuals (Da ison,
2015) and, in his case, se es a ep esen a ional pu pose, in
con as o a cons uc i e pu pose ha I ha e obse ed so a .
Companies wi h a highe CO2 in ensi y o e enue
(“log(CO2_Emissions_Re )”) end o use mo e en i onmen-
al pic u es and ewe Co po a e Excellence ones, possibly as
a s a egic mo e o po ay hemsel es as en i onmen ally
esponsible despi e hei la ge en i onmen al impac .
I obse e con as ing pa e ns o companies sensi i e
o social and en i onmen al issues (E_X_S) and hose only
en i onmen ally sensi i e (ENV). En i onmen ally sensi i e
i ms seem o u ilize mo e Co po a e Excellence images, pe -
haps indica ing hei capabili y o add ess en i onmen al
challenges. This choice migh be made because ce ain in-
dus ies canno clea ly show di ec en i onmen al bene i s
due o hei ope a ional na u e; ins ead, hey demons a e
compe ency in managing such challenges and add essing
inhe en indus y issues. Explaining he e ec o companies
being sensi i e in bo h dimensions (Using mo e en i onmen-
al and less Co po a e Excellence pic u es) is mo e complex,
as heo e ically, he same line o easoning could apply o
hem. Inspec ing he ypes o companies in my sample ha
all in o his ca ego y, I see ha hese ope a e in basic ma-
e ials (e.g., mining and me al ab ica ion). Again, gi en
he inhe en limi a ions o hei indus ies, hese companies
ace s uc u al cons ain s ha p e en subs an ial changes
h ough inno a ion alone. The p ocesses hey ely on a e
la gely ixed, and meaning ul imp o emen s in hei en i-
onmen al impac o en ely on b oade ad ancemen s in
T. Janßen /Junio Managemen Science 10(4) (2025) 1009-1027 1025
high-densi y g een ene gy, which is beyond hei di ec con-
ol. In his con ex , po aying hemsel es as pionee s o
en i onmen al inno a ion may no be seen as au hen ic due
o he na u e o hei ope a ions. Consequen ly, hese i ms
migh op o gene ic en i onmen al images in hei CSR
epo s, possibly o c ea e a pe cep ion o en i onmen al
awa eness. This s a egy could be seen as a emp ing o
align wi h socie al expec a ions and con ey a commi men
o en i onmen al goals, e en when ope a ional changes a e
limi ed.
Again, con a y o wha I expec ed, I see no co ela ion
be ween ESG pe o mance (o he di e en single sco es)
and he ype o pic u es used.
8.2. Limi a ions
Se e al ac o s educe he gene alizabili y o he esul s:
he sampled da a, he (classi ica ion) me hodology, and he
esea ch design (in addi ion o no p e ious simila s udy).
A en ion mus be paid o hese limi a ions as hey a e c u-
cial when classi ying he esul s ob ained. In he ollowing, I
will discuss hese limi a ions and p o ide sugges ions on how
u u e esea ch migh deal wi h hem.
8.2.1. Ga he ed Sample
In c ea ing my sample, I in oduced a selec ion bias by
only conside ing companies ha p oduce sepa a e CSR e-
po s. The decision o publish a sepa a e epo could ha e al-
eady a ec ed he epo ing beha io s I am s udying h ough
eg ession analyses and migh , in u n, be in luenced by some
o he cha ac e is ics o a iables unde conside a ion in my
s udy. Fu he , he di ec ma ching app oach migh no lead
o op imal sample g oups since i only ma ches indus y and
e enue; o he impo an ac o s could ha e been igno ed.
Mo eo e , he composi ion o hese g oups changes e e y
yea , especially in he yea s be o e 2016, when ewe com-
panies ei he did no epo a all o issued in eg a ed e-
po s. In addi ion, my seconda y da a has missing obse a-
ions, which u he limi s he sample size o some eg es-
sions. Fo u u e s udies, including in eg a ed epo s om a
la ge ini ial sample and using PSM o g oup c ea ion could
be bene icial. Ensu ing a consis en da ase ha includes he
same companies each yea would also be an ad an age i he
sample size is su icien .
8.2.2. Pic u e Classi ica ion
My unique app oach o con en analysis allowed o he
analysis o a signi ican numbe o pic u es wi h compa a-
ble low e o , bu i has limi a ions. I canno apply ypi-
cal e o es s wi hou manually classi ying pic u es and hus
lack a benchma k o "co ec " classi ica ion. This leads o
unce ain y ega ding he accu acy o he o med clus e s.
The echnical e alua ions, like he Silhoue e Sco e, indi-
ca ed a he low alues (0.162), wi h he sco e anging om
-1 o 1, and 1 being he bes sco e, sugges ing ha clus e ing
was echnically weak. While i does no necessa ily mean he
classi ica ion was ine ec i e, as i is easonable ha , e.g., a
pic u e depic ing an o ice scene would be qui e simila o pic-
u es depic ing a co po a e p o essional, he p ocess could,
ne e heless, bene i om a deepe analysis o he wo ds ha
mos s ongly in luence he embeddings and clus e o ma-
ions. Fi s , inspec ing a subse o he pic u es and iden i-
ying whe he he cu en impo ance o single labels accu-
a ely cap u es he di e ences ha a e o in e es , o i o he
labels a e be e sui ed. Second, adjus ing he weigh s o sin-
gle labels o mee he equi emen s could lead o mo e ac-
cu a e clus e s. This could eplace he manual adjus men s I
ha e done a e he ini ial clus e ing, educing he possibil-
i y o manual e o s. Addi ionally, I only conside ed pic u e
con en wi hou add essing a ibu es like colo and size o
he depic ion o emo ions (e.g., Ga cía-Sánchez and A aújo-
Be na do, 2020; In e nizzi e al., 2022), which could en-
hance ca ego y di e en ia ion when added as ano he laye .
As he Google Vision API can also analyze a pic u e’s colo
composi ion, his could be in eg a ed wi h an au oma ed ap-
p oach.
8.2.3. Resea ch Design
My eg essions could only cap u e a small po ion o he
da ase a iabili y, signaled by low alues o R-squa ed (<
0.2). While his may be due o in insic a iabili y in he
ou come a iables, i could also a ise om excluding some
impo an a iables no conside ed du ing he analysis due
o limi ed esea ch in his a ea. The backwa d elimina ion
me hod also has limi a ions. Al hough i o e s a sys ema ic
way o selec ing a iables o he model, i migh ha e missed
ou on some combina ions ha could explain ou comes mo e
e ec i ely. Fu u e s udies a e ecommended o del e in o
di e en se s o a iables along wi h hei in e ac ions o ei-
he con i m he pe sis ence o he obse ed e ec s o e eal
new explana o y ones. Mo eo e , I used only a no mal linea
model wi hou conside ing ime and uni ixed e ec s when
unning eg essions on he numbe o pic u es and pic u e
con en . This could be included in a panel eg ession in u-
u e s udies, ideally wi h mo e e enly dis ibu ed obse a-
ions o e he yea s.
9. Conclusion and Indica ion o Fu u e Resea ch
Only a ew s udies ha e examined he use o pho og aphs
in CSR epo s so a ; wi h my s udy, I con ibu e o his ield
by explo ing de e minan s ega ding he amoun and ype
o pic u es used. I ha e done so by using a semi-au oma ed
classi ica ion app oach, h ough which I could u ilize a com-
pa ably la ge se o pic u es and eg ession models.
In he pape , I analyzed why images a e employed in CSR
Repo s, he di e en pu poses hey can se e, and wha ac-
o s migh in luence he decision on how many and wha ype
o pic u es a e used. The cen al heme o he esea ch ques-
ion was whe he isuals a e used s a egically o cons uc a
posi i e image o a company, po en ially independen o i s
ac ual pe o mance.
I ound ha indus y sensi i i y and measu able indica-
o s, such as CO2 emissions pe e enue, a ec he amoun
T. Janßen /Junio Managemen Science 10(4) (2025) 1009-10271026
and con en o pic u es used in CSR epo s. These indings
sugges ha images a e cons uc i e elemen s in imp ession
managemen , o e shadowing o mi iga ing ad e se impac s
and indus y laws. Howe e , in some cases, hey could also
se e as desc ip i e elemen s, e.g., when companies engaged
in en i onmen al ini ia i es end o use mo e en i onmen al
pic u es.
Fu he , I also assessed he impac o he Non-Financial
Repo ing Di ec i e (NFRD) on using isuals in CSR epo s.
I disco e ed ha he NFRD has a mode a ing e ec , educing
he numbe o pic u es used, which migh enhance he ans-
pa ency o hese epo s and po en ially lead hem o ocus
mo e on he ac ual disclosu e o ele an in o ma ion.
Gi en my indings, I sugges ha S akeholde s mus be
conscious o he possibili y ha companies s a egically use
images in CSR epo s. Employed as imp ession managemen
echniques, hey ypically show an idealized image o he o -
ganiza ion. Unde s anding his aspec is c ucial, as he in-
luence o isuals on pe cep ion is o en subconscious, and
being awa e o i can assis in mi iga ing hese subconscious
e ec s.
Fu he esea ch in his a ea could add ess he limi a ions
o his s udy and es whe he hese esul s hold o di e en
samples o explo e di e en a iables and combina ions ha
may a ec he use o images. I would also be bene icial o
examine he impac o he NFRD on he con en o isuals
in CSR epo s. Fu he mo e, as in his pape , I only in es i-
ga ed he pa e ns in luencing he usage o images, i would
also be in e es ing o s udy whe he and o wha ex en he
amoun and con en o pic u es in luence he alua ion o a
company’s ESG pe o mance and pe cep ion o sus ainabili y
e o s.
Re e ences
Accen u e. (2022). Unlocking he Global Pa hways o Resilience, G ow h,
and Sus ainabili y o 2030: The 12 h Uni ed Na ions Global
Compac -Accen u e CEO S udy, 2022. h ps://www.accen u e
.com/con en /dam/accen u e/ inal/accen u e-com/documen
/Accen u e-CEO-S udy-Uni ed-Na ions-Global-Compac .pd
Ba on, D. P. (2001). P i a e Poli ics, Co po a e Social Responsibili y, and In-
eg a ed S a egy. Jou nal o Economics & Managemen S a egy,
10(1), 7–45. h ps://doi.o g/10.1111/j.1430-9134.2001.00007
.x
Benabou, R., & Ti ole, J. (2010). Indi idual and Co po a e Social Responsi-
bili y. Economica,77(305), 1–19. h ps://doi.o g/10.1111/j.146
8-0335.2009.00843.x
BlackRock. (2020). Sus ainabili y goes mains eam, 2020. h ps://img.lal .c
o/cms/2021/05/28202727/black ock-sus ainabili y-su ey.pd
Boi al, O. O., & Hen i, J.-F. (2017). Is Sus ainabili y Pe o mance Compa a-
ble? A S udy o GRI Repo s o Mining O ganiza ions. Business &
Socie y,56(2), 283–317. h ps://doi.o g/10.1177/0007650315
576134
Boi al, O. O. (2013). Sus ainabili y epo s as simulac a? A coun e -accoun
o A and A+GRI epo s. Accoun ing, Audi ing & Accoun abili y
Jou nal,26(7), 1036–1071. h ps://doi.o g/10.1108/AAAJ-04-
2012-00998
Bone i, P., Cho, C. H., & Michelon, G. (2023). En i onmen al Disclosu e
and he Cos o Capi al: E idence om he Fukushima Nuclea
Disas e . Eu opean Accoun ing Re iew, 1–29. h ps://doi.o g/10
.1080/09638180.2023.2203410
B amme , S., & Milling on, A. (2005). Co po a e Repu a ion and Philan-
h opy: An Empi ical Analysis. Jou nal o Business E hics,61(1),
29–44. h ps://doi.o g/10.1007/s10551-005-7443-4
B anco, M. C., & Rod igues, L. L. (2008). Fac o s In luencing Social Respon-
sibili y Disclosu e by Po uguese Companies. Jou nal o Business
E hics,83(4), 685–701. h ps://doi.o g/10.1007/s10551-007-9
658-z
Business Round able. (2019). Business Round able Rede ines he Pu pose
o a Co po a ion o P omo e ‘An Economy Tha Se es All Ame -
icans’. h ps://www.business ound able.o g/business- ound abl
e- ede ines- he-pu pose-o -a-co po a ion- o-p omo e-an-econo
my- ha -se es-all-ame icans
By d, J., Hickman, K., Bake , C., & Cohanie , B. (2017). Co po a e Social
Responsibili y Repo ing in Con o e sial Indus ies. h ps://doi
.o g/10.2139/ss n.2894789
Ca le ion, T. A., & Hsiang, S. M. (2016). Social and economic impac s o
clima e. Science,353(6304). h ps://doi.o g/10.1126/science.a
ad9837
Cho, C., Phillips, J., Hageman, A., & Pa en, D. (2009). Media ichness, use
us , and pe cep ions o co po a e social esponsibili y: An ex-
pe imen al in es iga ion o isual web si e disclosu es. Accoun -
ing, Audi ing and Accoun abili y Jou nal,22(6), 933–952. h ps:
//doi.o g/10.1108/09513570910980481
Chong, S., Na ayan, A. K., & Ali, I. (2019). Pho og aphs depic ing CSR: cap-
u ed eali y o c ea i e illusion? Paci ic Accoun ing Re iew,31(3),
313–335. h ps://doi.o g/10.1108/PAR-10-2017-0086
Ch is ensen, H. B., Hail, L., & Leuz, C. (2021). Manda o y CSR and sus ain-
abili y epo ing: economic analysis and li e a u e e iew. Re iew
o Accoun ing S udies,26, 1176–1248. h ps://doi.o g/10.1007
/s11142-021-09609-5
Da ison, J. (2008). Rhe o ic, epe i ion, epo ing and he “do .com” e a:
wo ds, pic u es, in angibles. Accoun ing, Audi ing & Accoun abil-
i y Jou nal,21(6), 791–826. h ps://doi.o g/10.1108/0951357
0810893254
Da ison, J. (2010). [In] isible [in] angibles: Visual po ai s o he business
éli e. Accoun ing, O ganiza ions and Socie y,35(2), 165–183. h
ps://doi.o g/10.1016/j.aos.2009.03.003
Da ison, J. (2015). Visualising accoun ing: an in e disciplina y e iew and
syn hesis. Accoun ing and Business Resea ch,45(2), 121–165. h
ps://doi.o g/10.1080/00014788.2014.987203
Da ison, J., & Wa en, S. (2009). Imag[in]ing accoun ing and accoun abili y.
Accoun ing, Audi ing & Accoun abili y Jou nal,22(6), 845–857. h
ps://doi.o g/10.1108/09513570910980436
Dye , T., Lang, M., & S ice-Law ence, L. (2017). The E olu ion o 10-K Tex-
ual Disclosu e: E idence om La en Di ichle Alloca ion. Jou -
nal o Accoun ing & Economics (JAE), Fo hcoming. h ps://doi.o
g/10.2139/ss n.2741682
E ickson, S. L., Webe , M., & Sego ia, J. (2011). Using Communica ion The-
o y o Analyze Co po a e Repo ing S a egies. Jou nal o Business
Communica ion,48(2), 207–223. h ps://doi.o g/10.1177/002
1943611399728
Fiech e , P., Hi z, J.-M., & Lehmann, N. (2022). Real E ec s o a Widesp ead
CSR Repo ing Manda e: E idence om he Eu opean Union’s
CSR Di ec i e. Jou nal o Accoun ing Resea ch,60(4), 1499–1549.
h ps://doi.o g/10.1111/1475-679X.12424
Game schlag, R., Mölle , K., & Ve bee en, F. (2011). De e minan s o olun-
a y CSR disclosu e: empi ical e idence om Ge many. Re iew o
Manage ial Science,5, 233–262. h ps://doi.o g/10.1007/s1184
6-010-0052-3
Ga cía-Sánchez, I.-M., & A aújo-Be na do, C.-A. (2020). Wha colou is he
co po a e social esponsibili y epo ? S uc u al isual he o ic,
imp ession managemen s a egies, and s akeholde engagemen .
Co po a e Social Responsibili y and En i onmen al Managemen ,
27(2), 1117–1142. h ps://doi.o g/10.1002/cs .1869
Ga i, L., Pizze i, M., & Seele, P. (2021). G een lies and hei e ec on in en-
ion o in es . Jou nal o Business Resea ch, 228–240. h ps://doi
.o g/10.1016/j.jbus es.2021.01.028
Global Repo ing Ini ia i e. (2006). Sus ainabili y epo ing guidelines. Am-
s e dam, The Ne he lands.
Google. (2024). Cloud Vision documen a ion. h ps://cloud.google.com/ i
sion/docs?hl=de
T. Janßen /Junio Managemen Science 10(4) (2025) 1009-1027 1027
G ougiou, V., Dedoulis, E., & Le en is, S. (2016). Co po a e Social Respon-
sibili y Repo ing and O ganiza ional S igma: The Case o “Sin”
Indus ies. Jou nal o Business Resea ch,69(2), 905–914. h ps:
//doi.o g/10.1016/j.jbus es.2015.06.041
Hahn, R., & Kühnen, M. (2013). De e minan s o sus ainabili y epo ing: a
e iew o esul s, ends, heo y, and oppo uni ies in an expand-
ing ield o esea ch. Jou nal o Cleane P oduc ion,59, 5–21. h
ps://doi.o g/10.1016/j.jclep o.2013.07.005
Ha pe , D. (2002). Talking abou pic u es: A case o pho o elici a ion. Visual
S udies,17(1), 13–26. h ps://doi.o g/10.1080/147258602201
37345
He, C., & Lo us, J. (2014). Does en i onmen al epo ing e lec en i on-
men al pe o mance? Paci ic Accoun ing Re iew,26(1/2), 134–
154. h ps://doi.o g/10.1108/PAR-07-2013-0073
H asky, S. (2012). Visual disclosu e s a egies adop ed by mo e and less
sus ainabili y-d i en companies. Accoun ing Fo um,36(3), 154–
165. h ps://doi.o g/10.1016/j.acc o .2012.02.001
Huang, C.-L., & Kung, F.-H. (2010). D i e s o En i onmen al Disclosu e and
S akeholde Expec a ion: E idence om Taiwan. Jou nal o Busi-
ness E hics,96, 435–451. h ps://doi.o g/10.1007/s10551-010-
0476-3
Humai a, H., & Rasyidah, R. (2018). De e mining The App opia e Clus e
Numbe Using Elbow Me hod o K-Means Algo i hm. h ps://d
oi.o g/10.4108/eai.24-1-2018.2292388
In e nizzi, A. C., Bellucci, M., Acu i, D., & Mane i, G. (2022). Fo m and sub-
s ance: Visual con en in CSR epo s and in es o s’ pe cep ions.
Psychology & Ma ke ing,39(5), 974–989. h ps://doi.o g/10.10
02/ma .21635
Ki zmuelle , M., & Shimshack, J. (2012). Economic Pe spec i es on Co po-
a e Social Responsibili y. Jou nal o Economic Li e a u e,50(1),
51–84. h ps://doi.o g/10.1257/jel.50.1.51
Legend e, T. S., Ja is, N., Kang, Y., Jamal, G., & Jackson, J. (2020). Rescu-
ing impe ec p oduce: The e ec s o s igma disclosu e s a egies,
con ollabili y, and aes he ics. In e na ional Jou nal o Hospi ali y
Managemen ,85, 102443. h ps://doi.o g/10.1016/j.ijhm.2019
.102443
Nielsen, A. E., & Thomsen, C. (2018). Re iewing co po a e social espon-
sibili y communica ion: a legi imacy pe spec i e. Co po a e Com-
munica ions: An In e na ional Jou nal,23(4), 492–511. h ps://d
oi.o g/10.1108/CCIJ-04-2018-0042
Posne , M. I., Nissen, M. J., & Klein, R. M. (1976). Visual dominance: An
in o ma ion-p ocessing accoun o i s o igins and signi icance.
Psychological Re iew,83(2), 157–171. h ps://doi.o g/10.103
7/0033-295X.83.2.157
P es on, A. M., W igh , C., & Young, J. J. (1996). IMAG[IN]ING ANNUAL RE-
PORTS. Accoun ing, O ganiza ions and Socie y,21(1), 113–137.
h ps://doi.o g/10.1016/0361-3682(95)00032-5
Rämö, H. (2011). Visualizing he Ph one ic O ganiza ion: The Case o Pho-
og aphs in CSR Repo s. Jou nal o Business E hics,104, 371–
387.
Ri e os-Ga ilanes, J. M. (2023). Tes ing Pa allel T ends in Di e ences-in-
Di e ences and E en S udy Designs: A Resea ch App oach Based
on P e-T ea men Pe iod Signi icance. Jou nal o Resea ch, Inno-
a ion and Technologies (JoRIT),2(4), 226–237. h ps://doi.o g
/10.57017/jo i . 2.2(4).07
Se a eim, G., & Amel-Zadeh, A. (2017). Why and How In es o s Use ESG
In o ma ion- E idence om a Global Su ey. Financial Analys s
Jou nal,74(3), 87–103. h ps://doi.o g/10.2139/ss n.2925310
Siano, A., Volle o, A., Con e, F., & Amabile, S. (2017). “Mo e han wo ds”:
Expanding he axonomy o g eenwashing a e he Volkswagen
scandal. Jou nal o Business Resea ch,71, 27–37. h ps://doi.o g
/10.1016/j.jbus es.2016.11.002
Song, K., Tan, X., Qin, T., Lu, J., & Liu, T.-Y. (2020). MPNe : Masked and
Pe mu ed P e- aining o Language Unde s anding. h p://a xi
.o g/pd /2004.09297 2
S olowy, H., & Paugam, L. (2018). The expansion o non- inancial epo ing:
an explo a o y s udy. Accoun ing and Business Resea ch,48(5),
525–548. h ps://doi.o g/10.1080/00014788.2018.1470141
Suchman, M. C. (1995). Managing Legi imacy: S a egic and Ins i u ional
App oaches. The Academy o Managemen Re iew,20(3), 571–
610. h ps://doi.o g/10.2307/258788
Ta e, W. L., Ell am, L. M., & Ki cho , J. L. (2009). CORPORATE SOCIAL RE-
SPONSIBILITY REPORTS:A THEMATIC ANALYSIS RELATED TO
SUPPLYCHAIN MANAGEMENT. Jou nal o Supply Chain Manage-
men ,46(1), 19–44. h ps://doi.o g/10.1111/j.1745-493X.200
9.03184.x
Townsend, C., & Shu, S. B. (2010). When and how aes he ics in luences
inancial decisions. Jou nal o Consume Psychology,20(4), 452–
458. h ps://doi.o g/10.1016/j.jcps.2010.06.013
T e sky, B. (1974). Eye ixa ions in p edic ion o ecogni ion and ecall.
Memo y & Cogni ion,2, 275–278. h ps://doi.o g/10.3758/BF0
3208995
Vasue, E. S., & Howe, A. C. (1989). The e ec o isual and e bal modes o
p esen a ion on child en’s e en ion o images and wo ds. J. Res.
Sci. Teach.,26(5), 401–407. h ps://doi.o g/10.1002/ ea.36602
60505
Wang, Y. (2019). Single T aining Dimension Selec ion o Wo d Embed-
ding wi h PCA. P oceedings o he 2019 Con e ence on Empi ical
Me hods in Na u al Language P ocessing and he 9 h In e na ional
Join Con e ence on Na u al Language P ocessing (EMNLP-IJCNLP),
3597–3602. h ps://doi.o g/10.18653/ 1/D19-1369
Wo ell, E., Be ns ein, L., Roy, J., P ice, L., & Ha nisch, J. (2009). Indus-
ial ene gy e iciency and clima e change mi iga ion. Ene gy E -
iciency,2, 109–123. h ps://doi.o g/10.1007/s12053-008-903
2-8
Yilmaz, Ö., Achenie, L. E. K., & S i as a a, R. (2007). Sys ema ic uning
o pa ame e s in suppo ec o clus e ing. Ma hema ical Bio-
sciences,205, 252–270. h ps://doi.o g/10.1016/j.mbs.2006.09
.013
