Lukas, Ch is ian; Neube , Max-F ede ik; Schöndube, Jens Robe
A icle — Published Ve sion
Explo ing decision-making: expe imen al obse a ions on
p ojec selec ion and he impac o jus i ica ion p essu e
Jou nal o Managemen and Go e nance
P o ided in Coope a ion wi h:
Sp inge Na u e
Sugges ed Ci a ion: Lukas, Ch is ian; Neube , Max-F ede ik; Schöndube, Jens Robe (2024) :
Explo ing decision-making: expe imen al obse a ions on p ojec selec ion and he impac o
jus i ica ion p essu e, Jou nal o Managemen and Go e nance, ISSN 1572-963X, Sp inge US, New
Yo k, NY, Vol. 29, Iss. 3, pp. 735-775,
h ps://doi.o g/10.1007/s10997-024-09717-9
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/330803
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/
Vol.:(0123456789)
Jou nal o Managemen and Go e nance (2025) 29:735–775
h ps://doi.o g/10.1007/s10997-024-09717-9
Explo ing decision‑making: expe imen al obse a ions
onp ojec selec ion and heimpac o jus i ica ion p essu e
Ch is ianLukas1 · Max‑F ede ikNeube 2· JensRobe Schöndube3
Accep ed: 29 July 2024 / Published online: 2 Sep embe 2024
© The Au ho (s) 2024
Abs ac
In his expe imen al in es iga ion, we explo e he impac o jus i ica ion on p ojec
choices. In oducing a no el elemen , we implemen asymme ic payo schemes
commonly employed in business, signi ying dis inc payo dis ibu ions o he
i m (p incipal) and he manage (agen ). The agen has o choose one p ojec om
wo op ions ha di e in hei isk- e u n p o iles. The ou comes o ou expe imen
subs an ia e ou hypo hesis, indica ing ha a manda e o jus i ica ion dec eases he
p obabili y o agen s selec ing he p ojec wi h highe isk and e u n. The deg ee o
his educ ion appea s o hinge on he na u e o jus i ica ion. Inc eased p o i sha es
o he agen o a p ojec ecommenda ion om he p incipal can pa ially coun e -
balance he dis o ion in he p ojec choice.
Keywo ds Agency· Beha io al accoun ing· Expe imen · Incen i es· Jus i ica ion·
P ojec selec ion
JEL Classi ica ion C72· C91· D81· M40· M52
1 In oduc ion
I is a s anda d p ac ice o manage s o p o ide a a ionale o hei decisions and
elucida e he ou comes o hei choices o ac ions. In business, examples o jus i i-
ca ions encompass a ious scena ios such as employees epo ing o supe io s, op
* Ch is ian Lukas
[email p o ec ed]
Jens Robe Schöndube
schoendube@con olling.uni-hanno e .de
1 F ied ich-Schille -Uni e si ä Jena, Ca l-Zeiss-S asse 3, 07743Jena, Ge many
2 O o- on-Gue icke-Uni e si ä Magdebu g, Uni e si ä spla z 2, 39106Magdebu g, Ge many
3 Ins i u e o Manage ial Accoun ing, Leibniz Uni e si ä Hanno e , Königswo he Pla z 1,
30167Hanno e , Ge many
736
C.Lukas e al.
managemen add essing inqui ies du ing in es o calls, o he boa d commen ing
on company pe o mance in gene al mee ings. The equi emen o jus i y decisions
and ou comes in hese and simila si ua ions may lead o pe sonal cos s, mani es ing
as p essu e, discom o , o e en se e e s ess o indi iduals asked wi h de ending
hemsel es (F imanson e al., 2021; Robe s, 2009). Fo ins ance, a he uppe ech-
elons o managemen , he CEO bea s accoun abili y o he boa d o di ec o s. Con-
sequen ly, he CEO mus a ionalise majo decisions, pa icula ly hose o a s a e-
gic na u e ha impac he company’s long- e m success, and add ess (nega i e) pas
ou comes be o e he boa d. Howe e , he p essu e o p o ide jus i ica ions is no
exclusi e o he op ie . Taking, o ins ance, a p ojec manage who mus explain
delays o budge o e uns in comple ing a p ojec .
The expe imen al esea ch on jus i ica ion p edominan ly u ilises se ings in ol -
ing a single-pe son (Vieide , 2009, 2011), single-pe son se ings wi h hypo he ical
second pa ies (Bauch & Weißenbe ge , 2020); Feh enbache e al., 2020), o wo-
pe son se ings wi h symme ic payo s uc u es (Pahlke e al., 2012).1
Howe e , in a business con ex , payo s uc u es commonly exhibi asymme y,
leading o unequal payo sha es o he decision-make , i.e., he manage , and he
i m. Aligned wi h he p ima y goal o his s udy, we explo e whe he and in wha
manne jus i ica ion in luences p ojec decisions wi hin a con ex ep esen a i e o
con ol issues in i ms. In his scena io, he au ho i y o make p ojec choices is del-
ega ed o a manage whose compensa ion is ied o he p ojec ’s success. Howe e ,
ac ing as he esidual claiman , he i m ypically ecei es a sha e o he success
dis inc om he manage ’s.
The second aim o ou s udy is connec ed o he conclusions d awn by Vieide
(2009) and Pahlke e al. (2012), which p opose ha jus i ica ions heigh en he p ob-
abili y o op ing o isky p ojec s ha could esul in losses. Gi en ha analy ical
e idence (Lukas e al. (2019), o he model in he p esen pape ) sugges s a con a y
p edic ion, indica ing less inclina ion owa ds isk- aking, we examine ou asymme -
ic payo scheme conce ning he decision make ’s p opensi y.
Ou hi d objec i e explo es whe he he na u e o wha he decision-make has o
jus i y makes a di e ence. Insigh s om he pe o mance e alua ion li e a u e indi-
ca e ha he me e expec a ion o de ending one’s ac ions can be p o oundly s ess ul
o indi iduals asked wi h jus i ying hei ac ions and ou comes (F imanson e al.,
2021; Messne , 2009; Robe s, 2009). The way indi iduals pe cei e hese pe sonal
cos s may a y. Ne e heless, hei impac p obably also hinges on he jus i ica ion,
speci ically, whe he one has o jus i y ou comes o decisions. We con ibu e o he
li e a u e by in es iga ing how h ee dis inc jus i ica ion egimes o ypes com-
monly employed in managemen p ac ices in luence he pe cei ed cos s o jus i ica-
ion. These egimes di e in ha he indi idual is equi ed o jus i y (1) he decision,
(2) he ou come, o (3) he ou come when i alls below a h eshold.
Ano he ac o a ec ing he cos s o jus i ica ion may be supe io s’ p e e ed
cou ses o ac ion, such as sha eholde s’ a ou ed in es men s a egy o he le el
1 Pollmann e al. (2014) simila ly u ilised asymme ic payo s, whe e he agen ’s ma e ial payo depends
on he p incipal’s decision o ewa d he agen be o e o a e he p ojec choice.
737
Explo ing decision‑making: expe imen al obse a ions on…
o isk he i m o di ision is willing o bea o i s e u ns. Op ing o a decision
aligned wi h he p e e ed s a egy is he mos easily de ensible choice (Te lock,
1985), whe eas a ied p e e ences a e likely o escala e he cos s associa ed wi h
jus i ica ion. While hese e ec s a e in ui i ely unde s andable, as o ou knowledge
cu o da e, no speci ic e idence is de i ed om a se ing wi h asymme ic payo
schemes as in ou s udy. Hence, ou ou h objec i e is o in es iga e how he com-
munica ion o a p e e ence in e ac s wi h asymme ic pay and ypes o jus i ica ion
conce ning pe cei ed cos s o jus i ica ion and p ojec choice.
To add ess ou esea ch inqui ies, we conduc ed a compu e ised labo a o y
expe imen in ol ing 360 unde g adua e and g adua e s uden pa icipan s a Leib-
niz Uni e si y Hanno e (Ge many) in se e al sessions in 2015 and 2019. We es ab-
lished an agency si ua ion whe e he agen is accoun able o choosing a p ojec . The
p incipal ( i m owne ) is he esidual claiman o he p ojec ’s payo and equi es
he agen o jus i y he p ojec choice o pe o mance i i alls below a p ede e -
mined le el. The p incipal-agen pai s engage in in e ac ions ac oss mul iple deci-
sion ounds. In each ound, he agen chooses be ween wo a ailable p ojec s: a
s anda d p ojec and a isiona y p ojec wi h a highe mean e u n and a iance.
We ep esen hese p ojec s as basic lo e ies wi h wo dis inc bu equally p obable
ou comes (low/high). No ably, he ex-an e e icien ( isiona y) high- isk/high- e u n
p ojec may esul in a loss o he p incipal.
The expe imen in ol es h ee manipula ions: i s , we modi y he agen ’s payo s
om he a ailable p ojec s (adjus ing his/he a iable pay); second, we change he
equi emen o jus i y decision-making o pe o mance; and hi d, we al e he p in-
cipal’s oppo uni y o communica e p ojec p e e ences ( ecommenda ions) o he
agen .
We o mula e ou hypo heses using a s aigh o wa d analy ical model connec ed
o Lukas e al. (2019). This model is si ua ed wi hin he amewo k o managemen
con ol al e na i es p oposed by Me chan and Van de S ede (2007). I in eg a es
esul s con ol (ou come-con ingen pay), ac ion con ol (jus i ica ion), and pe son-
nel con ol (p ojec ecommenda ion cla i ies he i m’s expec a ion). The expe i-
men al esul s essen ially suppo ou hypo heses. He e a e ou key indings: (i) he
p esence o a jus i ica ion equi emen dec eases he p obabili y o agen s selec -
ing mo e luc a i e and iskie p ojec s, (ii) p incipals’ ini ial ecommenda ions and
inc eased a iable pay o such p ojec s coun e ac he impac o jus i ica ion, esul -
ing in mo e equen selec ion o such p ojec s, and (iii) decision jus i ica ion elici s
he highes compliance wi h ecommenda ions o such p ojec s.
We make wo con ibu ions o he expe imen al li e a u e on jus i ica ion e ec s.
The i s con ibu ion ela es o he (a)symme y o he payo scheme ele an in
a se ing wi h a jus i ica ion equi emen . The second con ibu ion pe ains o he
e en igge ing he jus i ica ion. Conce ning he i s con ibu ion, we inco po a e
an asymme ic payo scheme in a simpli ied manage - i m scena io. Con a y o
s udies u ilising symme ic payo s (Pahlke e al., 2012) o single-pe son se ings
(Vieide , 2009), ou indings show ha jus i ica ion dec eases he p obabili y o
pa icipan s op ing o high- isk/high- e u n choices. Ou s udy also di e ges om
Pollmann e al. (2014) who employ a ewa d-based jus i ica ion app oach whe e
he decision-make does no ac i ely jus i y he decision o ou come. Ins ead, he
738
C.Lukas e al.
p incipal ewa ds ei he he decision o he ealised ou come, esul ing in dec eased
isk ole ance. In con as , ou esea ch explo es he in e ac i e e ec s o jus i ica-
ion and asymme ic payo schemes on p ojec choices. As pa o ha in es iga-
ion, we can demons a e ha jus i ica ion does no consis en ly align wi h he p in-
cipal’s bes in e es s. Conside ing he e idence ha decision-making on behal o
o he s is ypically linked wi h a dec ease in loss a e sion (e.g., Chak a a y e al.,
2011; Polman, 2012; Ande sson e al., 2016), one migh an icipa e he iskies deci-
sions in ou expe imen al condi ions wi h jus i ica ion. Howe e , ou obse a ions
e eal a mo e o less opposi e ou come.
Conce ning he second con ibu ion, we explo e he e ec s o di e en “ igge -
ing e en s” o jus i ica ions on decisions. I is an empi ical ques ion i i makes a
di e ence o decision-making whe he he decision i sel , low ou comes o losses
ollowing a decision igge he jus i ica ion equi emen . P io li e a u e ocuses
on decision jus i ica ion (Vieide , 2011; Feh enbache e al., 2020). We in es iga e
low ou come jus i ica ion and loss jus i ica ion in addi ion o decision jus i ica ion.
Gi en ou esul s, he a ious jus i ica ion egimes appea o in luence p ojec selec-
ion and compliance wi h p ojec ecommenda ions om supe iso s in dis inc
manne s. Ou expe imen inds ha o highe isk s a egies, decision jus i ica ion
esul s in highe compliance a es han jus i ica ions o ou comes. This esul may
ha e implica ions o business as i sugges s ha i ms could bes align managemen
decisions wi h he i ms’ p e e ed s a egy by equi ing a jus i ica ion o decisions
a he han ou comes o decisions. Consequen ly, ou indings con ibu e o com-
p ehending di e se (jus i ica ion- ela ed) managemen con ols, as ou lined by Me -
chan and O ley (2007).
The es o his pape is o ganised as ollows. The subsequen sec ion p o ides an
o e iew o ela ed li e a u e. Sec ion3 es ablishes he heo e ical amewo k and
in oduces a s aigh o wa d model ha can de i e es able hypo heses. Sec ion 4
ou lines he expe imen , and Sec .5 p esen s i s esul s. The concluding discussion
in Sec .6 e alua es he indings o ou s udy.
2 Rela ed li e a u e
The body o li e a u e on jus i ica ion p essu e and jus i ica ion e ec s is expand-
ing, as summa ised by Pa il e al. (2014).2 Ou s udy aligns wi h he expe imen al
li e a u e explo ing jus i ica ion o choices in ol ing isky al e na i es o p ojec s.
In pa icula , i is ela ed o he wo k o Vieide (2009), Pahlke e al. (2012), and
Pahlke e al. (2015). Vieide (2009) obse es in a single-pe son se ing ha loss
a e sion diminishes i he decision-make is equi ed o explain hei choice a e
he ac . In essence, jus i ica ion aises he p obabili y o making isky decisions
ha could esul in losses. Simila indings a e epo ed in Pahlke e al. (2012) and
Pahlke e al. (2015), who employ a wo-pe son se ing wi h symme ic payo s and
2 The e is a ela ion be ween jus i ica ion and accoun abili y. Le ne and Te lock (1999) e e o accoun -
abili y as he expec a ion ha indi iduals may be obliged o jus i y hei ac ions o o he s.
739
Explo ing decision‑making: expe imen al obse a ions on…
show ha jus i ica ion educes loss a e sion while lea ing o he elemen s o isk
a i ude una ec ed. Symme ic payo s imply ha he pe son who decides and he
o he passi e ecipien ecei e he same yield o each po en ial choice. The asym-
me ic payo s uc u e in ou pape enables us o also examine how a ia ions in he
agen ’s pay o pe o mance a ec isky p ojec choices unde jus i ica ion p essu e.
By inco po a ing a ious ypes o jus i ica ion, ou s udy is ela ed o he li e a-
u e analysing i and how di e en jus i ica ion a angemen s in e ac and whe he
hey a ec decision-making di e en ly. Siegel-Jacobs and Ya es (1996) expe imen-
ally in es iga ed how di e en jus i ica ion ypes, ou come jus i ica ion and deci-
sion jus i ica ion, a ec judgemen s abou indi idual a i udes o o he pe sons. The
esul s sugges ha decision jus i ica ion incen i ises people o conside ele an
in o ma ion in mo e de ail, while ou come jus i ica ion only p oduces addi ional
noise in he indi iduals’ judgemen s.
Jus i ica ion a angemen s a e also pa icula ly ele an in p ojec managemen .
Mac Donald e al. (2020) conduc ed in e iews wi h p ojec manage s and e ealed
ha hese manage s unde go a ious e ec s o jus i ica ion. In esponse o he need
o jus i ica ion, p ojec manage s cul i a e skills o acili a e decisions, an icipa e
p oblems, and manage mul iple p io i ies. Mi and Rezania (2023) analysed da a
om a su ey o p ojec manage s. They show ha p ojec manage s’ jus i ica ion
o he p ojec managemen decision p ocess mode a es he e ec o he manage s’
in e ac i e use o p ojec managemen con ol sys ems on p ojec pe o mance ia
eam lea ning. While Leong (1991) conside s jus i ica ion o he decision p ocess o
p ojec managemen and jus i ica ion o p ojec ou comes as accoun abili y a ange-
men s ha a e unning pa allel, Rezania e al. (2019) obse e di e ences in he
s eng h o bo h jus i ica ion ypes among o ganisa ions.
Finally, ou in es iga ion aligns wi h s udies explo ing bo h jus i ica ion’s
posi i e and nega i e e ec s on decision-making. Va ious expe imen al s udies a
he indi idual le el ha e ex ensi ely documen ed he posi i e impac s o jus i ica-
ion. Fo ins ance, jus i ica ion p essu e has been shown o dec ease p e e ence
e e sals (Vieide , 2011), mi iga e o e con idence o o de e ec s (Ash on, 1990;
Je mias, 2006), diminish he in luence o posi i e a ec i e eac ions (Feh en-
bache e al., 2020), and add ess inaccu a e judgemen s and a ou i ism (Ash on,
1992; Bauch & Weißenbe ge , 2020). Addi ionally, s udies by Webb (2002) and
A nold (2015) in es iga ed he impac o pe cei ed p essu e o jus i y decisions
and inancial p essu e on budge ing decisions. They obse e ha such p essu e
ends o educe slack and enhance coope a ion in decision-making p ocesses. The
expe imen al esul s o Ash on (1990) indica e ha pe o mance p essu e may
ha m o imp o e pe o mance depending on he exis ence o a decision aid. The
a ailabili y o a decision aid can hu pe o mance as i changes he na u e o he
decision make ’s ask. As we add he p incipal’s ecommenda ion o he p ojec
choice o ou expe imen , we also conside some “decision aid” in ou in es i-
ga ion. Howe e , he decision aid is no based on a s a is ical eg ession like in
Ash on (1990) bu i is p o ided by he supe io who has own in e es s. Thus, we
complemen he abo e indings by analysing how a decision aid om a supe io
who induces he jus i ica ion p essu e a ec s he decision make ’s choice.
740
C.Lukas e al.
3 Fo mula ion o model andhypo hesis de elopmen
Theo e ical amewo k In business, nume ous i ms a e o e seen by manage s
a he han hei owne s, g an ing hese manage s signi ican decision-making
au ho i y. Fi ms employ a ious managemen con ols o ensu e ha manage s’
ac ions align wi h he owne s’ objec i es. Me chan and Van de S ede (2007)
s a e ha hese con ols can be ca ego ised in o esul s, ac ion, pe sonnel, and
cul u al con ols. The in e play o hese con ols is essen ial, and when i ms u i-
lise di e en con ols while conside ing po en ial in e ac ions, hey employ man-
agemen con ol as a sys em (G abne & Moe s, 2013).
Resul s con ols a e widesp ead and commonly mani es as ou come-con in-
gen pay o managemen . Ac ion con ols include delega ing decision igh s
o manda ing jus i ica ions o decisions and ou comes. Pe sonnel con ols a e
means designed o communica e he i m’s expec a ions o wha i “wan s,” while
cul u al con ols in ol e elemen s such as shaping he o ganisa ion’s iden i y.
In ou model, a i m engages a manage and p esen s a con ac ea u ing ou -
come-con ingen pay. The manage aces a decision be ween wo dis inc p ojec s.
The selec ion o a p ojec , coupled wi h a andom s a e o na u e, de e mines he
p ojec ’s ou come and, consequen ly, he payo s o bo h he manage and he
i m. Ou model inco po a es and in eg a es h ee ypes o con ols: a iable pay
con ingen on he p ojec ’s ou come, se ing as a esul s con ol; jus i ica ion,
unc ioning as an ac ion con ol; and a p ojec ecommenda ion, ope a ing as a
pe sonnel con ol. In o mula ing he managemen con ol sys em, he model i m
conside s he in e ac ion be ween hese con ols. The i m’s p ima y aim is o
p omp he managemen o make he desi ed p ojec choice a he lowes possible
cos . The unde lying amewo k o he model d aws on agency heo y, whe e we
designa e he i m as he p incipal and he manage as he agen .
The se ing We examine a p incipal-agen scena io in which he p incipal del-
ega es he decision ega ding a p ojec o an agen . I is assumed ha bo h con-
ac ing pa ies a e isk-neu al. The ou come, deno ed as
xi
o P ojec
i=A,B
, is
con ingen on he ealised s a e o na u e. The e a e wo equally likely s a es o
na u e deno ed as: {s a e 1, s a e 2}. I s a e 1 occu s, he ou come o P ojec i is
Li
, while in s a e 2 i is
Hi
, wi h
Hi>Li
. We ca ego ise P ojec A as he “s anda d
p ojec ” and P ojec B as he “ isiona y p ojec .” In he e en o s a e 1 ealisa-
ion, he g oss ou come o he isiona y p ojec o he p incipal is less han ha
o he s anda d p ojec ,
LB<LA
. Addi ionally, we assume ha P ojec B has a
highe expec ed ou come and highe ou come a iance. Hence,
wi h
HB
>
HA
.
The agen ’s compensa ion, deno ed as
wi
and con ingen on he selec ion o P ojec
i, co esponds o a bonus con ac , a common p ac ice in many i ms. Speci ically, he
compensa ion unc ion
wi=( ,si)
comp ises a ixed paymen and a bonus ( a e)
si
.
E
(
xA
)
=0.5
(
HA+LA
)
<0.5
(
HB+LB
)
=E
(
xB
)
Va
(x
A
)=0.25(H
A
−L
A
)2<0.25(H
B
−L
B
)2=Va (x
B
)
,
741
Explo ing decision‑making: expe imen al obse a ions on…
The ixed paymen emains una ec ed by he p ojec choice o ou come. Alongside
he ixed paymen , he agen is en i led o a bonus o
si
mone a y uni s pe uni o he
p ojec ou come, p o ided he ou come is posi i e. This a angemen implies he use
o esul s con ols by he p incipal. The bonus a e
si
can be in e p e ed as he agen ’s
sha e o he (posi i e) ou come o P ojec i.
We assume
LB
<
LA=0
and
Hi
>
0
o
i=A,B
o s eamline he analysis. In o -
mal e ms, he compensa ion con ac o he agen comp ises a menu o wo compen-
sa ion unc ions
{wA,wB}
om which he agen chooses one by implemen ing P ojec
i. The e o e, he agen ’s expec ed compensa ion, gi en p ojec choice i is exp essed as:
We assume ha he agen is p o ec ed by limi ed liabili y, necessi a ing ha
si≥0
and
≥0
mus be sa is ied.
Jus i ica ion p essu e The agen is asked wi h he esponsibili y o p ojec selec ion.
In esponse, he p incipal equi es a jus i ica ion om he agen , con ingen on poo
p ojec pe o mance o o he chosen p ojec . The jus i ica ion se es as an ac ion con-
ol implemen ed by he p incipal. In p ac ice e ms, he jus i ica ion p o ides he agen
wi h a chance o cla i y de ia ions om an icipa ed esul s, elabo a e on ac o s ha
impeded he p ojec ’s p ope implemen a ion, o p o ide e idence suppo ing he iew
ha a pa icula p ojec choice was op imal based on p e-p ojec analysis. While his
may lead o posi i e ou comes o bo h he agen and he company, he p ocess unde-
niably induces s ess and discom o o he agen (Messne , 2009; F imanson e al.,
2021). S ess is no ably p obable when an explana ion is equi ed ollowing subpa pe -
o mance. Ne e heless, e en i he jus i ica ion pe ains o he decision a he han i s
ou come, he e is a basis o assume ha he ul ima e esul in luences how he agen ’s
decision is assessed by supe io s o sha eholde s (Lipe, 1993).
When he agen decides while he ou come emains unce ain, he agen will con-
side he po en ial o subop imal p ojec pe o mance. We deno e he psychological
s ess and e o associa ed wi h jus i ica ion as jus i ica ion p essu e
JPi
o P ojec i.
JPi
signi ies he expec ed cos s o jus i ica ion, and is exp essed as ollows:
The unc ion
JC(xi),xi=Li,Hi
, ep esen s he jus i ica ion cos s when he ou come
xi
occu s a e choosing P ojec i. The indica o a iable
𝜄
x
i∈{0, 1}
is ze o i he
p incipal does no eques a jus i ica ion o ou come
xi
; i a jus i ica ion is equi ed,
𝜄
x
i
=1
. In ui ion sugges s ha
JC(xi)
is a mono one dec easing unc ion - he highe
(1)
E(wi)= +0.5siHi.
(2)
JP
i
=0.5
[
𝜄
Hi
JC(H
i
)+𝜄
Li
JC(L
i
)
].
742
C.Lukas e al.
he ealised ou come
xi
he less s ess ul i is o jus i y he esul o decision.3 Es ab-
lishing he ollowing ela ions hen is s aigh o wa d:
I he p incipal equi es a jus i ica ion o a loss,
𝜄LB
=1
; i a low ou come neces-
si a es jus i ica ion,
𝜄LA
=
𝜄
LB
=1
; all o he indica o a iables a e se o ze o. Gi en
Eqs.(3) and (4), i is hen easy o es ablish he ollowing ela ionship:
I he agen is equi ed o jus i y a loss o a low ou come, he an icipa ed jus i ica ion
cos s o choosing op ion B a e consis en ly highe han o op ion A. Fu he mo e,
suppose
JC(xi)
is su icien ly s ic ly con ex, indica ing ha jus i ying he po en ial
loss om P ojec B is su icien ly s ess ul. In ha case, he inequali y in (5) emains
alid e en when he agen has o jus i y he p ojec choice ega dless o he ou come.
I he p incipal can indica e a p e e ence o a pa icula p ojec using pe son-
nel con ols, his will p obably impac jus i ica ion cos s. Fo ins ance, i he p inci-
pal p e e s en u es wi h highe isk, such as P ojec B in ou model, jus i ying he
po en ial loss om P ojec B is likely less bu densome. In a b oade sense, when
he p incipal signals a p e e ence o P ojec i, jus i ying he selec ion o i o a spe-
ci ic ou come is expec ed o be less s ess ul han insi ua ions wi hou he signaled
p e e ence. Con e sely, indica ing a p e e ence o P ojec j should aise he cos s o
jus i ying i. To in eg a e he impac o he ecommenda ion, we modi y Eq.(2) as
ollows:
He e,
RECj
indica es he p incipal’s p e e ence o ecommenda ion o P ojec j,
communica ed o he agen . The a iable
𝜌ji
e lec s he e ec o he ecommenda-
ion. We assume
0
<𝜌
AA =
𝜌
BB
<
1
. I he selec ed P ojec i aligns wi h he ec-
ommended P ojec j, meaning
i=j
, i esul s in a e ical downwa d shi o he
ini ial jus i ica ion cos unc ion. Con e sely, i he agen does no choose he ecom-
mended P ojec j, jus i ica ion cos s inc ease,
𝜌AB =
𝜌
BA
>
1
. Following his eason-
ing, he subsequen ela ions a e de i ed:
(3)
JC(LB)>JC(LA),
(4)
JC(HB)
<
JC(HA).
(5)
JPB
>
JPA.
(6)
JP
i(RECj)=0.5
[
𝜄H
i
𝜌jiJC(Hi)+𝜄L
i
𝜌jiJC(Li)
],
3 An al e na i e a ionale o a dec easing jus i ica ion cos unc ion s ems om he p inciple o loss
a e sion. I is widely acknowledged ha indi iduals end o loss a e sion, whe ein nega i e (mone a y)
alues ca y g ea e weigh han posi i e alues o equal magni ude (B ink & Rankin, 2013; Sawe s
e al., 2011; T e sky & Kahneman, 1991) Consequen ly, agen s, pe cei ing a loss o be mo e impac ul
on he p incipal han he co esponding gain, ace highe jus i ica ion cos s when jus i ying a loss han a
gain o equi alen magni ude. Gi en ha
LB
<
LA=0
, whe e he low ou come unde P ojec B signi ies
a loss, he jus i ica ion cos unc ion
JC(xi)
would exhibi a con ex dec easing end caused by loss a e -
sion.
749
Explo ing decision‑making: expe imen al obse a ions on…
o he de e minan s o jus i ica ion p essu e. Figu e1 p o ides a summa y o he
expe imen .
In each egime, we implemen he jus i ica ion manipula ion h ough a com-
pu e cha . The agen is manda ed o en e he jus i ica ion, and he p incipal has he
oppo uni y o espond. We (co ec ly) an icipa e ha agen pa icipan s expe ience
su icien discom o when hey jus i y hemsel es.
Ou second manipula ion seeks o unde s and he e ec s o communica ion o
owne s’ p e e ences ( ecommenda ions) o p ojec decisions. This is execu ed by
enabling up on communica ion o he p incipal’s p e e ence in ea men s REC,
LOS-REC, DEC-REC, and LOW-REC. Technically, in e e y decision ound, he
p incipal’s p ojec p e e ence is p esen ed on he agen ’s compu e sc een be o e he
agen makes a p ojec decision. The agen has he libe y o ei he adhe e o o dis-
ega d he ecommenda ion. No addi ional communica ion, such as h ough cha , is
allowed.
Ou hi d manipula ion in ol es a ying he p incipal and agen payo sha es i
P ojec B is chosen (Table3 in he appendix). The escala ing ou come sha e o he
agen e lec s he inc ease in a iable pay ha migh be necessa y o incen i ise p o-
jec s wi h highe expec ed e u ns (and e u n a iance). This a ia ion allows us o
po ay P ojec B as ela i ely mo e a ac i e o he agen han P ojec A ega ding
expec ed compensa ion. The manipula ion o ou come sha es enhances he in e nal
alidi y o he expe imen .
Implemen ing he h ee manipula ions, he ini ial subs an ial pa o ou analysis
concen a es on examining how hese manipula ions in luence he agen s’ equen-
cies o selec ing he p ojec wi h highe isk and e u n. Co espondingly, he num-
be o agen s’ selec ions o P ojec B (ChoiceB) is he p ima y a iable o in e es .
The second signi ican aspec o ou in es iga ion cen e s on he psychological
e ec s o he manipula ions, speci ically, jus i ica ion p essu e (
JPi
in he model).
Fig. 1 O e iew o he expe imen ’s se up and ea men s (bold solid a ows = ea men s LOW, REC-
LOW; bold na owly dashed a ows = ea men s DEC, REC-DEC; bold widely dashed a ow = ea -
men s LOS, REC-LOS)
750
C.Lukas e al.
In line wi h he de ini ion o Le ne and Te lock (1999), we posi ha only agen s in
jus i ica ion condi ions expe ience jus i ica ion cos s. To assess i s pe cep ion and
ex en , we examine da a om a compu e ised pos -expe imen al ques ionnai e ha
manda es agen s in jus i ica ion condi ions o epo hei expe ience wi h he jus i i-
ca ion equi emen s. As measu emen s a e each decision likely in luence beha io
in subsequen ounds, we collec ed he da a a e comple ing he expe imen . The
ques ionnai e comp ises se en ques ions whe e subjec s a e hei expe ience wi h
jus i ica ion on a 9-poin scale. We calcula e C onbach’s alpha coe icien o e alua e
whe he he i ems a e concep ually ela ed. Wi h a alue o 0.84, he in e nal con-
sis ency o ou ques ionnai e’s scale is conside ed good. Following s anda d p ac-
ice, we ope a ionalise he cons uc wi h he jus i ica ion p essu e a iable JP, ep-
esen ing he a e age sco es agen s achie e on he co esponding i ems.
To ensu e he in e nal alidi y o ou expe imen , we employ s anda dised ques-
ionnai es ha a e acknowledged o hei eliabili y and alidi y. Addi ionally, we
u ilise a ious es s and eg ession models o e i y he obus ness o he esul s.
Se e al con ols complemen he da a o ou main a iables. Alongside sociodemo-
g aphic in o ma ion (Age, Sex), we ga he da a on he subjec s’ isk a i ude (Will-
Risk) h ough a p e-expe imen al ques ionnai e. The ques ionnai e is cons uc ed
based on he Ge man Socio-Economic Panel (SOEP). Fu he mo e, we conside
inequali y a e sion since o he - ega ding p e e ences could impac decision-making
in ou expe imen . We asce ain he a iable o inequali y a e sion (Inequali yF)
using a es om Fo in e al. (2007), ha ca ego ises pa icipan s in o h ee classes
o inequali y a e sion (low, medium, and high).
O he in luences ha me i conside a ion s em om he epea ed in e ac ion o
he pa icipan s (mul iple decision se s, ixed ma ching). The i s such in luence is
epu a ion building. Fo ins ance, as he p incipal obse es he agen ’s ack eco d
o decisions, i could be possible ha he agen aims o shape he p incipal’s pe -
cep ion o he /him. Howe e , we do no belie e ha agen s ha e he incen i e o
manage hei epu a ion. I he agen is unawa e o he p incipal’s p e e ence, i is
unclea which image is “ igh .” E en i he p incipal’s ecommenda ion (p e e ence)
is known, i is unce ain whe he an image as a “B-decision make ” o “A-decision
make ” enhances he agen ’s u ili y.
As a second ac o , we examine he pa icipan s’ po en ial o coope a i e o
e alia o y beha io in cha communica ion ea men s. Acco dingly, we analyse
cha con en s o di e en ia e be ween undesi ed communica ion and beha io s in lu-
enced by jus i ica ion, ac ual ecommenda ions, o he pay scheme. An ins ance o
undesi ed collusion is a p incipal communica ing he /his p e e ence o he agen
h ough he jus i ica ion cha in ea men s wi h jus i ica ion bu wi hou a ecom-
menda ion. Ano he example in ol es p incipals and agen s e ealing hei names
ia cha and ag eeing o sha e he payo a e he expe imen . Iden i ying suspicious
cha s o a p incipal-agen pai in a gi en ound excludes all obse a ions a e he
751
Explo ing decision‑making: expe imen al obse a ions on…
pa icula ound.6 In o al, 76 ou o 1080 agen - ound obse a ions a e elimina ed
due o collusi e beha io . The majo i y o elimina ions a e iden i ied unde deci-
sion jus i ica ion (36), ollowed by low ou come jus i ica ion (21) and loss jus i ica-
ion (19). We a ibu e he a ia ion be ween he ea men s o he ac ha he e is
he mos equen oppo uni y o collusion unde decision jus i ica ion, ollowed by
low ou come jus i ica ion and loss jus i ica ion. No dispu es o ac s o e alia ion a e
iden i ied.
The expe imen was conduc ed in he Leibniz Labo a o y o Expe imen al Eco-
nomics a Leibniz Uni e si y Hanno e . The so wa e h oo (Bock e al., 2014)
was employed o o ganisa ion and adminis a ion p ocesses. The expe imen was
p og ammed using he so wa e zT ee (Fischbache , 2007). 360 unde g adua e and
g adua e s uden s om a ious ields pa icipa ed, esul ing in an o e all sample o
180 p incipal-agen pai s. The p opo ion o emale subjec s in he expe imen is
40.56%. The gende dis ibu ions o p incipals and agen s a e simila wi hin each
ea men and do no a y signi ican ly be ween ea men s. The highes p opo ion
o emale agen s is ound in he loss jus i ica ion condi ion (50%). On a e age, he
s uden s a e 23.84 yea s o age and a end cou ses in he 5 h semes e . Rega ding
con en , 49.72% o he pa icipan s a e en olled in STEM cou ses, while he es a e
dis ibu ed be ween economics and managemen , eaching, and some o he ields.
Expe imen al sessions las ed app oxima ely 60min, and ea nings a e aged 11.23
Eu o.7
A e a i ing in he lab, pa icipan s ecei ed w i en ins uc ions con aining all
ele an de ails abou he expe imen . A ideo ilm was played in which an expe i-
men e (who was no p esen du ing he expe imen ) ead he comple e ins uc ions,
and explana o y sc eensho s o he upcoming expe imen we e p esen ed. This was
ollowed by pa icipan s eading he w i en ins uc ions a hei assigned sea s.
Any cla i ying ques ions we e add essed a he pa icipan s’ sea s. P io o he com-
mencemen o he ac ual expe imen , pa icipan s we e equi ed o answe se e al
con ol ques ions o ensu e a ho ough unde s anding o he expe imen al si ua ion.
Ou expe imen pa icipan s ecei ed compensa ion h ough an ini ial endowmen o
30 Tale (3 Tale = 1 Eu o) and he payo om a speci ic decision ound. Ins ead
o all ounds, paying o only one ound was done o a oid weal h e ec s (Cha ness
e al., 2016). The ound ele an o he payo was publicly and andomly selec ed
a e he expe imen . The ini ial endowmen o each pa icipan gua an eed ha he
sum o he ini ial endowmen and he payo in he ele an decision ound could no
be nega i e, meaning no pa icipan could incu a mone a y loss in he expe imen .
6 To main ain neu ali y, he p ocess o analysing cha con en s and il e ing obse a ions was conduc ed
by mul iple hi d pa ies. These indi iduals we e un amilia wi h he hypo heses and we e unin e es ed in
he expe imen ’s ou comes.
7 Addi ional de ails ega ding he da ase can be p o ided upon eques .
752
C.Lukas e al.
5 Resul s o heexpe imen
Conce ning Hypo hesis 1, we examine how decision-making is in luenced by jus i i-
ca ion. Figu e2 illus a es he P ojec -B choices o agen s wi hou ecommenda ion.8
Agen s choose he isky P ojec B less equen ly when jus i ica ion is p esen .
This sugges s ha he manipula ion in he expe imen (jus i ica ion yes/no) was
e ec i e. To de e mine whe he hese di e ences a e s a is ically signi ican , we
conduc ed a - es . The Shapi o-Wilk es con i ms no mali y o he co espond-
ing a iables. We ind he di e ence be ween agen s ope a ing unde any jus i ica-
ion (wi hou a ecommenda ion) and he baseline s a is ically signi ican ( - es :
=2.511
,
d =522
,
p=0.012
). This esul p o ides ini ial suppo o Hypo hesis
1.
Asse ing ha he bes decision aligns wi h he p incipal’s in e es , no neces-
sa ily maximising expec ed alue, we con ol o he p incipals’ p e e ences when
assessing he impac o jus i ica ion. Al hough ou model p edic s ha p incipals
p e e P ojec B, i is easonable o assume ha he e migh be p incipals a o ing
he less isky in es men (P ojec A). Again, ocusing only on ea men s wi hou a
ecommenda ion, Fig.3 illus a es he agen s’ decisions.
I appea s ha jus i ica ion educes agen s’ selec ions o P ojec B e en when
p incipals p e e he isky P ojec B ( - es :
=2.097
,
d =265
,
p=0.037
). When
conside ing p incipal-agen pai s whe e he p incipals p e e P ojec A, he e ec
o jus i ica ion on P ojec B choices shows a simila end bu does no each a
Fig. 2 Desc ip i e s a is ics o P ojec B choices (ChoiceB) o agen s unde no (Jus Dum=0) and any
jus i ica ion (Jus Dum=1), excluding ea men s wi h ecommenda ions. The ba cha displays he mean
a es a which agen s choose P ojec B in a ious decision se s. The pe cen age alues abo e each ba
indica e he a e age selec ion a es o P ojec B ac oss all decision se s. The numbe o obse a ions
is p esen ed a he bo om o each ba . E o ba s a e included o ep esen he 95% con idence in e als
8 A comp ehensi e compila ion o ound-by- ound esul s is a ailable in he appendix, speci ically in
Tables4 and 5.
753
Explo ing decision‑making: expe imen al obse a ions on…
signi ican le el ( - es :
=−1.367
,
d =255
,
p=0.173
). In a p elimina y sum-
ma y, i seems ha jus i ica ion gene ally esul s in less isky choices by agen s.
When p incipals p e e highe - isk (in es men ) s a egies wi hou he agen s being
awa e o his p e e ence, he jus i ica ion wo ks agains he in e es s o he p inci-
pals. Howe e , when p incipals p e e lowe - isk s a egies, jus i ica ion does no
appea o wo k agains hei in e es s.
To be e unde s and he agen s’ decisions, we pe o med logis ic eg ession anal-
yses9 o he agen s’ choices o P ojec B and Table2 p esen s he esul s. Conce n-
ing Hypo hesis 1, models (1) and (2) a e pe inen , ocusing on agen s who mus
jus i y decisions o ou comes wi hou con olling o he ype o jus i ica ion. While
model (1) includes he a iables o in e es , model (2) also inco po a es a ious con-
ol measu es.
The odds a ios o Jus Dum indica e a educ ion in he likelihood o a P ojec
B choice due o jus i ica ion. Fo example, he a io o 0.542 sugges s ha unde
jus i ica ion (compa ed o he baseline), we an icipa e inding only 0.542 agen s
selec ing P ojec B o e e y agen choosing P ojec A. This magni ude emains
Fig. 3 Desc ip i e s a is ics o P ojec B selec ions o agen s ope a ing unde no jus i ica ion o any
jus i ica ion (no dis inc ion be ween ypes), con olling o uncommunica ed p ojec p e e ences o p in-
cipals. Ba cha s, di e en ia ed by ligh e (no jus i ica ion) and da ke (wi h jus i ica ion) g ay, illus-
a e he impac on P ojec Bchoices (ChoiceB) by agen s in ea men s wi hou (Jus Dum = 0) and wi h
(Jus Dum = 1) jus i ica ion. The wo le ba cha s depic he impac o jus i ica ion on P ojec B choices
(ChoiceB) agen s make when hei p incipals p e e P ojec B (P e B). The wo ba cha s on he igh
show he e ec on agen s’ P ojec B choices (ChoiceB) when p incipals p e e P ojec A (P e A). T ea -
men s wi h ecommenda ions, whe e p incipal p e e ences a e communica ed, a e excluded. The pe cen -
age numbe s abo e each ba ep esen he mean a es o selec ing P ojec B ac oss all decision se s, wi h
he numbe o obse a ions displayed a he bo om o each ba . E o ba s a e included o ep esen he
95% con idence in e als
9 Non-linea eg ession analysis is equen ly used in psychology esea ch when he dependen a iable
is bina y (Gomila, 2021). Mul iple linea eg ession models con i m ou esul s. See also Woold idge
(2002) o insigh s in o app op ia e models.
754
C.Lukas e al.
Table 2 RE logi models on he
agen s’ P ojec B-choices
Random e ec s logi models on he agen ’s P ojec Bchoices includ-
ing da a om all decision se s and ea men s. The dependen a ia-
ble ChoiceB is a bina y a iable signaling o each ound i he agen
selec s P ojec B. Jus Dum is a bina y a iable signaling whe he an
agen mus jus i y he decision o ou come (independen o he ype
ChoiceB (1) (2)
Jus Dum 0.542 0.624
(0.206) (0.222)
Rec
A 0.424* 0.537
(0.201) (0.252)
B 26.052*** 28.910***
(20.727) (21.275)
Jus Dum#Rec
1 A 0.647 0.421
(0.403) (0.259)
1 B 0.341 0.233*
(0.312) (0.200)
E[Sha eB]
40% 1.994** 2.001**
(0.647) (0.650)
45% 2.057** 2.032**
(0.593) (0.587)
60% 7.391*** 7.438***
(3.044) (3.056)
80% 57.813*** 60.535***
(50.735) (50.850)
Va [Sha eB] 0.975*** 0.974***
(0.006) (0.006)
WillRisk 1.123
(0.081)
Inequali yF
Medium 0.171***
(0.113)
High 0.201***
(0.078)
Age 0.982
(0.023)
Male 1.596*
(0.435)
Semes e 1.093**
(0.044)
Obse a ions (n) 1.004 1.004
Wald chi2 79.11 101.37
P ob > chi2
<0.001
<0.001
755
Explo ing decision‑making: expe imen al obse a ions on…
consis en when con olling o indi idual cha ac e is ics such as isk a i ude and
sociodemog aphic de ails. Bo h models na owly miss he signi icance h eshold
o he main e ec , howe e , he in e ac ion be ween Jus Dum and Rec eaches
s a is ical signi icance in Model (2). I seems ha jus i ica ion signi ican ly a ec s
p ojec decisions when ecommenda ions ha e been made be o ehand. Figu e4
p o ides an o e iew o hese e ec s.
We no ice ha he dec ease in p edic ed p obabili ies is mo e p onounced a e
a ecommenda ion ( iangle and squa e ends) han wi hou a ecommenda ion
(ci cle ends). By calcula ing he ma gins, we disco e ha jus i ica ion educes
he p edic ed p obabili y o a P ojec B choice a e a P ojec B ecommenda ion
o jus i ica ion). Rec is a ca ego ical a iable indica ing whe he he
agen ecei es a ecommenda ion o P ojec A o B. I s base le el
is “no ecommenda ion.” E[Sha eB] is a ca ego ical a iable indi-
ca ing he agen ’s expec ed p o i sha e (in pe cen ) in s a e 2 when
selec ing P ojec B. I s base le el is “30%.” Va [Sha eB] indica es
he a iance o he agen ’s payo sha e (in absolu e numbe s) when
selec ing P ojec B. Con ol a iables a e he subjec s’ age (Age), sex
(Male), s udy p og ess (Semes e ), inequali y a e sion (Inequali yF),
and isk a i ude (WillRisk). Coe icien s a e p esen ed in exponen i-
a ed o m, i.e., as odds a ios. S anda d e o s (in pa en heses) a e
clus e ed a he indi idual le el. The cons an s a e included in he
models bu no epo ed
*
p
<
0.1
**
p
<
0.05
***
p
<
0.01
Table 2 (con inued)
Fig. 4 Ma ginsplo o p edic ed p obabili ies o P ojec B choices by agen s unde no jus i ica ion o any
jus i ica ion (no dis inc ion be ween ypes), including P ojec B and P ojec A ecommenda ions. The
line wi h iangles (squa es) depic s p edic ed p obabili ies a e P ojec B (P ojec A) ecommenda ions.
The line wi h ci cles illus a es p obabili ies o agen s ecei ing no ecommenda ions. The pe cen age
numbe s nex o each line ep esen he co esponding p edic ed p obabili ies. The analysis is based on
1.004 obse a ions, and e o ba s a e included o ep esen 95% con idence in e als
756
C.Lukas e al.
by
15.36%
(
z=−2.590
,
p=0.010
) and by
21, 52%
(
z=−2.670
,
p=0.008
) a e
a P ojec A ecommenda ion.
Based on he esul s om ou eg ession models and es s, we iden i y jus i i-
ca ion as a c ucial ac o in he agen s’ decision-making. In line wi h he p edic-
ion o he heo e ical model, jus i ica ion seems o diminish he a ac i eness
o p ojec s wi h high isks and e u ns. This e ec appea s o pe sis e en in he
p esence o ecommenda ions. Thus, o Hypo hesis 1, we s a e:
Resul 1 Expe imen al e idence suppo s Hypo hesis 1, jus i ica ion leads o a less
equen choice o he high- isk/high- e u n p ojec (P ojec B).
Shi ing ou ocus o Hypo hesis 2, Fig.5 illus a es desc ip i e da a. Depend-
ing on he p esence o absence o a jus i ica ion equi emen , i aligns he P ojec
B choices o agen s in ea men s wi hou ecommenda ions wi h he P ojec B
choices o agen s in ea men s wi h P ojec B ecommenda ions. Ou heo e ical
amewo k posi s ha when he p incipal makes a P ojec B ecommenda ion, i
signals a willingness o bea losses o he agen . This, in u n, esul s in lowe jus-
i ica ion p essu e and has he po en ial o coun e ac he nega i e e ec o jus i i-
ca ion on P ojec B choices, as p edic ed in Hypo hesis 1 and documen ed abo e.
Fig. 5 Desc ip i e s a is ics o P ojec B choices (ChoiceB) o agen s wi h and wi hou P ojec B ecom-
menda ions. Ba cha s, di e en ia ed by ligh (wi hou ecommenda ions) and da k (wi h ecommenda-
ions) g ay, depic he impac o P ojec B ecommenda ions (RecDum=0 s. Rec=B). The pe cen age
alues abo e each ba ep esen he o e all a es a which agen s choose P ojec B ac oss all decision
se s, wi h he numbe o obse a ions lis ed a he bo om o each ba . E o ba s ep esen he 95% con i-
dence in e als. The wo ba cha s demons a e he e ec o P ojec B ecommenda ions when agen s a e
no equi ed o jus i y decisions o ou comes (Jus Dum=0). A he same ime, he wo ba cha s on he
igh show he e ec unde jus i ica ion (Jus Dum=1). The plo s do no include he decisions o agen s
ecei ing a ecommenda ion o choose he less isky P ojec A
757
Explo ing decision‑making: expe imen al obse a ions on…
E en hough ecommenda ions a e no legally binding, i is e iden ha agen s
end o ollow hem. We de ine his endency as compliance. Fo ins ance, a ound
95% o pa icipan s chose P ojec B a e ecei ing he co esponding ecommenda-
ion in he REC ea men , whe eas only 58% op ed o P ojec B in he BL (see Rec-
Dum=0 s. Rec=B compa ison when Jus Dum=0). These di e ences a e s a is i-
cally signi ican ( - es s:
=−6.287
,
d =245
,
p<0.001
(Jus Dum=0);
=−8.022
,
d =506
,
p<0.001
(Jus Dum=1)).
Table 2 p esen s addi ional e idence ega ding he in luence o ecommenda-
ions: he coe icien s associa ed wi h P ojec B ecommenda ions ou weigh hose o
all o he decision d i e s.10 Fo Hypo hesis 2, we hus s a e:
Resul 2 Expe imen al e idence suppo s Hypo hesis 2, indica ing ha ecommen-
da ions om p incipals a o ing P ojec B inc ease he possibili y o agen s choos-
ing he high- isk/high- e u n p ojec (P ojec B).
Conce ning Hypo hesis 3, inc easing he agen ’s ou come sha e, i.e., ewa d-
ing he high ou come wi h a la ge bonus, is expec ed o boos he p obabili y
o P ojec B choices. Acco ding o ou model, a highe bonus compensa es o
he agen ’s jus i ica ion p essu e and he discom o po en ially causing losses
on he p incipal. This mechanism could shi he balance in a o o P ojec B.
Obse a ions om he expe imen align wi h his idea, as e iden om Table2,
whe e he expec ed bonuses o P ojec B s a is ically signi ican ly mo i a e agen
Fig. 6 Desc ip i e s a is ics o pe cei ed jus i ica ion p essu e (JP) among all agen s unde a ious
ypes o jus i ica ion wi hou ecommenda ions: decision jus i ica ion (DEC), low ou come jus i ica ion
(LOW), and loss jus i ica ion (LOS). The numbe s abo e each ba ep esen he mean jus i ica ion p es-
su e sco es, wi h he numbe o obse a ions indica ed a he bo om o each ba . E o ba s a e included
o depic he 95% con idence in e als
10 Addi ionally, Table5 and Fig.8 in he appendix e eal ha ecommenda ions a o ing P ojec A om
some p incipals a e also e ec i e.
758
C.Lukas e al.
pa icipan s o choose P ojec B (indica ed by coe icien s o E[Sha eB]). The
logis ic eg essions in he baseline (BL) u he demons a e signi ican coe i-
cien s. This sugges s ha P ojec B choices (and he appa en “willingness” o
cause losses o he p incipal) a e po en ially in luenced by he agen ’s p o i
sha e, e en in he absence o jus i ica ion p essu e and ecommenda ions (Table6
in he appendix). We conclude as ollows:
Resul 3 Expe imen al e idence suppo s Hypo hesis 3, sugges ing ha a highe
p o i sha e o he agen mi iga es he impac o jus i ica ion and inc eases he like-
lihood o op ing o he high- isk/high- e u n p ojec (P ojec B).
To es Hypo hesis 4(a), we u ilise ou measu e o jus i ica ion p essu e o com-
pa e he sco es o agen s ope a ing unde di e en ypes o jus i ica ion. We p edic
he highes jus i ica ion p essu e sco e unde decision jus i ica ion o ei he p ojec
[see (14)-(15)]. The esul s o e e y ype o jus i ica ion a e p esen ed in Fig.6.
The jus i ica ion p essu e sco es a e oughly on he same le el o each o he
h ee jus i ica ion egimes. We ind g ea e a iance in pe cei ed jus i ica ion p es-
su e unde decision jus i ica ion compa ed o he low ou come and loss jus i ica-
ion. Gi en ha he Shapi o-Wilk es does no con i m no mali y o each a ia-
ble, non-pa ame ic es ing using he wo-sided Wilcoxon-Mann–Whi ney es o
independen samples (WMW) is employed. The esul s o his es con i m ha he
di e ences do no achie e s a is ically signi ican le els (WMW: DEC s. LOW:
z=−0.923
,
p=0.356
; DEC s. LOS:
z=−0.309
,
p=0.758
; LOW s. LOS:
Fig. 7 Desc ip i e s a is ics o p ojec choices o agen s ecei ing speci ic ecommenda ions unde a i-
ous ypes o jus i ica ion: no jus i ica ion (REC), decision jus i ica ion (DEC-REC), low ou come jus i i-
ca ion (LOW-REC), and loss jus i ica ion (LOS-REC). The pe cen age numbe s abo e each ba indica e
he mean compliance a es o agen s wi h he p incipals’ p ojec ecommenda ions ac oss all decision
se s. The numbe o obse a ions is p esen ed a he bo om o each ba , wi h e o ba s ep esen ing 95%
con idence in e als
765
Explo ing decision‑making: expe imen al obse a ions on…
Ins uc ions o he expe imen
766
C.Lukas e al.
Payo dis ibu ion
See Appendix Table3.
Table 3 Payo dis ibu ion (in expe imen al cu ency Tale ) be ween he p incipal and he agen o each
p ojec , s a e o na u e and decision se
The dis ibu ion is equal in all ea men s. Subjec s a e only able o obse e he payo dis ibu ion o he
cu en decision ound, no he dis ibu ion o p e ious o subsequen decision se s (i.e., he i s payo
dis ibu ion is only obse able in ound one, he second dis ibu ion only in ound wo, e c.). Expec ed
alues and a iances a e also no displayed in he expe imen
767
Explo ing decision‑making: expe imen al obse a ions on…
P ojec choices and p e e ences, ea men s wi hou ecommenda ions
See Appendix Table4.
Table 4 Agen s’ p ojec choices and p incipals’ p ojec p e e ences o each decision se in ea men s
wi hou ecommenda ions
BL e e s o baseline, DEC o decision jus i ica ion, LOW o low ou come jus i ica ion, and LOS o loss
jus i ica ion. Only agen s we e esponsible o p ojec selec ions. We also eco ded he p incipals’ p e -
e ences in each decision se in ea men s wi hou ecommenda ions. The uppe numbe s in each cell
display he absolu e equencies, he lowe numbe s illus a e he ela i e a es in pe cen . No e ha some
decisions o agen s a e elimina ed due o subjec s communica ing p e e ences o e he jus i ica ion cha
o engaging in o he kinds o illegal collusion (see oo no e 6). Thus, numbe s wi hin some ea men s
can change be ween decision se s
Agen s’
choices
BL LOS DEC LOW
P ojec A P ojec B P ojec A P ojec B P ojec A P ojec B P ojec A P ojec B
Se 1 15 13 21 9 7 12 12 11
53.57% 46.43% 70.00% 30.00% 36.84% 63.16% 52.17% 47.83%
Se 2 13 15 16 14 7 9 9 9
46.43% 53.57% 53.33% 46.67% 43.75% 56.25% 50.00% 50.00%
Se 3 7 21 13 15 1 10 6 13
25.00% 75.00% 46.43% 53.57% 09.09% 90.91% 31.58% 68.42%
Se 4 14 14 18 8 5 4 14 5
50.00% 50.00% 69.23% 30.77% 55.56% 44.44% 73.68% 26.32%
Se 5 11 17 17 7 6 7 12 8
39.29% 60.71% 70.83% 29.17% 46.15% 53.85% 60.00% 40.00%
Se 6 10 18 12 11 5 5 9 9
35.71% 64.29% 52.17% 47.83% 50.00% 50.00% 50.00% 50.00%
P incipals’
p e e ences
BL LOS DEC LOW
P ojec A P ojec B P ojec A P ojec B P ojec A P ojec B P ojec A P ojec B
Se 1 7 21 10 20 8 11 10 13
25.00% 75.00% 33.33% 66.67% 42.11% 57.89% 43.48% 56.52%
Se 2 12 16 13 17 4 15 11 12
42.86% 57.14% 43.33% 56.67% 21.05% 78.95% 60.87% 39.13%
Se 3 10 18 14 16 5 14 14 9
35.71% 64.29% 46.67% 53.33% 26.32% 73.68% 60.87% 39.13%
Se 4 13 15 14 16 10 9 13 10
46.43% 53.57% 46.67% 53.33% 52.63% 47.37% 56.52% 43.48%
Se 5 17 11 15 15 11 8 14 9
60.71% 39.29% 50.00% 50.00% 57.89% 42.11% 60.87% 39.13%
Se 6 20 8 15 15 12 7 20 3
71.43% 28.57% 50.00% 50.00% 63.16% 36.84% 86.96% 13.04%
768
C.Lukas e al.
P ojec choices and p e e ences, ea men s wi h ecommenda ions
See Appendix Fig.8 and Table5.
Fig. 8 Dis ibu ion o p ojec ecommenda ions (Rec) o p incipals in all ea men s. The pe cen age
numbe s abo e each ba ep esen he mean a es a which p incipals ecommended P ojec B ac oss all
decision se s. A he bo om o each ba s ands he numbe o obse a ions. The e o ba s ep esen 95%
con idence in e als
769
Explo ing decision‑making: expe imen al obse a ions on…
Va iables in es s and eg essions (1)
Va iable Desc ip ion
Jus Dum Dummy a iable = 1 i subjec ope a es in an jus i ica ion ea men (no dis inc ion
be ween he ypes o jus i ica ion)
Age Yea s o age
Table 5 Agen s’ p ojec choices and p incipals’ p ojec p e e ences o each decision se in ea men s
wi h ecommenda ions
REC e e s o he ea men wi h ecommenda ions, DEC-REC o decision jus i ica ion wi h ecommen-
da ions, LOW-REC o low ou come jus i ica ion wi h ecommenda ions, and LOS-REC o loss jus i i-
ca ion wi h ecommenda ions. Only agen s we e esponsible o p ojec selec ions. Fo ea men s wi h
ecommenda ions p incipals had he possibili y o make a ecommenda ion p io o he p ojec selec ion.
The uppe numbe s in each cell display he absolu e quan i ies, he lowe numbe s illus a e he ela i e
a es in pe cen
Agen s’
choices
REC LOS-REC DEC-REC LOW-REC
P ojec A P ojec B P ojec A P ojec B P ojec A P ojec B P ojec A P ojec B
Se 1 10 18 14 9 5 10 9 5
35.71% 64.29% 60.87% 39.13% 33.33% 66.67% 64.29% 35.71%
Se 2 8 20 9 14 2 13 7 7
28.57% 71.43% 39.13% 60.87% 13.33% 86.67% 50.00% 50.00%
Se 3 6 22 8 15 4 11 5 9
21.43% 78.57% 34.78% 65.22% 26.67% 73.33% 35.71% 64.29%
Se 4 12 16 13 10 6 9 9 5
42.86% 57.14% 56.52% 43.48% 40.00% 60.00% 64.29% 35.71%
Se 5 8 20 15 8 6 9 10 4
28.57% 71.43% 65.22% 34.78% 40.00% 60.00% 71.43% 28.57%
Se 6 11 17 13 10 3 12 9 5
39.29% 60.71% 56.52% 43.48% 20.00% 80.00% 64.29% 35.71%
P incipals’
p e e ences
REC LOS-REC DEC-REC LOW-REC
P ojec A P ojec B P ojec A P ojec B P ojec A P ojec B P ojec A P ojec B
Se 1 12 16 13 10 7 8 7 7
42.86% 57.14% 56.52% 43.48% 46.67% 53.33% 50.00% 50.00%
Se 2 9 19 10 13 5 10 5 9
32.14% 67.86% 43.48% 56.52% 33.33% 66.67% 35.71% 64.29%
Se 3 11 17 11 12 4 11 7 7
39.29% 60.71% 47.83% 52.17% 26.67% 73.33% 50.00% 50.0%
Se 4 21 7 15 8 6 9 10 4
75.00% 25.00% 65.22% 34.78% 40.00% 60.0% 71.43% 28.57%
Se 5 17 11 14 9 7 8 9 5
60.71% 39.29% 60.87% 39.13% 46.67% 53.33% 64.29% 35.71%
Se 6 19 9 17 6 4 11 9 5
67.86% 32.14% 73.91% 26.09% 26.67% 73.33% 64.29% 35.71%
770
C.Lukas e al.
Va iable Desc ip ion
ChoiceB Dummy a iable = 1 i he agen chooses P ojec B in a speci ic decision ound o he
expe imen
F eqRe-
ealedP in-
RiskA e
Numbe o imes he p incipal e ealed isk a e sion, i.e., p e e ences o a oiding he
isky p ojec , ia cha ( a iable is used o exclude cases om s a is ics, es s, and eg es-
sions)
F eqRe-
ealedP in-
RiskTol
Numbe o imes he p incipal e ealed isk ole ance, i.e., p e e ences o an in es men
in he isky p ojec , ia cha ( a iable is used o exclude cases om s a is ics, es s, and
eg essions)
ID1 Indi idual iden i ica ion numbe o subjec
ID2 Team iden i ica ion numbe o subjec
Inequali yF Le el o inequali y a e sion indica ed by a measu e om Fo in e al., 2007 (0 = low a e -
sion; 1 = medium a e sion; 2 = high a e sion)
JP Mean sel assessmen sco e in ques ions 1, 3, 5, 6, 8, and 14 o he pos -expe imen al
ques ionnai e conce ning he pe cep ion o jus i ica ion p essu e (see pos -expe imen al
ques ionnai e); i ems 10 o 13 we e excluded om ou cons uc s as hese ques ions
cap u e changes in jus i ica ion p essu e o e he cou se o he expe imen , a he han
he pe cep ion i sel ; mo eo e , i em 15 was elimina ed because i showed only small
co ela ions wi h he o he i ems and i s exclusion inc eased C onbach’s alpha
Male Dummy a iable = 1 i subjec = male
Va iables in es s and eg essions (2)
Va iable Desc ip ion
E[Sha eB] Ca ego ical a iable o he agen ’s expec ed payo sha e (in pe cen ) o o al i m
p o i in Tale (expe imen al cu ency) i s/he chooses P ojec B and s a e 2 ealises
(30%; 40%; 45%; 60%; 80%)
P eReli Measu e indica ing whe he subjec s pe cei e ha hey ha e gi en co ec in o ma-
ion in he p e-expe imen al ques ionnai e
Rec Ca ego ical a iable o ecommenda ion in a decision ound (0 = no ecommenda-
ion; A = P ojec A- ecommenda ion; B = P ojec B- ecommenda ion)
Reli Measu e indica ing whe he subjec s pe cei e ha hey ha e gi en co ec in o ma-
ion in he pos -expe imen al ques ionnai e
RecDum Dummy a iable = 1 i subjec ecei es a P ojec ecommenda ion (no dis inc ion
be ween he ype o ecommenda ion)
Re P inRiskA e Dummy a iable = 1 i he p incipal communica ed isk a e sion, i.e., p e e ences
o a oiding he isky p ojec , in he p e ious decision ound ia cha ( a iable is
used o exclude cases om s a is ics, es s, and eg essions)
Re P inRiskTol Dummy a iable = 1 i he p incipal communica ed isk ole ance, i.e., p e e ences
o an in es men in he isky p ojec , in he p e ious decision ound ia cha ( a i-
able is used o exclude cases om s a is ics, es s, and eg essions)
Round Decision se o he expe imen
Semes e Cu en leng h o s udy measu ed in numbe o semes e s/ e ms
S a e S a e o na u e (bad = 1; good = 2)
Subjec Ca ego ical a iable o ole o he subjec (1 = agen ; 2 = p incipal)
T ea men Condi ion he subjec is pa o (1 = BL; 2 = LOS; 3 = REC; 4 = LOS-REC; 5 =
DEC; 6 = DEC-REC; 7 = LOW; 8 = LOW-REC)
Va [Sha eB] Va iance o he subjec ’s payo sha e om P ojec B
WillRisk Measu e o willingness o ake isks om SOEP (see ques ionnai e isk a i ude)
771
Explo ing decision‑making: expe imen al obse a ions on…
See Appendix Table6.
Table 6 RE logi models on he
agen s’ P ojec B-choices in BL
Random e ec s logi models on he agen s’ P ojec B-choices
including da a om he baseline (BL), i.e., no jus i ica ion and no
ecommenda ions. The dependen a iable ChoiceB is a bina y
a iable signaling o each ound i he agen selec s P ojec B.
E[Sha eB] is a ca ego ical a iable indica ing he agen ’s expec ed
p o i sha e (in pe cen ) in s a e 2 when selec ing P ojec B. I s base
le el is “30%.” Va [Sha eB] indica es he a iance o he agen ’s
payo sha e (in absolu e numbe s) when selec ing P ojec B. Con-
ol a iables a e he subjec s’ age (Age), sex (Male), s udy p og ess
(Semes e ), inequali y a e sion (Inequali yF), and isk a i ude (Will-
Risk). Coe icien s a e p esen ed in exponen ia ed o m, i.e., as odds
a ios. S anda d e o s (in pa en heses) a e clus e ed a he indi idual
le el. The cons an s a e included in he models bu no epo ed
*
p
<
0.1
**
p
<
0.05
***
p
<
0.01
ChoiceB (1) (2)
E[Sha eB]
40% 1.969 1.971
(1.607) (1.608)
45% 1.773 1.773
(1.207) (1.206)
60% 7.810* 7.830*
(8.819) (8.863)
80% 29.968 30.128
(74.218) (74.792)
Va [Sha eB] 0.982 0.982
(0.016) (0.016)
WillRisk 1.200
(0.198)
Inequali yF
Medium 0.453
(0.507)
High 0.332
(0.316)
Age 0.929
(0.073)
Male 1.690
(1.109)
Semes e 1.303*
(0.147)
Obse a ions (n) 168 168
Wald chi2 6.89 14.69
P ob > chi2 0.2287 0.1973
772
C.Lukas e al.
Acknowledgemen s We acknowledge help ul discussions and commen s om Timo hy W. Shields a he
12 h Wo k- shop on Accoun ing and Economics in Tilbu g, 2016, Ca he ine Roux a he 17 h GEABA
Symposium in Basel, 2016, and Robe G asse a he 2019 ENEAR mee ing in Maas ich . We also
would like o hank Se geja Slapnica , Ba ba a Schöndube-Pi chegge , Ca ina Keldenich and Ch is ine
Lücke o commen s and ideas. We a e also g a e ul o Kay Blau us, Ma in Fochmann, and Nadja Foch-
mann o hei suppo in conduc ing he expe imen .
Funding Open Access unding enabled and o ganized by P ojek DEAL. The dean’s o ice o he Facul y
o Economics and Managemen o he Leibniz Uni e si ä Hanno e suppo ed he esea ch and pa ially
inanced he expe imen wi h 2.000 eu os. The e was no u he in ol emen o he acul y o o he hi d
pa ies.
Decla a ions
Con lic o in e es The au ho s ha e no ele an inancial o non- inancial in e es s o disclose.
E hical app o al App o al o conduc expe imen s wi h human pa icipan s in he Leibniz Labo a o y o
Expe imen al Economics (i s-Pool) has been g an ed by he dean’s o ice o he Facul y o Economics and
Managemen o he Leibniz Uni e si ä Hanno e .
Open Access This a icle is licensed unde a C ea i e Commons A ibu ion 4.0 In e na ional License,
which pe mi s use, sha ing, adap a ion, dis ibu ion and ep oduc ion in any medium o o ma , as long
as you gi e app op ia e c edi o he o iginal au ho (s) and he sou ce, p o ide a link o he C ea i e
Commons licence, and indica e i changes we e made. The images o o he hi d pa y ma e ial in his
a icle a e included in he a icle’s C ea i e Commons licence, unless indica ed o he wise in a c edi line
o he ma e ial. I ma e ial is no included in he a icle’s C ea i e Commons licence and you in ended
use is no pe mi ed by s a u o y egula ion o exceeds he pe mi ed use, you will need o ob ain pe mis-
sion di ec ly om he copy igh holde . To iew a copy o his licence, isi h p://c ea i ecommons.o g/
licenses/by/4.0/.
Re e ences
Adelbe g, S., & Ba son, C. D. (1978). Accoun abili y and helping: When needs exceed esou ces. Jou nal
o Pe sonali y and Social Psychology, 36(4), 343–350. h ps:// doi. o g/ 10. 1037/ 0022- 3514. 36.4. 343
Ag awal, A., & Mandelke , G. N. (1987). Manage ial incen i es and co po a e in es men and inancing
decisions. The Jou nal o Finance, 42(4), 823–837.
Ande sson, O., Holm, H. J., Ty an, J. R., & Wengs öm, E. (2016). Deciding o o he s educes loss a e -
sion. Managemen Science, 62(1), 29–36. h ps:// doi. o g/ 10. 1287/ mnsc. 2014. 2085
A nold, M. C. (2015). The e ec o supe io s’ exogenous cons ain s on budge nego ia ions. The
Accoun ing Re iew, 90(1), 31–57. h ps:// doi. o g/ 10. 2308/ acc - 50864
Ash on, R. H. (1990). P essu e and pe o mance in accoun ing decision se ings: Pa adoxical e ec s o
incen i es, eedback, and jus i ica ion. Jou nal o Accoun ing Resea ch, 28, 148–180. h ps:// doi.
o g/ 10. 2307/ 24912 53
Ash on, R. H. (1992). E ec s o jus i ica ion and a mechanical aid on judgmen pe o mance. O gani-
za ional Beha io and Human Decision P ocesses, 52(2), 292–306. h ps:// doi. o g/ 10. 1016/ 0749-
5978(92) 90040-E
Ba on, J., & He shey, J. C. (1988). Ou come bias in decision e alua ion. Jou nal o pe sonali y and social
psychology, 54(4), 569.
Bauch, K. A., & Weißenbe ge , B. E. (2020). The e ec s o accoun abili y on a o i ism in subjec i e
pe o mance e alua ions: An eye- acking s udy. SSRN. h ps:// doi. o g/ 10. 2139/ ss n. 36463 01
Ben-Ne , A., & Pu e man, L. (2009). T us , communica ion and con ac s: An expe imen . Jou nal o
Economic Beha io & O ganiza ion, 70(1–2), 106–121.
Bock, O., Bae ge, I., & Nicklisch, A. (2014). h oo : Hambu g egis a ion and o ganiza ion online ool.
Eu opean Economic Re iew, 71, 117–120. h ps:// doi. o g/ 10. 1016/j. eu oe co e . 2014. 07. 003
773
Explo ing decision‑making: expe imen al obse a ions on…
B and s, J., Coope , D. J., & Ro , C. (2019). Communica ion in labo a o y expe imen s. In Handbook o
esea ch me hods and applica ions in expe imen al economics. Edwa d Elga Publishing.
B ink, A. G., & Rankin, F. W. (2013). The e ec o isk p e e ence and loss a e sion on indi idual beha -
io unde bonus, penal y, and combined con ac ames. Beha io al Resea ch in Accoun ing, 25(2),
145–170.
Chak a a y, S., Ha ison, G. W., Ha u y, E. E., & Ru s öm, E. E. (2011). A e you isk a e se o e o he
people’s money? Sou he n Economic Jou nal, 77(4), 901–913.
Chang, W., A anaso , P., Pa il, S. V., Melle s, B. A., & Te lock, P. E. (2017). Accoun abili y and adap i e
pe o mance unde unce ain y: A long- e m iew. Judgmen and Decicion Making, 12(6), 610–626.
Chang, L. J., Cheng, M. M., & T o man, K. T. (2013). The e ec o ou come and p ocess accoun abili y
on cus ome -supplie nego ia ions. Accoun ing, O ganiza ions and Socie y, 38(2), 93–107. h ps://
doi. o g/ 10. 1016/j. aos. 2012. 12. 002
Cha ness, G., Gneezy, U., & Halladay, B. (2016). Expe imen al me hods: Pay one o pay all. Jou nal o
Economic Beha io & O ganiza ion, 131, 141–150. h ps:// doi. o g/ 10. 1016/j. jebo. 2016. 08. 010
Coles, J. L., Daniel, N. D., & Na een, L. (2006). Manage ial incen i es and isk- aking. Jou nal o Finan-
cial Economics, 79(2), 431–468. h ps:// doi. o g/ 10. 1016/j. j ine co. 2004. 09. 004
Dalla Via, N., Pe ego, P., & an Rinsum, M. (2019). How accoun abili y ype in luences in o ma ion
sea ch p ocesses and decision quali y. Accoun ing, O ganiza ions and Socie y, 75, 79–91. h ps://
doi. o g/ 10. 1016/j. aos. 2018. 10. 001
Feh enbache , D. D., Kaplan, S. E., & Moulang, C. (2020). The ole o accoun abili y in educing he
impac o a ec i e eac ions on capi al budge ing decisions. Managemen Accoun ing Resea ch, 47,
100650. h ps:// doi. o g/ 10. 1016/j. ma . 2019. 100650
Fischbache , U. (2007). z-T ee. Zu ich oolbox o eady-made economic expe imen s. Expe imen al Eco-
nomics, 10(2), 171–178. h ps:// doi. o g/ 10. 1007/ s10683- 006- 9159-4
Fo in, B., Lac oix, G., & Ville al, M. C. (2007). Tax e asion and social in e ac ions. Jou nal o Public
Economics, 91(11–12), 2089–2112. h ps:// doi. o g/ 10. 1016/j. jpube co. 2007. 03. 005
F imanson, L., Ho nbach, J., & Ha mann, F. G. (2021). Pe o mance e alua ions and s ess: Field e i-
dence o he ho monal e ec s o e alua ion equency. Accoun ing, O ganiza ions and Socie y, 95,
101279. h ps:// doi. o g/ 10. 1016/j. aos. 2021. 101279
Gomila, R. (2021). Logis ic o linea ? Es ima ing causal e ec s o expe imen al ea men s on bina y ou -
comes using eg ession analysis. Jou nal o Expe imen al Psychology: Gene al, 150(4), 700.
G abne , I., & Moe s, F. (2013). Managemen con ol as a sys em o a package? Concep ual and empi i-
cal issues. Accoun ing, O ganiza ions and Socie y, 38(6–7), 407–419.
Hall, A. T., F ink, D. D., & Buckley, M. R. (2017). An accoun abili y accoun : A e iew and syn hesis
o he heo e ical and empi ical esea ch on el accoun abili y. Jou nal o O ganiza ional Beha io ,
38(2), 204–224. h ps:// doi. o g/ 10. 1002/ job. 2052
Je mias, J. (2006). The in luence o accoun abili y on o e con idence and esis ance o change: A esea ch
amewo k and expe imen al e idence. Managemen Accoun ing Resea ch, 17(4), 370–388.
Kim, S., & T o man, K. T. (2015). The compa a i e e ec o p ocess and ou come accoun abili y in
enhancing p o essional scep icism. Accoun ing & Finance, 55(4), 1015–1040. h ps:// doi. o g/ 10.
1111/ ac i. 12084
Langhe, Bd., an Osselae , S. M., & Wie enga, B. (2011). The e ec s o p ocess and ou come accoun -
abili y on judgmen p ocess and pe o mance. O ganiza ional Beha io and Human Decision P o-
cesses, 115(2), 238–252. h ps:// doi. o g/ 10. 1016/j. obhdp. 2011. 02. 003
Le eb e, M., & Vieide , F. M. (2013). Reining in excessi e isk- aking by execu i es: The e ec o
accoun abili y. Theo y and Decision, 75(4), 497–517. h ps:// doi. o g/ 10. 1007/ s11238- 012- 9335-2
Leong, C. (1991). Accoun abili y and p ojec managemen : A con e gence o objec i es. In e na ional
Jou nal o P ojec Managemen , 9(4), 240–249. h ps:// doi. o g/ 10. 1016/ 0263- 7863(91) 90033-R
Le ne , J. S., & Te lock, P. E. (1999). Accoun ing o he e ec s o accoun abili y. Psychological Bulle in,
125(2), 255–275.
Lipe, M. G. (1993). Analyzing he a iance in es iga ion decision: The e ec o ou comes, men al
accoun ing, and aming. The Accoun ing Re iew, 68(4), 748–764.
Lukas, C., Neube , M. F., & Schöndube, J. R. (2019). Accoun abili y in an agency model: P ojec selec-
ion, e o incen i es, and con ac design. Manage ial and Decision Economics, 40(2), 150–158.
h ps:// doi. o g/ 10. 1002/ mde. 2989
Mac Donald, K., Rezania, D., & Bake , R. (2020). A g ounded heo y examina ion o p ojec manag-
e s’ accoun abili y. In e na ional Jou nal o P ojec Managemen , 38(1), 27–35. h ps:// doi. o g/ 10.
1016/j. ijp o man. 2019. 09. 008
774
C.Lukas e al.
Me chan , K. A., & O ley, D. T. (2007). A e iew o he li e a u e on con ol and accoun abili y. In C. S.
Chapman, A. G. Hopwood, & M. D. Shields (Eds.), Handbooks o managemen accoun ing esea ch
(Vol. 2, pp. 785–802). Else ie . h ps:// doi. o g/ 10. 1016/ S1751- 3243(06) 02013-X
Me chan , K. A., & Van de S ede, W. A. (2007). Managemen con ol sys ems: Pe o mance measu e-
men , e alua ion and incen i es. Pea son Educa ion.
Messne , M. (2009). The limi s o accoun abili y. Accoun ing, O ganiza ions and Socie y, 34(8), 918–938.
Mi , F. A., & Rezania, D. (2023). P ojec leade ’s in e ac i e use o con ols, eam lea ning beha iou and
IT p ojec pe o mance: The mode a ing ole o p ocess accoun abili y. Leade ship and O ganiza-
ion De elopmen Jou nal, 44(6), 742–770. h ps:// doi. o g/ 10. 1108/ LODJ- 12- 2022- 0553
Pahlke, J., S asse , S., & Vieide , F. M. (2012). Risk- aking o o he s unde accoun abili y. Economics
Le e s, 114(1), 102–105. h ps:// doi. o g/ 10. 1016/j. econl e . 2011. 09. 037
Pahlke, J., S asse , S., & Vieide , F. M. (2015). Responsibili y e ec s in decision making unde isk.
Jou nal o Risk and Unce ain y, 51(2), 125–146. h ps:// doi. o g/ 10. 1007/ s11166- 015- 9223-6
Pa il, S. V., Vieide , F., & Te lock, P. E. (2014). P ocess e sus ou come accoun abili y. In M. Bo ens, R.
E. Goodin, & T. Schillemans (Eds.), The Ox o d Handbook o Public Accoun abili y. Ox o d Uni-
e si y P ess. h ps:// doi. o g/ 10. 1093/ ox o dhb/ 97801 99641 253. 013. 0002
Pa il, S. V., Te lock, P. E., & Melle s, B. A. (2017). Accoun abili y sys ems and g oup no ms: Balancing
he isks o mindless con o mi y and eckless de ia ion. Jou nal o Beha io al Decision Making,
30(2), 282–303. h ps:// doi. o g/ 10. 1002/ bdm. 1933
Pollmann, M. M., Po e s, J., & T au mann, S. T. (2014). Risk aking by agen s: The ole o ex-an e and ex-
pos accoun abili y. Economics Le e s, 123(3), 387–390. h ps:// doi. o g/ 10. 1016/j. econl e . 2014. 04. 004
Polman, E. (2012). Sel -o he decision making and loss a e sion. O ganiza ional Beha io and Human
Decision P ocesses, 119(2), 141–150. h ps:// doi. o g/ 10. 1016/j. obhdp. 2012. 06. 005
Rezania, D., Bake , R., & Nixon, A. (2019). Explo ing p ojec manage s’ accoun abili y. In e na-
ional Jou nal o Managing P ojec s in Business, 12(4), 919–937. h ps:// doi. o g/ 10. 1108/
IJMPB- 03- 2018- 0037
Robe s, J. (2009). No one is pe ec : The limi s o anspa ency and an e hic o ‘in elligen ’ accoun abili y.
Accoun ing, O ganiza ions and Socie y, 34(8), 957–970. h ps:// doi. o g/ 10. 1016/j. aos. 2009. 04. 005
Sawe s, K., W igh , A., & Zamo a, V. (2011). Does g ea e isk-bea ing in s ock op ion compensa ion
educe he in luence o p oblem aming on manage ial isk- aking beha io ? Beha io al Resea ch
in Accoun ing, 23(1), 185–201. h ps:// doi. o g/ 10. 2308/ b ia. 2011. 23.1. 185
Siegel-Jacobs, K., & Ya es, J. (1996). E ec s o p ocedu al and ou come accoun abili y on judgmen
quali y. O ganiza ional Beha io and Human Decision P ocesses, 65(1), 1–17. h ps:// doi. o g/ 10.
1006/ obhd. 1996. 0001
Te lock, P. E. (1983). Accoun abili y and complexi y o hough . Jou nal o Pe sonali y and Social Psy-
chology, 45(1), 74–83. h ps:// doi. o g/ 10. 1037/ 0022- 3514. 45.1. 74
Te lock, P. E. (1985). Accoun abili y: A social check on he undamen al a ibu ion e o . Social Psy-
chology Qua e ly, 48(3), 227–236.
Te lock, P. E., Ski ka, L., & Boe ge , R. (1989). Social and cogni i e s a egies o coping wi h accoun a-
bili y: Con o mi y, complexi y, and bols e ing. Jou nal o Pe sonali y and Social Psychology, 57(4),
632–640. h ps:// doi. o g/ 10. 1037/ 0022- 3514. 57.4. 632
T e sky, A., & Kahneman, D. (1991). Loss a e sion in iskless choice: A e e ence-dependen model.
Qua e ly Jou nal o Economics, 106(4), 1039–1061. h ps:// doi. o g/ 10. 2307/ 29379 56
Vieide , F. M. (2009). The e ec o accoun abili y on loss a e sion. Ac a Psychologica, 132(1), 96–101.
h ps:// doi. o g/ 10. 1016/j. ac psy. 2009. 05. 006
Vieide , F. M. (2011). Sepa a ing eal incen i es and accoun abili y. Expe imen al Economics, 14(4),
507–518. h ps:// doi. o g/ 10. 1007/ s10683- 011- 9279-3
Webb, R. (2002). The impac o epu a ion and a iance in es iga ions on he c ea ion o budge slack.
Accoun ing, O ganiza ions and Socie y, 27(4–5), 361–378. h ps:// doi. o g/ 10. 1016/ S0361- 3682(01)
00034-4
Weimann, J., & B osig-Koch, J. (2019). Me hods in expe imen al economics. Sp inge . h ps:// doi. o g/ 10.
1007/ 978-3- 319- 93363-4
Woold idge, J. M. (2002). Econome ic analysis o c oss sec ion and panel da a. The MIT P ess.
Publishe ’s No e Sp inge Na u e emains neu al wi h ega d o ju isdic ional claims in published maps
and ins i u ional a ilia ions.
