Vol.:(0123456789)
RESEARCH
Towa ds imp o ing agili y inpublic adminis a ion
HannaLooks1 · JannikFangmann2 · Jö gThomaschewski2 ·
Ma ía-JoséEscalona1 · E a-Ma iaSchön2
Accep ed: 18 No embe 2023 / Published online: 20 Janua y 2024
© The Au ho (s) 2024
Abs ac
In imes o c ises, such as he e ugee c isis o he co ona pandemic, he wo kload in public
adminis a ions inc eases because o he demands o ci izens o sho - e m legal changes.
In addi ion, he e is an inc easing need o digi aliza ion o o be able o eac lexibly o
changes. Agile p ocess models and agile p ac ices a e app op ia e o o e come hese chal-
lenges. The objec i e o his pape is o in es iga e how public adminis a ions can meas-
u e hei deg ee o agili y o iden i y po en ial o imp o ing i . The au ho s conduc ed a
desc ip i e single-case s udy which included mul iple uni s o analysis in a public admin-
is a ion in Ge many. The case s udy was suppo ed by hei ques ionnai e o measu ing
he deg ee o agili y. One ou come o his s udy is a concep ual amewo k ha can be used
o d i e agile ans o ma ion in public adminis a ions by con inuously measu ing agili y.
The e o e, a ques ionnai e o measu ing agili y a eam le el in public adminis a ions has
been de eloped. The applica ion o he ques ionnai e in h ee eams p o ide insigh s in o
dys unc ionali y in he in e disciplina y eams as well as op imiza ion po en ial in e ms o
a ini y o change. The adop ion o agili y in public adminis a ion is a challenge, gi en ha
esis ance o change is s ill p e alen . A ans o ma ional change is a cons an jou ney, and
he e o e, he measu emen o p og ess plays an impo an ole in he con inuous imp o e-
men o an o ganiza ion. The applied app oach deli e s high po en ial o imp o emen in
e ms o agili y and p o ides in e es ing insigh s o bo h p ac i ione s and academics.
Keywo ds Agile· Agile ans o ma ion· Agile alues· e-Go e nmen · Measu emen o
agili y· Public adminis a ion· Team le el
1 In oduc ion
The wo kload in public adminis a ions inc eases because o he demands o ci izens, c isis
si ua ions, o sho - e m changes in legisla ion. In hese imes, in pa icula , public admin-
is a ions can be o e loaded and mus ha e he abili y o eac o changes lexibly. This
being said, public adminis a ions a e o en clea ly de ined in e ms o hei s uc u e and
*Hanna Looks
[email p o ec ed]
1 Uni e si y o Se ille, Se ille, Spain
2 Uni e si y o Applied Science Emden/Lee , Emden, Ge many
284
1 3
o ganiza ion and changes can only be implemen ed wi h a lo o bu eauc a ic e o . New
de elopmen s a e managed by means o es ablished mechanisms o go e nance, wi h s a-
bili y and accoun abili y as he main alues (Janssen & an de Voo , 2016). Mo eo e ,
decisions a e made on he basis o ixed c i e ia. The con ac ing o p ojec s o ex e nal
se ice p o ide s is usually egula ed by objec i e p ocu emen p ocedu es. Fu he mo e,
ex ensi e documen a ion is o en used in ela ion o p ojec managemen . Requi emen s
end o al eady be comple ely de ined a he beginning o a p ojec and he scope is ixed
(Fangmann e al., 2020). These condi ions in public adminis a ions lead o challenges such
as s ong hie a chies, insu icien communica ion, and e en anxie y wi h ega d o change.
The use o agile p ocess models can be a possible app oach o su moun he challenges
in espec o he need o lexibili y as well as he special obs acles in public adminis-
a ions desc ibed abo e. The p ac ice o agile alues p omo es a cul u e o con inuous
communica ion and eedback. In addi ion, he adop ion o agile p ac ices p omises capa-
bili ies like enhancing he abili y o manage changing p io i ies o inc easing p oduc i i y
(Digi al.ai, 2020). Mo eo e , i has been shown in he pas ha agile p ocess models can
be success ully applied in he ield o public adminis a ion. On he one hand, To ecilla
e al. (2013) p esen lessons lea ned om applying an agile amewo k based on Sc um
(Schwabe & Su he land, 2020) o so wa e de elopmen in public adminis a ion in Spain.
On he o he hand, Ka aj and Li le (2013) show how he adop ion o Lean and Kanban
(Ande son, 2010) changes he way o wo king in a public adminis a ion in Canada.
Agile p ocess models ha e hei o igins in so wa e de elopmen . The deli e ies o p o-
jec s in public adminis a ions do no always include so wa e, o so wa e is only a pa
he eo a lo o he ime. The e o e, sou ce code is a ely c ea ed in public adminis a ions,
bu s anda d so wa e is usually adap ed. E en in his con ex , so wa e quali y plays an
impo an ole. To main ain a high-quali y so wa e p oduc , a ious quali y ac o s mus
be me a a ce ain le el (Jain e al., 2018). Quali y echniques o adi ional and hea y-
weigh so wa e p oduc de elopmen a e based on inspec ions and e iews a he end o a
de elopmen p ocess, whe eas quali y assu ance echniques o agile, ligh weigh de elop-
men a e based on ou ine ac i i ies o he eam (Jain e al., 2016). In agile eams ha wo k
wi h sho i e a ions and sho imelines, i is especially impo an o moni o and imp o e
he quali y o he so wa e. Th ough consis en cus ome eedback, an inc ease in so wa e
quali y can be achie ed (Mashmool e al., 2021). In addi ion, agile eams mus de elop an
unde s anding o how di e en quali y aspec s (e.g., use - iendliness, eliabili y, pe o -
mance, and compa ibili y) can be inco po a ed in o hei so wa e de elopmen .
Public adminis a ions o en imes ace complex p oblems like e-go e nmen o digi ali-
za ion p ojec s. So as o be able o sol e complex p oblems, people need o p obe i s , hen
sense and espond, in o de o espond o he complexi y (Snowden & Boone, 2007). In
his con ex , agile p ocess models like Sc um (Schwabe & Su he land, 2020) o Kanban
(Ande son, 2010) can play an impo an ole because hey p o ide empi ical app oaches,
which allow o ganiza ions o op imize p edic abili y and con ol isk caused by i e a i e
and inc emen al app oaches.
The aim o his pape is o p esen he indings o a case s udy app oach on how public
adminis a ions can measu e hei deg ee o agili y o iden i y po en ial o imp o ing i . To
his end, he ollowing esea ch ques ion (RQ) is posed: How can agili y in public adminis-
a ions be measu ed?
Fo his pu pose, he au ho s ha e de eloped scales o measu e agili y in non-so wa e
en i onmen s. In pa icula , he au ho s ha e adap ed hei p e iously de eloped ques ion-
nai e o measu ing agili y (Looks e al., 2021) o he con ex o public adminis a ion, o
he gene al condi ions o be aken in o accoun .
285
The pape is o ganized as ollows: Sec .2 p esen s he backg ound and ela ed wo ks. In
Sec .3, he au ho s a e going o p esen hei esea ch objec i e and he esea ch me hod
which guided his s udy. Sec ion4 demons a es he esul s and he indings conce ning
he measu emen o agili y in public adminis a ions a e summa ized. Sec ion5 is going
o discuss he meaning o indings and implica ions o his s udy as well as documen he
limi a ions o he esea ch. Sec ion6 o e s conclusions and p esen s he u u e esea ch in
his ield o esea ch.
2 Backg ound and ela ed wo k
The use o agile p ocess models is an empi ical app oach and con inuous imp o emen
is one o i s main agile alues. To his end, he agile ans o ma ion o an o ganiza ion
is an ongoing jou ney wi hou a de ined end da e. In o de o ensu e ha con inuous
imp o emen can also be e ec i ely e alua ed, his p ocess should be accompanied by
egula measu emen s o agili y. In he ollowing, he au ho s a e going o p esen he
backg ound o imp o ing agili y by means o measu emen , as well as explain he ela ed
wo k ound on his opic. In addi ion, he au ho s a e going o sho ly ou line hei p e i-
ous wo ks in his esea ch ield, as hese a e used as he heo e ical ounda ion o con-
duc ing he case s udy.
2.1 Con ex o public adminis a ion
Ci izens’ expec a ions wi h ega d o public adminis a ion a e changing, and echni-
cal de elopmen s in he con ex o digi iza ion a e leading o g owing demands on pub-
lic adminis a ion. This esul s in high expec a ions o public adminis a ion in e ms o
ci izen communica ion, pa icipa ion, and anspa ency. The inc easing demands and he
esul ing opics and p ojec s usually do no i in o he p e ious s uc u e o public admin-
is a ion because hey a e new in he way hey a ise o in e ms o con en and mo e ac oss
he c oss-sec ion o di e en a eas. These opics canno be embedded in he ac ual line
s uc u e o public adminis a ion, as hey ha e o be deal wi h ac oss depa men s and
wi hin se e al esponsibili ies and hie a chies (Ba oni z e al., 2018). As a esul , he wo k
o employees in public adminis a ion is no longe exclusi ely limi ed o he adminis a i e
asks assigned o hem, bu p ojec and eamwo k is inc easingly coming in o ocus. Me h-
odological knowledge o p ojec wo k is o en no a ailable and a e in luenced by co e-
sponding in luencing ac o s. Public adminis a ions a e o en bu eauc a ically s uc u ed
and o ganized. Due o he ixed alloca ion o esponsibili ies, he o icial hie a chy, admin-
is a i e egula ions, and documen a ion as he main ea u es o adminis a i e o gani-
za ion, hey allow li le lexibili y and can lead o us a ion and demo i a ion among
employees (Bogumil, 2008). Agili y can be a c ucial imp o emen ac o o hese men-
ioned challenges in eam-o ien ed p ojec wo k in public adminis a ions. The in luencing
ac o s clea ly show he need o a ge ed suppo in he agile ans o ma ion p ocess and
demons a e ha ins umen s o igina ing om pu e so wa e de elopmen canno be ans-
e ed o his con ex wi hou app op ia e adap a ion. Public adminis a ion is he e o e a
sui able con ex o an ini ial case s udy, because a pa icula impo ance mus be a ached
o a ge ed suppo in he agile ans o ma ion p ocess ha akes he in luencing ac o s
in o accoun .
286
2.2 Imp o ing agili y bymeans o measu emen
Some s udies epo esul s conce ning he measu emen o success o agile ans o ma-
ion o o measu ing agili y. The 14 h Annual S a e o Agile Repo (Digi al.ai, 2020)
shows how success o agile ans o ma ions is measu ed in di e en ypes o o ganiza-
ions. The op measu es a e anked as ollows: cus ome o use sa is ac ion, business
alue, on- ime deli e y, quali y, and business objec i e achie ed.
The li e a u e also men ions some ma u i y models, which e alua e he agili y o an o gan-
iza ion and iden i y app oaches o imp o ing i . In 2007, Sidky e al. (2007) published he
Agile Adop ion F amewo k, which is based on he p inciples o he Agile Mani es o (Beck
e al., 2001). The amewo k includes he measu emen o agili y wi h he Sidky Agile Meas-
u emen Index (SAMI), and a p ocess model o in oducing agile p ac ices wi hin an o gani-
za ion. In 2007, Packlick (2007) de ines a ma u i y model—by way o he Agile Ma u i y
Map—consis ing o he componen s’ a ge a eas, goals, accep ance c i e ia, and ma u i y
le els. Goals and hei accep ance c i e ia a e p esen ed in he o m o use s o ies. In 2009,
Pa el and Ramachand an (2009) de eloped he agile ma u i y model. The model de ines i e
ma u i y le els as well as an adminis a ion p ocess based on sel -assessmen h ough assess-
men ques ions. The ma u i y le els a e used o de ine goals o he so wa e de elopmen
p ocess. In o de o achie e he goals, he model shows a eas o imp o emen a each ma u-
i y le el. In 2010, Bene ield (2010) c ea ed a model ha dis inguishes agile ma u i y o
se en dimensions: au oma ed eg ession es ing, code quali y me ics, au oma ed deploymen
and backup, au oma ed builds and con igu a ion managemen bes p ac ices, in e locked
deli e y and in e ace in eg a ion es ing, es -d i en de elopmen (TDD), and pe o mance
and scalabili y es ing. Each o hese dimensions is assessed on he basis o i e le els o
ma u i y. In 2011, Yin e al. (2011) de eloped he sc um ma u i y model, which is based on
i e ma u i y le els. Fo each le el o ma u i y, conc e e objec i es and p ac ices a e de ined,
which ha e o be implemen ed by he o ganiza ion in o de o achie e he a ge . Fon ana
e al. (2015) coined he P og essi e Ou comes F amewo k in 2015. The amewo k dis in-
guishes six p og essi e ou comes: p ac ices, eam, deli e ies, equi emen s, p oduc , and cus-
ome . Fo each o hese ou comes, ma u i y le els a e de ined, a ying in numbe be ween
h ee and ou . In 2007, Qume e al. (2007) p o ided a ma u i y model called Agile Adop-
ion and Imp o emen Model (AAIM). This is an ex ension o he Agile So wa e Solu ion
F amewo k (ASSF) (Qume & Hende son-Selle s, 2009). The ASSF desc ibes—on a high
le el o abs ac ion—an app oach o he c ea ion and adap a ion o an agile so wa e de el-
opmen p ocess.
In addi ion o he a o emen ioned ma u i y models, he e a e also ques ionnai es ha
measu e agili y. In 2009, So and Scholl (2009) published he Pe cep i e Agile Measu e-
men ques ionnai e. The ques ionnai e measu es he pe cei ed agili y independen o
he conc e e, echnical ools ha a e used in so wa e de elopmen . The i ems o he
ques ionnai e measu e eigh ac o s: i e a ion planning, i e a i e de elopmen , con inu-
ous in eg a ion and es ing, s and-up mee ings, cus ome access, cus ome accep ance
es s, e ospec i es, and co-loca ion. In 2015, G en e al. (2015) de eloped a ques ion-
nai e ha can be used o ex end he Agile Adop ion F amewo k (Sidky e al., 2007). The
ques ionnai e eplaces he p ocedu e o analyzing o ganiza ional eadiness, acco ding
o Sidky e al. (2007). In con as o he o ganiza ional eadiness, he ques ionnai e is
no used o de e mine he agile po en ial, bu o in es iga e he cu en deg ee o agili y.
The a o emen ioned models o imp o ing and measu ing agili y ha e he disad-
an age ha hey do no ake in o accoun he con ex o he pa icula o ganiza ion.
287
The e o e, hese wo ks a e no specialized in he con ex o public adminis a ions
desc ibed in he in oduc o y sec ion. Especially in agile p ocess models, con ex plays
an impo an ole. Fo ins ance, he e a e s udies ha show ha he cul u e o an o gani-
za ion has an impac on he use o agile p ac ices (Gup a e al., 2019). Mo eo e , some
o he models e e ed o ha e been de eloped o so wa e de elopmen p ojec s. In he
con ex o public adminis a ion, howe e , so wa e is no de eloped in e e y p ojec .
The e o e, some o he models canno be applied o imp o e agili y in public adminis-
a ion wi hou adap a ion.
2.3 Measu ing agili y in hecon ex o public adminis a ion
Resea ch ega ding he measu emen o agili y in he con ex o public adminis a ion is a he
limi ed. In 2018, Me gel e al. (2018) conduc ed a sys ema ic li e a u e e iew ega ding agile
go e nmen . They iden i ied ou applica ion a eas o agili y in go e nmen (agile so wa e
de elopmen , agile p ojec managemen , agile acquisi ion, and agile e alua ion). Me gel e al.
(2018) s a e ha he ca ego y o agile e alua ion is la gely unde explo ed in a go e nmen al
con ex . They ound wo pape s which deal wi h he opic o e alua ion. One o hese, he
wo k by Dahma deh and Pou shahabi (2011) is ela ed o his wo k. They p o ide a knowl-
edge-based amewo k o he measu emen and assessmen o public sec o agili y and use
Kea ney’s Model o Agile Go e nmen as a basis o c ea ing hei scales. In con as o ou
wo k, Dahma deh and Pou shahabi (2011) iew he opic o agili y om a di e en pe spec-
i e (agile manu ac u ing) and use a di e en basis o hei scales.
2.4 Analysis o he esea ch gap
The analysis o ela ed wo k e eals ha he e is a need o suppo in agile ans o ma-
ion. So a , he au ho s ha e no been able o ind any wo k ha deals wi h he applica ion
o models o measu ing o imp o ing agili y in he con ex o public adminis a ion ha
is based on agile alues. Conside ing he ela ed wo ks in mo e de ail, only he wo k o
Dahma deh and Pou shahabi (2011) om 2011 deals wi h a measu emen o agili y in he
con ex o public adminis a ion, bu his model does no conside agile alues. Pa icu-
la ly due o he special ea u es o public adminis a ions men ioned in Sec .2.1, he main
ocus should be placed on alue o ien a ion in he con ex o a ge ed agile ans o ma ion.
Fu he mo e, o an applica ion o he ques ionnai e o measu ing agili y in public admin-
is a ion, a linguis ic adap a ion o his con ex mus be made be o e. Because his con ex
is di e en o so wa e de elopmen , which was aken in o accoun while de eloping he
ini ial ques ionnai e. Mo eo e , due o he di e si y o subjec a eas and asks in public
adminis a ions, eams, and hei p ojec s a e also o a e y di e en na u e. Fo his ea-
son, i mus be possible o adap he ool o he use -speci ic con ex o an indi idual eam
wi hin an adminis a ion. Known models ei he do no ake agile alues in o accoun o
hey ha e no been adap ed o he con ex o public adminis a ion.
3 Resea ch objec i e and esea ch me hod
The objec i e o his pape is o in es iga e how public adminis a ions can measu e hei
deg ee o agili y o iden i y po en ial o imp o ing i . To his end, we pose he ollowing
esea ch ques ion (RQ):
288
3.1 How can agili y inpublic adminis a ions be measu ed?
To answe his esea ch ques ion, an app op ia e esea ch me hodology mus be selec ed.
In his ega d, he au ho s chose o pe o m a case s udy. The app oach o use a case s udy
allows us o collec da a in p ac ice and o be e unde s and he con ex . Complex phenom-
ena wi h hei espec i e con ex s a e in es iga ed in case s udies (Bax e & Jack, 2008;
Go in e al., 2019; Yin, 2003). Acco ding o Bax e and Jack (2008), he esea ch ques ion
ela es o how-ques ions and he au ho s wan o co e con ex ual condi ions because he
bounda ies a e no clea be ween he phenomenon and i s con ex . Fu he mo e, case s ud-
ies a e no only a alid ool o gene a ing a heo y o answe he esea ch ques ion. They
can also be used o es and elabo a e he cons uc ed heo y (Go in e al., 2019). This
wo k is he e o e going o p esen he indings o a desc ip i e single-case s udy embedded
mul iple uni s o analysis (Yin, 2003), in which he agile ans o ma ion is suppo ed by he
ques ionnai e o measu ing agili y (Looks e al., 2021). The s udy comp ises h ee main
pa s: s udy planning, da a collec ion, and da a analysis. The ac i i ies o hese pa s a e
summa ized in Fig.1, oge he wi h he p elimina y wo k on his issue.
3.2 S udy con ex
In he ollowing, he au ho s a e going o explain he gene al con ex o hei case s udy and
hen desc ibe he di e ences be ween he embedded uni s (Team A, Team B, and Team C).
The in e ac ion be ween he en i ies is shown in Fig.2.
In Ge many, he adminis a ions o he egional au ho i ies o he ede al go e nmen ,
he ede al s a es and he local go e nmen s, which consis o o ganiza ional uni s o
di e en ypes, a e e e ed o as public adminis a ions (Reichel, 2010). Public admin-
is a ions in Ge many can be di ided in o planning, law-p epa ing and law-en o cing
o ganiza ional uni s (“au ho i ies”), as well as o ganiza ional uni s wi h se ice and p o-
duc ion cha ac e (“es ablishmen s”). As an adminis a i e pa o he execu i e b anch,
public adminis a ion se es o implemen and conc e ize poli ical decisions made by
he legisla u e and is hus accoun able o he go e nmen . Consequen ly, he adminis a-
ion is a cen al componen o he au ho i y-sha ing o ganiza ion o he mode n con-
s i u ional s a e (Reichel, 2010). The case s udy was conduc ed wi h h ee eams o a
Fig. 1 O e iew o he esea ch me hod
289
local go e nmen . As pa o he execu i e au ho i y, he local go e nmen consis s o
he municipal council and he adminis a i e s a , who a e led by democ a ically legi i-
mized mayo s. The di e ences be ween he h ee local go e nmen eams ha con ib-
u ed o he case s udy a e explained below.
•Embedded Uni Team A. Team A has i e eam membe s. I is a cen al uni ha is
esponsible o a digi aliza ion p ojec . Fu he mo e, Team A is supposed o mo i-
a e, suppo , and es ablish he use o agile wo king p ac ices in he cou se o he
adminis a i e mode niza ion. The eam al eady wo ks wi h agile p ac ices, such
as Kanban and egula s and-up mee ings. Team A assesses i sel in ad ance as
al eady agile.
•Embedded Uni Team B. Team B has i e eam membe s. The eam belongs o he
In o ma ion echnology depa men and is echnically in ol ed in he de elopmen
o digi al p oduc s. Rega ding he implemen a ion o digi iza ion p ojec s, he e is no
comple ely independen de elopmen o so wa e bu a he collabo a ion wi h ex e nal
se ice p o ide s. In e ms o o ganiza ional s uc u e and cul u e, i has a a he classic
se up wi h indi idual agile p ac ices in use. Depending on he p ojec , he eam decides
whe he i wan s o use agile o classic me hods o implemen a ion.
•Embedded Uni Team C. Team C has eigh eam membe s. This eam is esponsible o
de ining o ganiza ional aspec s such as he adap a ion o wo k p ocesses in he admin-
is a ion in acco dance wi h he law. Team C uses agile me hods and agile p ac ices
wi hou de eloping so wa e. Thus, Team C shows ha agile ways o wo king can be
used in public adminis a ion beyond so wa e de elopmen . I is a c oss- unc ional p o-
jec eam ha wo ks au onomously wi hou a hie a chy. The eam uses a sc um-like
amewo k wi h he oles p oduc owne , and sc um mas e . Sc um e en s a e ca ied
ou wi h ime adjus men s. In addi ion, a sc um boa d and icke s a e used. The sp in s
a e no always homogeneous, bu al e na e be ween s anda d and special sp in s.
Fig. 2 Gene al se ings o he case s udy
290
3.3 S udy planning
In he planning phase o his esea ch, i was al eady clea , based on a li e a u e e iew and
he p ac ical expe ience o he au ho s, ha he agile ans o ma ion in public adminis a-
ion had no ye p og essed e y a . The e o e, he au ho s could no selec jus any public
adminis a ion o he case s udy. The au ho s de ined he selec ion c i e ia as ollows:
•people al eady wo king oge he in eams;
•ini ial expe ience wi h he applica ion o agile me hods and agile p ac ices is a ailable;
• he public adminis a ion wan s o d i e agile ans o ma ion o wa d;
•decision make s suppo he implemen a ion o he case s udy.
The selec ion o a sui able public adminis a ion was made on a olun a y basis. Fo
his pu pose, a call o pa icipa ion was launched ia he newsle e o an o ganiza ion o
he es ablishmen o agile me hods in Ge man public adminis a ions. This o ganiza ion
consis s o an associa ion o p ac i ione s om public adminis a ion (municipal, can onal,
and ede al adminis a ion) as well as om adminis a ion-o ien ed se ice companies. The
au ho s ecei ed some esponses o hei eques o sea ch; howe e , only one adminis a-
ion ul illed he c i e ia abo e. In his ega d, he au ho s also con ac ed he public admin-
is a ion, which we e examined in he case s udy. Ini ial discussions wi h employees in he
s udied public adminis a ion showed ha i s expe ience wi h agile p ocess models and
p ac ices was al eady a ailable, and mo eo e , ha a s ep owa ds an agile ans o ma ion
had al eady been aken.
3.4 Da a collec ion
The case s udy was conduc ed du ing he pe iod om 2020/04/27 o 2020/07/16. Table1
shows an o e iew o he di e en me hods used o collec da a. All s udies we e con-
duc ed emo ely, so no ace- o- ace con ac wi h s udy pa icipan s was necessa y.
Agile and Lean pa adigms ocus on con inuous imp o emen . In he p e ious wo k, he
au ho s ha e in es iga ed how o ganiza ions could imp o e hei alue deli e y by means
o applying agile p ac ices o p oduc de elopmen (Schön e al., 2015, 2017a, b, 2020).
In his con ex , measu emen plays an impo an ole. Fo his eason, he au ho s ha e
de eloped a ques ionnai e o measu ing agili y o suppo agile ans o ma ion. The ques-
ionnai e o measu ing agili y measu es he cu en s a e o agili y in so wa e de elop-
men (Looks e al., 2021). The scales a e de eloped using he agile alues as a basis. The
ques ionnai e o measu ing agili y was o me ly de eloped o suppo agile ans o ma ion
in he con ex o so wa e de elopmen as well as digi al p oduc de elopmen . Applying i
o public adminis a ions equi es p io al e a ions, because so wa e is no de eloped only
in his pa icula con ex and p ojec s o en do no deli e a digi al p oduc as a esul . In
a i s s ep he au ho s adap ed a ques ionnai e o measu ing agili y (Looks e al., 2021)
o he con ex o public adminis a ion by means o empi ical expe alida ions in se -
e al i e a ions. This ac i i y was conduc ed in he pe iod om 2020/04/27 o 2020/05/24
(see Table1). In he i s ound, 20 expe s om di e en public adminis a ions e alua ed
he ques ionnai e’s i ems in e ms o hei sui abili y. The expe s judged ha 17 o he 28
i ems had o be adap ed because he e ms used we e speci ically connec ed o so wa e
de elopmen and hus unsui able o public adminis a ions. The i ems conside ed o be
291
Table 1 O e iew o da a collec ion
Ac i i y o da a collec ion Da a sou ce Type o da a Timeline
Adap ion o he ques ionnai e o measu ing agili y o he con ex o
public adminis a ion by expe alida ion Da a shee in Mic oso Excel Quali a i e and quan i a i e da a 2020/04/27–2020/05/24
2 wo kshops in oducing s udy and ques ionnai e o measu ing agili y Documen ed in w i en o m Quali a i e da a, ag eed p eselec ion o
weigh ing ques ions (see Table3)2020/05/27
Applying o he ques ionnai e o measu ing agili y in 3 eams Online ques ionnai e Quan i a i e da a 2020/06/29–2020/07/16
298
Table 6 S a is ical da a o Team A—all s a is ical esul s ela e o he “answe ed da a poin s”
Dimension Communica i e Change-a ine Imp o emen -o ien ed Sel -o ganized P oduc -d i en I e a i e
Team membe s (N o al) 5 5 5 5 5 5
Ques ions pe dimension 8 4 4 4 4 4
Answe ed da a poin s 40 20 20 19 18 20
Mean 1.28 0.65 1.85 0.82 0.58 0.95
Weigh ing 3 3 2 2 0 -1
S anda d de ia ion 1.12 1.32 1.21 1.02 1.36 1.23
Va iance 1.26 1.74 1.47 1.04 1.86 1.51
Con idence 0.98 1.16 1.06 0.90 1.20 1.08
299
4.3 Findings ela ed oTeam C
Figu e5 shows a compa ison o he calcula ed mean alues ( a ing ob ained h ough he
ques ionnai e) o he dimensions o Team C o he de ined weigh ing alues (ob ained
om he wo kshops be o e). The mean alue and all s a is ical esul s de e mined o a
dimension is calcula ed o e all pa icipan esponses and hus o e he answe ed da a
poin s o a eam in he dimension. The e o ba s ep esen he calcula ed 95% con i-
dence in e al.
Mo eo e , Table8 p esen s he s a is ical da a o Team C. The e alua ion o he ques-
ionnai e esul s o Team C shows ha he imp o emen -o ien ed dimension was a ed he
highes o all dimensions, wi h a mean o 2.31, and exceeds he p e iously de ined weigh -
ing alue “impo an ” (nume ical alue 2). The i e a i e and he p oduc -d i en dimensions
show a highe mean compa ed o he es ablished weigh ing alue. The mean alues o he
communica i e and he change-a ine dimensions a e signi ican ly lowe compa ed o he
co esponding weigh ing alues.
The de ailed analysis o he da a indica es an ou lie in he sel -o ganized dimension.
The da a om he pa icipan ’s assessmen o he i ems de ia es signi ican ly om he
assessmen s o he o he pa icipan s.
Fig. 4 Ra ing compa ed o weigh ing o Team B
300
Table 7 S a is ical da a o Team B. All s a is ical esul s ela e o he “answe ed da a poin s”
Dimension Communica i e Change-a ine Imp o emen -
o ien ed Sel -o ganized P oduc -d i en I e a i e
Team membe s (N o al) 5 5 5 5 5 5
Ques ions pe dimension 8 4 4 4 4 4
Answe ed da a poin s 39 20 16 15 14 18
Mean 0.74 0.30 0.94 0.63 0.92 -0.45
Weigh ing 3 3 2 2 0 -1
S anda d de ia ion 1.04 0.91 1.04 0.67 1.04 0.4
Va iance 1.09 0.84 1.07 0.45 1.09 0.16
Con idence 0.91 0.80 1.02 0.66 1.02 0.35
301
4.4 C oss‑case analysis
The ollowing is a c oss-case analysis o he ques ionnai e esul s o he h ee eams
unde in es iga ion. Figu e6 shows he mean alues de e mined o he h ee eams in
he espec i e dimensions and he associa ed 95% con idence in e als. Figu e6 illus-
a es ha , o he communica i e dimension, Team A was able o ecei e he high-
es mean alue in e ms o agili y. On he o he hand, in he change-a ine dimension,
Team C showed he g ea es p og ess in agile ans o ma ion. Mo eo e , in he i e a-
i e dimension, Team B achie ed he lowes mean alue. O e all, his dimension was
weigh ed wi h he scale alue “ a he unimpo an ” (nume ical alue − 1). All h ee
eams exceeded his p e iously de ined weigh ing. The sel -o ganized dimension
achie ed a mean alue > 0 in all h ee eams. Fu he mo e, Team C showed good p o-
g ess in he agile ans o ma ion o his dimension. In he p oduc -d i en dimension,
Team C achie ed he highes mean alue compa ed o he o he eams. In he imp o e-
men -o ien ed dimension, all h ee eams showed good p og ess in agile ans o ma ion.
Team C was able o exceed he p e iously de ined weigh ing alue o “impo an .”
5 Discussion
In his sec ion, he au ho s a e going o discuss he meaning o he indings and answe
he esea ch ques ion, o how agili y can be measu ed in public adminis a ions.
Fig. 5 Ra ing compa ed o weigh ing o Team C
302
Table 8 S a is ical da a o Team C. All s a is ical esul s ela e o he “answe ed da a poin s”
Dimension Communica i e Change-a ine Imp o emen -
o ien ed Sel -o ganized P oduc -d i en I e a i e
Team membe s (N o al) 8 8 8 8 8 8
Ques ions pe dimension 8 4 4 4 4 4
Answe ed da a poin s 64 31 32 32 27 31
Mean 1.03 1.18 2.31 1.97 1.08 1.10
Weigh ing 3 3 2 2 0 -1
S anda d de ia ion 0.54 0.82 0.39 0.9 0.65 0.61
Va iance 0.30 0.67 0.15 0.8 0.42 0.38
Con idence 0.38 0.57 0.27 0.62 0.45 0.43
303
Fi s , he au ho s a e going o c i ically e lec on he con ibu ions o his wo k o
show how hey can ex end he exis ing body o knowledge. The au ho s deli e ed a
ques ionnai e o measu e agili y in public adminis a ions a eam le el. Wi h an expe
alida ion, he ques ionnai e o measu ing agili y in so wa e de elopmen (Looks
e al., 2021) was adap ed o he con ex o public adminis a ion. Fo his pu pose, 17 o
he 28 i ems o he o iginal ques ionnai e had o be eph ased. This shows ha he pub-
lic adminis a ion con ex di e s om he so wa e de elopmen con ex . The adop ion
o agili y in public adminis a ion goes beyond so wa e de elopmen as desc ibed in
Sec s.2.1 and 2.3. Compa ed o he amewo k o Dahma deh and Pou shahabi (2011),
he ques ionnai e is based on agile alues. As a as is known, his is he i s ime ha
i has been possible o measu e he agili y o a eam in public adminis a ion. Wi h he
help o he ques ionnai e, eams a e now able o c i ically sha e hei unde s anding o
agili y based on me ics. This is a g ea p og ess, i he au ho s conside ha eamwo k
is no ye used as a s anda d in public adminis a ions (see Sec .2.1). The common
answe ing o he weigh ing ques ions wi hin a eam (see Table2), in pa icula , enables
he speci ic eam o ake a c i ical look a i s own way o wo king and means a s ep
owa ds a sha ed unde s anding o agili y wi hin he eam.
Subsequen ly, he au ho s a e going o discuss he analysis o he su ey conduc ed
o measu e agili y a eam le el. The esul s pe dimension (communica i e, change-
a ine, i e a i e, sel -o ganized, p oduc -d i en, and imp o emen -o ien ed) a e dis-
cussed. In some cases, he au ho s a e going o di e one le el deepe and discuss indi-
idual i ems o a dimension. I he esul s o he h ee eams a e analyzed in mo e
de ail, po en ial o op imiza ion can be iden i ied.
Fig. 6 C oss-case compa ison be ween eams. The e o ba s ep esen he calcula ed 95% con idence in e al
304
5.1 Meaning o indings
The e alua ion o Team A e eals a conside able need o ac ion in he change-a ine
dimension. The mean alue de e mined o his dimension is 0.65 and ep esen s
he la ges di e ence wi h ega d o he de ined weigh ing o he dimension wi h
he scale alue “pa icula ly impo an ” (nume ical alue 3) (see Sec .4.1). Simila
o his, Team B shows po en ial o op imiza ion in he change-a ine dimension (see
Sec .4.2). In he Team B, he highes de ia ion om he weigh ing alue is ep e-
sen ed by he change-a ine dimension as well, wi h a mean alue o 0.3. This dimen-
sion was gi en he scale alue “pa icula ly impo an ” (nume ical alue 3) and hus
indica es he s onges need o ac ion. Team C shows mean alues > 1 in all dimen-
sions (see Sec .4.3). This al eady indica es a solid le el o agili y. The weigh ing al-
ues a e almos eached in he dimensions. A need o ac ion can be de i ed conce ning
he communica i e and he change-a ine dimension, since he mean alue is below he
weigh ing alue and a high weigh ing wi h he scale alue “pa icula ly impo an ” was
p e iously de ined. In all h ee eams, he imp o emen -o ien ed dimension was a ed
highes . This al eady indica es good p og ess in he a ea o imp o emen o ien a ion.
Wi h ega d o he assessmen o he ques ionnai e esul s o he indi idual i ems in he
imp o emen -o ien ed dimension, i is ob ious ha e ospec i es a e al eady being
ca ied ou and eam membe s a e ac i ely pa icipa ing in he con inuous imp o e-
men o he p ojec .
Due o he high a iance in he dimensions, a pa icipan -cen e ed e alua ion was
conduc ed in he indi idual dimensions o he eams. The pa icipan -cen e ed e alu-
a ion shows whe he he e a e any downwa d o upwa d ou lie s wi hin he eam and
in he espec i e dimension. Ou lie s may indica e p oblems in he eam dynamics.
In Team A, one pa icipan can be clea ly iden i ied as a downwa d ou lie wi hin he
ques ionnai e. In he communica i e, i e a i e, sel -o ganized, p oduc -d i en and
imp o emen -o ien ed dimension, he esul alues o his pa icipan de ia e signi i-
can ly om he esul alues o he o he pa icipan s. When looking a he ques ions
ega ding he a i ude and expe ience o he pa icipan s (see Table2), his pa ici-
pan a es himsel as an ad anced beginne in e ms o his compe ence in agile digi al
p oduc de elopmen . The pa icipan a es his hinking and ac ing in his daily wo k
as “no agile.” All o he membe s o Team A, on he o he hand, assess hei wo k as
“ a he agile.” Fu he mo e, i is ob ious ha ano he eam membe in Team A a ed
all dimensions highes . This pa icipan a es himsel as “compe en wi h ega d o
agile de elopmen o digi al p oduc s.” Based on hese esul s, Team A shows bo h
upwa d and downwa d ou lie s. This sugges s a di e en pe cep ion wi hin he eam
wi h ega d o he agile way o wo king and may indica e dys unc ionali ies o his
eam, which should be u he in es iga ed. Fu he in es iga ion o he possible dys-
unc ionali ies can be done, o ins ance, wi h he help o mode a ed g oup discus-
sions. In hese g oup discussions, he eam has he oppo uni y o c i ically e lec on
i s own agili y by discussing he measu emen esul s.
The achie emen o a low con idence in e al can hus show whe he he e is a
sha ed unde s anding in he small eams wi h ega d o he s a e o he agile ans o -
ma ion p ocess. The measu emen o agili y in a eam does no only mean a pu e meas-
u emen o agili y in six dimensions bu also se es as a basis o u he discussions in
he eam and hus as a s a ing poin o he concep ual amewo k.
305
5.2 How can agili y inpublic adminis a ions be measu ed?
A concep ual amewo k was de eloped o answe his esea ch ques ion (Bax e &
Jack, 2008). Figu e7 shows ou concep ual amewo k o how agili y can be measu ed
in public adminis a ions.
The le side o Fig.7 displays he ini ial si ua ion. A desc ip ion o he con ex in he
cu en e sion can be ound in Sec .3.2. In his case s udy, h ee eams we e in es iga ed
as embedded uni s. The au ho s used a ques ionnai e o measu e he deg ee o agili y o
hese h ee eams. E en he use o he ques ionnai e leads he pa icipa ing eams o an
in ensi e con on a ion and e lec ion o hei own wo king me hods wi h ega d o agili y.
The e alua ion o he ques ionnai e is also associa ed wi h he de elopmen o measu es
o imp o e agili y. Fo his pu pose, a con as is made be ween he objec i e (see Table5)
and he measu ed esul (see Sec s.4.1, 4.2, and 4.3). The de ia ions o hese wo alues
gi e an indica ion o opics ha should be in oduced and discussed and equi e u he
analysis. Addi ionally, measu es o imp o e agili y a e p oposed and applied in he con ex
unde s udy. The esul is he op imized con ex (see igh side o Fig.7) which uses hyb id
p ocess models o p oduc and se ice de elopmen . Hyb id p ocess models combine di -
e en me hods and p ac ices and a e made up o na u al p ocess e olu ion, which is mainly
d i en by expe ience, lea ning, and p agma ism (Kuh mann e al., 2017).
The measu emen o agili y showed ha all h ee embedded uni s ha e op imiza ion
po en ial in he change-a ine dimension. The e o e, measu es a e going o be de eloped
collabo a i ely wi h he indi idual eams in u he wo kshops o imp o e his impo an
aspec o he agile way o wo king. The p ocess o implemen ing he measu es mus be
accompanied by an expe (e.g. an agile coach), as i is p o en ha hie a chical cul u e has
a nega i e impac on he usage o social agile p ac ices (e.g. acili a e social in e ac ion,
collabo a ion, and di ec communica ion) as well as a nega i e impac on he usage o ech-
nical agile p ac ices (coding/ es ing-o ien ed so wa e enginee ing p ac ices) (Gup a e al.,
2019). Fu he mo e, a e applying he measu es o imp o e he change-a ine dimension, a
Fig. 7 Concep ual F amewo k o imp o ing agili y in he con ex o public adminis a ion
306
new measu emen o he agili y o each embedded uni is equi ed o e alua e whe he he
measu es show success. Mo eo e , he ques ionnai e e ealed ha he e may be dys unc-
ionali ies in Team A. Acco ding o Lencioni (2012), i is impo an o iden i y and esol e
dys unc ionali ies in eams in o de o p e en eam ailu e. The ailu e o a eam wi hin a
public adminis a ion does no suppo he imp o emen o agili y, gi en ha he ailu e
could be used as a nega i e example agains he adop ion o agile p inciples. Thus, i is
necessa y o c ea e a ecommended cou se o ac ion o dealing wi h he indings o he
applica ion o he ques ionnai e o measu ing agili y. All eams should ha e he oppo u-
ni y o de elop con inuously. A p e equisi e o his is a cul u al change in which he eam
membe s and hei in e ac ions a e p io i ized, wi h he aim o con inuously sha ing and
lea ning om each o he . The au ho s ecommend consul ing an agile coach o he imple-
men a ion o he a o emen ioned measu es o imp o e he public adminis a ion in e ms o
agili y. Agile coaches a e skilled in leade ship quali ies, p ojec managemen skills, as well
as echnical skills and ha e an expe ise in agile me hods. Agile coaches play a signi ican
ole in add essing challenges in an agile ans o ma ion, such as esis ance o change (S ay
e al., 2020). They a e esponsible o emo ing ba ie s o eam au onomy in agile eams
and make agile mee ings mo e aluable (S ay e al., 2020).
Public adminis a ions a e gene ally no classi ied as democ a ic ype o ganiza ions.
Democ a ic ype o ganiza ions a e mo e sui able o he use o agile p ocess models (Siakas
& Siakas, 2007). The e o e, he adop ion o agili y in public adminis a ion emains a chal-
lenge. The au ho s a e awa e o he ac ha ans o ma ional change is a cons an and e ol -
ing jou ney. Fo his eason, he measu emen o p og ess and he achie emen o a sha ed
unde s anding plays an impo an pa in he con inuous imp o emen o an o ganiza ion. The
change om a plan-d i en o a alue-d i en cul u e (Schön e al., 2017a), in pa icula , is a
di icul obs acle o a hie a chical en i onmen . Howe e , his change is absolu ely neces-
sa y o a success ul agile ans o ma ion.
5.3 Limi a ions
The au ho s a e awa e ha public adminis a ions belong o a s ic ly egula ed a ea and
ha he laws and egula ions a e highly coun y-dependen . The e o e, one could a gue ha
he esul s may no be ans e able. Wi h ega d o gene alizabili y, he au ho s can make
he ollowing s a emen s. The au ho s de eloped hei model o measu ing agili y in pub-
lic adminis a ion induc i ely based on he a o emen ioned s udies. The adap a ion o he
ques ionnai e o measu ing agili y o he con ex o public adminis a ion was e alua ed
by means o an expe su ey. The e o e, his pa o he esul s is applicable o o he cases
as well. I he concep ual amewo k is looked a mo e closely (see Fig.7), i becomes clea
ha he elemen s can be aced back o he li e a u e. The desc ip ion o he con ex o
public adminis a ion is based on he desc ip ion o a con ex in Yin (2003). Fu he mo e,
he h ee ci cles (agile wo king eam, measu e deg ee o agili y a eam le el, and apply
measu es o imp o e agili y) a e based on he build-measu e-lea n eedback loop by Ries
(2011), adap ed o he imp o emen o an embedded uni . Bo h elemen s can be applied by
any embedded uni . To his eason, he au ho s a e con iden ha he indings a e applica-
ble mo e widely.
In con as o his, he measu emen s wi h he ques ionnai e a e ailo ed o he h ee
eams s udied and a e di icul o be ans e ed o o he cases due o he speci ic con ex in
307
which he eams wo k. Only he measu es iden i ied a e ou comes which a e dependen on
he case s udied. Hence, he iden i ied issues in e ms o agili y migh no be ans e able o
o he eams.
6 Conclusion and u u e wo k
This pape p esen s he esul s o a case s udy, which aims o in es iga e how agili y can be
measu ed in he con ex o public adminis a ions. The e o e, he au ho s deli e ed a ques-
ionnai e o measu e a eam le el. Wi h he help o a ques ionnai e, eams a e now able o
c i ically sha e hei unde s anding o agili y based on me ics. The case s udy included an
applica ion o he ques ionnai e wi h h ee eams in a public adminis a ion in Ge many.
On he one hand, he au ho s collec ed quali a i e da a in wo wo kshops and ecei ed an
ag eed objec i e o he measu emen o agili y in he con ex s udied. The pa icipan s
o he s udy weigh ed he communica i e and he change-a ine dimension as pa icula ly
impo an , he imp o emen -o ien ed and he sel -o ganized dimension as impo an , he
p oduc -d i en dimension as neu al and he i e a i e dimension as a he unimpo an . On
he o he hand, he au ho s ga he ed quan i a i e da a by means o he ques ionnai e o
measu ing agili y a eam le el in public adminis a ions. The s a is ical analysis shows ha
all h ee eams show he need o op imiza ion in he change-a ine dimension. In addi ion,
he da a e eals ha he e may be dys unc ionali ies in Team A, which need o be analyzed
in de ail in u u e wo k. One u he ou come o his s udy is a concep ual amewo k,
which can be used o imp o e he agili y in public adminis a ions. The concep ual ame-
wo k is based on se e al componen s and allows o he op imiza ion o a gi en con ex in
e ms o agili y by means o measu emen s wi h a ques ionnai e. The e o e, he deg ee o
agili y is measu ed a eam le el.
This pape has se e al implica ions o bo h esea che s and p ac i ione s. The adop-
ion o he no ion o agili y in public adminis a ion emains a challenge, since he cul-
u e wi hin his kind o o ganiza ions is no he op imal s a ing poin o he use o agile
p ocess models and agile p ac ices. The au ho s a e awa e ha ans o ma ional change
is a cons an jou ney. The e o e, he measu emen o p og ess plays an impo an ole
in he con inuous imp o emen o an o ganiza ion. This app oach allows o iden i y
issues in e ms o agili y a eam le el based on measu emen s. The issues unco e ed a e
al eady p io i ized by aligning he objec i e in measu ing agili y be o ehand. Thus, i is
clea wha a eam can wo k on o imp o e i s own agili y. This pape p esen ed he i s
i e a ion o he applica ion o he ques ionnai e in a case s udy. Based on he analysis o
he esul s, imp o emen s can now be iden i ied and applied in he u u e. Fu he mo e,
a subsequen measu emen wi h he ques ionnai e should be ca ied ou in o de o com-
pa e he success o he measu es. In u u e wo k, a long- e m s udy will be conduc ed
o in es iga e he op imiza ion o indi idual eams by inco po a ing mul iple i e a ions
using he concep ual amewo k. The measu emen o eams also enables a compa ison
be ween agile eams in he public adminis a ion and agile eams in he p i a e sec o .
Fu he mo e, he ques ionnai e o measu ing agili y in public adminis a ions will be
imp o ed o enable he measu emen o agili y a he o ganiza ional le el and also allow
a compa ison be ween a ious o ganiza ions.
