Timmons, Shane; Ca oll, Eamonn; McGinni y, F ances
Resea ch Repo
Expe imen al es s o public suppo o disabili y policy
Resea ch Se ies, No. 159
P o ided in Coope a ion wi h:
The Economic and Social Resea ch Ins i u e (ESRI), Dublin
Sugges ed Ci a ion: Timmons, Shane; Ca oll, Eamonn; McGinni y, F ances (2023) : Expe imen al es s
o public suppo o disabili y policy, Resea ch Se ies, No. 159, The Economic and Social Resea ch
Ins i u e (ESRI), Dublin,
h ps://doi.o g/10.26504/ s159
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/298323
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/
EXPERIMENTAL TESTS OF PUBLIC SUPPORT FOR
DISABILITY POLICY
SHANE TIMMONS, EAMONN CARROLL AND FRANCES MCGINNITY
RESEARCH
SERIES
NUMBER 159
APRIL 2023
E
V
I
D
E
N
C
E
F
O
R
P
O
L
I
C
Y
EXPERIMENTAL TESTS OF PUBLIC SUPPORT
FOR DISABILITY POLICY
Shane Timmons
Eamonn Ca oll
F ances McGinni y
Ap il 2023
RESEARCH SERIES
NUMBER 159
A ailable o download om www.es i.ie
h ps://doi.o g/10.26504/ s159
The Economic and Social Resea ch Ins i u e
Whi ake Squa e, Si John Roge son’s Quay, Dublin 2
This Open Access wo k is licensed unde a C ea i e Commons A ibu ion 4.0 In e na ional License
(h ps://c ea i ecommons.o g/licenses/by/4.0/), which pe mi s un es ic ed use, dis ibu ion, and
ep oduc ion in any medium, p o ided he o iginal wo k is p ope ly c edi ed.
ABOUT THE ESRI
The mission o he Economic and Social Resea ch Ins i u e (ESRI) is o ad ance
e idence-based policymaking ha suppo s economic sus ainabili y and social
p og ess in I eland. ESRI esea che s apply he highes s anda ds o academic
excellence o challenges acing policymake s, ocusing on 12 a eas o c i ical
impo ance o 21s Cen u y I eland.
The Ins i u e was ounded in 1960 by a g oup o senio ci il se an s led by
D T. K. Whi ake , who iden i ied he need o independen and in-dep h esea ch
analysis o p o ide a obus e idence base o policymaking in I eland.
Since hen, he Ins i u e has emained commi ed o independen esea ch and i s
wo k is ee o any exp essed ideology o poli ical posi ion. The Ins i u e publishes
all esea ch eaching he app op ia e academic s anda d, i espec i e o i s
indings o who unds he esea ch.
The quali y o i s esea ch ou pu is gua an eed by a igo ous pee e iew p ocess.
ESRI esea che s a e expe s in hei ields and a e commi ed o p oducing wo k
ha mee s he highes academic s anda ds and p ac ices.
The wo k o he Ins i u e is dissemina ed widely in books, jou nal a icles and
epo s. ESRI publica ions a e a ailable o download, ee o cha ge, om i s
websi e. Addi ionally, ESRI s a communica e esea ch indings a egula
con e ences and semina s.
The ESRI is a company limi ed by gua an ee, answe able o i s membe s and
go e ned by a Council, comp ising 14 membe s who ep esen a c oss-sec ion o
ESRI membe s om academia, ci il se ices, s a e agencies, businesses and ci il
socie y. The Ins i u e ecei es an annual g an -in-aid om he Depa men o
Public Expendi u e and Re o m o suppo he scien i ic and public in e es
elemen s o he Ins i u e’s ac i i ies; he g an accoun ed o an a e age o 30 pe
cen o he Ins i u e’s income o e he li e ime o he las Resea ch S a egy. The
emaining unding comes om esea ch p og ammes suppo ed by go e nmen
depa men s and agencies, public bodies and compe i i e esea ch p og ammes.
Fu he in o ma ion is a ailable a www.es i.ie
THE AUTHORS
F ances McGinni y is an Associa e Resea ch P o esso a he Economic and Social
Resea ch Ins i u e (ESRI) and Adjunc P o esso o Sociology a T ini y College
Dublin (TCD). Shane Timmons is a Resea ch O ice a he ESRI and Adjunc
P o esso o Psychology a TCD. Eamonn Ca oll is a Pos doc o al Resea ch Fellow
a he ESRI.
ACKNOWLEDGEMENTS
This esea ch o ms pa o he Join Resea ch P og amme, which is unded by he
Na ional Disabili y Au ho i y, on he expe iences o pe sons wi h disabili ies ac oss
key policy a eas a he ESRI. We would like o hank s a a he NDA, in pa icula
Da id Hallinan and Rosalyn Tamming, o help ul eedback on he s udy design and
ma e ials. We a e also especially g a e ul o membe s o he Disabili y Ad iso y
Boa d o hei insigh s and inpu : Jacqui B owne, Fiona Fe is, Ed Ha pe , Sa ah
Moo head, Ma ia Ní Fhla ha a and Aoi e P ice. We also hank Ma C eigh on o
help ul commen s and ad ice on s udy design. We u he app ecia e help ul
commen s om colleagues a he ESRI a an in e nal semina on he indings.
This epo has been accep ed o publica ion by he Ins i u e, which does no i sel ake ins i u ional
policy posi ions. All ESRI Resea ch Se ies epo s a e pee e iewed p io o publica ion. The au ho s
a e solely esponsible o he con en and he iews exp essed. The iews and opinions exp essed in
his publica ion emain hose o he au ho s and do no ep esen he iews o opinions o he Na ional
Disabili y Au ho i y.

Table o con en s |
TABLE OF CONTENTS
EXECUTIVE SUMMARY .............................................................................................................. IX
CHAPTER 1 BACKGROUND TO THE STUDY .............................................................................. 1
1.1 Mo i a ion o he epo ................................................................................... 1
1.2 Challenges aced by disabled people in I eland .................................................. 3
1.3 A i udes owa ds disabled people ..................................................................... 5
1.4 Hidden a i udes and me hods o elici hem ..................................................... 7
1.5 Ina en ion o policy cos s and ade-o s ........................................................ 12
1.6 S udy o e iew .................................................................................................. 14
CHAPTER 2 DATA COLLECTION AND EXPERIMENTAL DESIGN ................................................ 15
2.1 Da a collec ion: pa icipan s ............................................................................. 15
2.2 Selec ion o issues o in es iga e ...................................................................... 16
2.3 Ma e ials and design ......................................................................................... 17
CHAPTER 3 LIST EXPERIMENT RESULTS ................................................................................... 23
3.1 Tes ing design e ec s ....................................................................................... 23
3.2 Inc eased wel a e paymen s ............................................................................. 24
3.3 Housing ............................................................................................................. 25
3.4 Pa king .............................................................................................................. 26
3.5 Familia i y wi h disabili y .................................................................................. 27
CHAPTER 4 POLICY TRADE-OFFS RESULTS .............................................................................. 31
4.1 Cos o li ing suppo ........................................................................................ 31
4.2 Suppo s o child en wi h disabili ies .............................................................. 33
4.3 Building wheelchai accessible in as uc u e .................................................. 34
4.4 Employmen ...................................................................................................... 36
4.5 Familia i y wi h disabili y .................................................................................. 38
CHAPTER 5 DISCUSSION AND IMPLICATIONS ......................................................................... 41
5.1 Suppo o social wel a e ................................................................................. 41
5.2 Accessibili y and in as uc u e......................................................................... 42
5.3 O he social issues ............................................................................................ 43
5.4 Familia i y wi h disabili y .................................................................................. 43
5.5 Limi a ions......................................................................................................... 44
5.6 Implica ions ....................................................................................................... 46
5.7 Conclusion ......................................................................................................... 47
REFERENCES ............................................................................................................................. 49
APPENDIX ................................................................................................................................. 57
i | Expe imen al es s o public suppo o disabili y policy
LIST OF TABLES
Table 2.1 Socio-demog aphic cha ac e is ics o pa icipan s ...................................................... 16
Table 2.2 Policy s a emen s ......................................................................................................... 21
Table 3.1 Chi-squa e es s o lis expe imen andomisa ion ...................................................... 24
Table 3.2a Logis ic eg ession models p edic ing di ec esponses and lis esponses o
inc easing wel a e paymen s ....................................................................................... 25
Table 3.2b Logis ic eg ession models p edic ing di ec esponses and lis esponses o
p io i ising disabled people o social housing ............................................................. 26
Table 3.3 Logis ic eg ession models p edic ing policy suppo om amilia i y wi h
disabili y ....................................................................................................................... 29
Table 4.1 Logis ic eg ession models p edic ing suppo o inc eased cos o li ing
suppo s ........................................................................................................................ 32
Table 4.2 Logis ic eg ession models p edic ing suppo o inc eased child suppo s ............... 34
Table 4.3 Logis ic eg ession model p edic ing suppo o building wheelchai accessible
in as uc u e................................................................................................................ 36
Table 4.4 Logis ic eg ession models p edic ing policy suppo om socio-demog aphic
cha ac e is ics ............................................................................................................... 38
Table 4.5 Logis ic eg ession models p edic ing policy suppo om amilia i y wi h
disabili y ....................................................................................................................... 39
Table A.1 Join dis ibu ions and design e ec es o wel a e paymen lis ............................. 57
Table A.2 Join dis ibu ions and design e ec es o social housing lis .................................. 57
Table A.3 Join dis ibu ions and design e ec es o ca pa king lis ....................................... 57
Table A.4 Logis ic eg ession o cos -o -li ing suppo : educa ion in e ac ion ............................ 59
LIST OF FIGURES
Figu e 2.1a Example lis expe imen design (social housing) ......................................................... 18
Figu e 2.1b Wel a e paymen s lis expe imen ............................................................................... 19
Figu e 2.1c ‘Fille ’ lis expe imen ................................................................................................... 19
Figu e 2.1d Pa king lis expe imen ................................................................................................. 20
Figu e 3.1 Pe cen age o pa icipan s endo sing sensi i e i ems in he h ee lis
expe imen s .................................................................................................................. 27
Figu e 3.2 Suppo o inc easing wel a e paymen s o disabled people by amilia i y wi h
disabili y and expe imen al ea men ......................................................................... 29
Figu e 4.1 Suppo o inc eased cos o li ing associa ed wi h ha ing a disabili y by e sion
...................................................................................................................................... 31
Figu e 4.2 Suppo o inc eased suppo s o child en wi h disabili ies by s a emen
e sion .......................................................................................................................... 33
Figu e 4.3 Suppo o building wheelchai accessible in as uc u e .......................................... 35
Figu e 4.4 Belie s ha people wi h disabili ies should wo k ......................................................... 37
Figu e 4.5 Suppo o policies by amilia i y wi h disabili y ......................................................... 40
Figu e A.1 Dis ibu ion o lis expe imen esponses .................................................................... 58
Abb e ia ions | ii
ABBREVIATIONS
ABC1 Chie income ea ne is in a highe , in e media e o junio
manage ial/p o essional ole
C2DE Chie income ea ne is in a manual o casual wo k ole o
unemployed
CSO Cen al S a is ics O ice
DPO Disabled Pe sons’ O ganisa ions
ESRI Economic and Social Resea ch Ins i u e
IHREC I ish Human Righ s and Equali y Commission
NDA Na ional Disabili y Au ho i y
UNCRPD Uni ed Na ions Con en ion on he Righ s o Pe sons wi h Disabili ies
2 | Expe imen al es s o public suppo o disabili y policy
ime, o i could ins ead e lec an inc easing pe cep ion ha posi i e esponses o
such ques ions a e socially expec ed.2 Ou i s aim was o iden i y he ex en o
‘social desi abili y’, he endency o su ey esponden s o al e hei esponses in
o de o p esen hemsel es in a posi i e ligh , in epo ed a i udes owa ds
disabili y.
This epo also comes a a ime o subs an ial e o m ac oss many policy a eas
cen al o disabled people’s li es. A epo on he cos o disabili y (Depa men o
Social P o ec ion, 2021) has spa ked public deba e abou how much o his cos
should be co e ed by he S a e and how much should be me p i a ely by
indi iduals and amilies. Recen esea ch on employmen among disabled people
shows ha he COVID-19 pandemic had a pa icula ly de imen al impac on his
g oup, exace ba ing an exis ing gap in employmen p ospec s (Eme son e al.,
2021; Eu o ound, 2022). In educa ion, he School Inclusion Model is designed o
make I ish schools mo e inclusi e, while social suppo and medical ca e o
child en wi h disabili ies a e also being e o med h ough he P og essing
Disabili ies p og amme. O e all, hese and o he policy de elopmen s aim o b ing
I eland in line wi h i s commi men s unde he Uni ed Na ions Con en ion on he
Righ s o Pe sons wi h Disabili ies (UNCRPD), adop ed in 2006 and o mally a i ied
by I eland in 2018.
These e o ms will chie ly a ec disabled people, and hei ou comes and
expe iences should be cen al o e alua ion. Howe e , hey come wi h cos s which
will be me by he gene al public. P omo ing a posi i e a i ude owa ds hese
de elopmen s among he public is hence an impo an componen o sus ainable
and e ec i e change. Ye s anda d su eys p io i ise simplici y o su ey i ems and
lea e as implied ha social policies en ail cos s. I esponden s ail o conside he
cos s o policies, suppo in su eys may be a i icially in la ed. Hence, ou second
aim was o de e mine whe he making di e en po en ial unding mechanisms o
policies explici in su ey s a emen s (e.g. a ax inc ease o und he policy)
in luences public suppo o hose policies.
A his junc u e we no e he a ia ion in concep ualisa ion o disabili y. The Uni ed
Na ions Con en ion on he Righ s o Pe sons wi h Disabili ies (UNCRPD) s a es ha :
pe sons wi h disabili ies include hose who ha e long- e m physical,
men al, in ellec ual o senso y impai men s which in in e ac ion wi h
a ious ba ie s may hinde hei ull and e ec i e pa icipa ion in socie y
on an equal basis wi h o he s.
_______________________________________________________________________________
2 Fo example C eigh on e al. (2015) ound ha while di ec ly-exp essed suppo o immig a ion declined in he Uni ed
S a es ollowing he 2008 economic c ash, anonymously exp essed suppo did no change. I was hus he ex en o
social desi abili y in esponding ha changed.

Backg ound o he s udy | 3
The 2016 Census eco ded whe he esponden s ha e a long-las ing di icul y o
condi ion, bu no necessa ily whe he any condi ion hey ha e hinde s hei
pa icipa ion in socie y. Mo eo e , how disabled people a e de ined in law,
esea ch o policy may no be he same as how hey iden i y hemsel es o how
hey a e pe cei ed by o he s. In pa icula , public pe cep ions o disabili y may
cen e on physical, isible disabili ies, and exclude in isible condi ions o men al
heal h issues. The ex en o his may also a y acco ding o amilia i y wi h
disabili y. We e u n o his poin in he concluding chap e .
The emainde o his chap e explo es he mo i a ion o he epo in g ea e
de ail. Fi s , some o he challenges aced by disabled people in I eland a e
conside ed, ollowed by an explo a ion o exis ing li e a u e on a i udes owa ds
disabled people wi h a ocus on esea ch in I eland whe e possible. App oaches
used o elici ‘hidden’ a i udes (i.e. a i udes ha a e held bu no di ec ly
exp essed in su eys) a e hen discussed. We highligh he ew in e na ional
s udies ha ha e employed hese me hods o in es iga e a i udes owa ds
disabili y bu ou ocus is on he ‘lis expe imen ’, which we exploi o his epo
bu , as a as we know, has no ye been used in his ield o esea ch anywhe e in
he wo ld. Finally, we highligh ele an esea ch om beha iou al science on how
people answe su ey ques ions.
1.2 CHALLENGES FACED BY DISABLED PEOPLE IN IRELAND
Acco ding o 2016 Census da a, 643,131 people o 13.5 pe cen o he I ish
popula ion iden i y as ha ing a long-las ing di icul y o condi ion o as
expe iencing di icul y wi h e e yday ac i i ies (CSO, 2017). This p opo ion is
p ojec ed o ha e inc eased by he 2022 Census da a, as he o al popula ion g ows
and ages.3 Disabled people hus ep esen a sizable mino i y g oup in I ish socie y
who ace many ba ie s o ull and equal pa icipa ion in socie y. These challenges
include bu a e no limi ed o he cos o li ing, employmen and access o he
se ices needed o li e a ull and independen li e. Impo an ly, disabled people
ep esen a di e se g oup, wi h needs ha depend on indi idual ci cums ances.
Tha said, e en hose wi h lowe le els o addi ional needs a e a isk o economic
disad an age, s emming om inc eased cos o li ing, lowe le els o employmen ,
and di icul y accessing app op ia e housing and educa ion.
The Cos o Disabili y in I eland Repo (Depa men o Social P o ec ion, 2021)
highligh s he addi ional spending needs aced by many people wi h disabili ies.
These include specialised ca e se ices and equipmen , home adap a ions,
medicines and mobili y cos s (e.g. addi ional use o axis), as well as mo e cos ly
e e yday expendi u e, such as inc eased hea ing cos s and specialised clo hing.
_______________________________________________________________________________
3 The 2022 da a a e likely o epo change in disabili y p e alence beyond changes in he popula ion as he wo ding o
he ele an ques ion was e ised. Responden s in 2022 we e asked whe he each long-las ing di icul o condi ion
impac s hem o a g ea ex en , o some ex en o does no impac hem, a he han a bina y ‘yes’ o ‘no’ esponse o
whe he hey ha e he condi ion as in 2016.
4 | Expe imen al es s o public suppo o disabili y policy
The ex en o hese cos s a ies om low o e y high, depending on indi idual
ci cums ances, bu a e es ima ed o be wi hin he ange o €9,482 o €11,734 pe
yea on a e age. O he s udies in I eland es ima e ha households wi h membe s
who ha e a disabili y ace addi ional cos s ha equa e o be ween 20 pe cen and
37 pe cen o he a e age household income (Cullinan e al., 2011). These cos s
a e a di ec esul o ha ing a disabili y.
The addi ional cos o li ing aced by disabled people is exace ba ed by di e ences
in employmen ou comes. Census da a om 2016 show ha jus one- hi d o
wo king age, disabled people iden i ied employmen as hei main economic s a us
compa ed o wo- hi ds o non-disabled people. This igu e a ies by disabili y
ype, om 15 pe cen o hose wi h an in ellec ual disabili y o 46 pe cen o hose
wi h dea ness o a se ious hea ing impai men , bu employmen le els o
indi iduals wi h all ypes o disabili ies emain signi ican ly below hose o pe sons
wi hou disabili ies (Kelly and Maî e, 2021; see also Wa son e al., 2013). Disabled
people a e a g ea e isk o po e y han non-disabled people. This applies bo h
when compa ing disabled people in employmen wi h non-disabled people in
employmen , and when compa ing disabled people no in employmen wi h non-
disabled people no in employmen (Kelly and Maî e, 2021).
The isk o economic disad an age o disabled people also ex ends o housing.
Analysis o disc imina ion (i.e. less a ou able ea men because o a p o ec ed
cha ac e is ic)4 in he housing ma ke in I eland shows ha people wi h disabili ies
a e wice as likely o epo being disc imina ed agains when compa ed o people
wi hou disabili ies, while con olling o hei employmen s a us (G o i e al.,
2018). This disc imina ion is pa icula ly p onounced among en e s, in line wi h
in e na ional esea ch (e.g. Tu ne , 2005). Ex ending om he abo e, people wi h
disabili ies a e a highe isk o expe iencing disc imina ion in accessing housing
and a e o e - ep esen ed among he homeless popula ion (G o i e al., 2018).
O he analyses show ha people wi h disabili ies expe ience highe le els o
housing dep i a ion and housing a o dabili y issues5 (Russell e al., 2021).
Gaps in socio-economic ou comes be ween disabled and non-disabled indi iduals
appea o begin in childhood. While educa ional a ainmen o disabled people in
I eland is be e han he EU a e age (42 pe cen s. 25 pe cen hold a leas a
pos -seconda y educa ion; Kelly and Maî e, 2021) and he e ha e been signi ican
e o s made o imp o e esou ces and inclusi i y in schools (Ring, o hcoming),
_______________________________________________________________________________
4 Gende , ma i al s a us, amily s a us, age, disabili y, ‘ ace’ (skin colou o e hnic g oup), sexual o ien a ion, eligious
belie , and/o membe ship o he T a elle Communi y a e ‘p o ec ed cha ac e is ics’ unde he legal de ini ion o
disc imina ion.
5 Housing dep i a ion is de ined based on ea u es o a home, speci ically accommoda ion cha ac e ised by one o mo e
o hese ou i ems: leaking oo /damp walls o /window ames o loo o ; lack o cen al hea ing; lack o double
glazing, and lack o ligh . Housing a o dabili y p oblems a e measu ed as, o example, hose expe iencing income
po e y a e housing cos s (household income lowe han 60 pe cen o he median equi alised income a e housing
cos s) o being in a ea s on en and mo gage epaymen s in he las 12 mon hs (see Russell e al., 2021 o u he
de ails).
Backg ound o he s udy | 5
ecen e idence highligh s challenges aced by child en wi h disabili ies. Da a om
he G owing Up in I eland (GUI) s udy show ha young people wi h special
educa ional needs and/o disabili ies a e less likely o like school, mo e likely o
s uggle academically and less likely o p og ess o highe educa ion (Ca oll e al.,
2022). Accessing suppo s bo h inside and ou side o he class oom and e en
en olling in an app op ia e se ing can also be a signi ican s uggle o disabled
young people, wi h ad e se e ec s on hei educa ional ou comes, gene al
wellbeing and wide amily (O’B ien, 2021). The educa ional challenges acing
disabled child en and young people a e compounded by he ac ha hey a e
mo e likely o expe ience economic ulne abili y and a end DEIS schools (schools
which ecei e a ge ed suppo s as hey se e a eas wi h high concen a ions o
socioeconomic disad an age) han hei pee s (Ca oll e al., 2022b).
Beyond economic disad an age and disc imina ion, many disabled people ace
e e yday challenges in accessing se ices. Exac issues depend on indi idual
ci cums ances and ype o disabili y, bu mobili y and anspo di icul ies a e
widesp ead. Fo example, some use o public anspo equi es disabled people
wi h pa icula impai men s and accessibili y needs o gi e no ice 24 hou s in
ad ance, a equi emen no necessa y o o he s in socie y. Access o buildings and
in as uc u e is ci ed as one o he bigges issues acing physically disabled people
(I ish Wheelchai Associa ion, 2020). Th ee-in- ou epo poo access o public
spaces due o s eps being he only op ion a main en ances. While in as uc u al
change depends on policy, public a i udes and beha iou ha e u he
implica ions o accessibili y. Anecdo al e idence and su ey esea ch by he
Disabled D i e s Associa ion o I eland highligh s illegal pa king in disabled bays by
hose wi hou pe mi s as a majo us a ion in he disabili y communi y (Hu on,
2020).
1.3 ATTITUDES TOWARDS DISABLED PEOPLE
The NDA ha e conduc ed egula na ional su eys in I eland on a i udes o
disabili y (NDA, 2017). These su eys we e ca ied ou ace- o- ace wi h a
na ionally ep esen a i e sample. Be ween 2001 and 2017, he e was a gene al
inc ease in he p opo ion ag eeing ha people wi h physical disabili ies,
in ellec ual disabili ies and au ism can pa icipa e ully in li e. Ag eemen has
gene ally been highe among younge people, sugges ing he e may be
gene a ional e ec s alongside a gene al shi in a i udes (ibid). The p opo ion o
people who ag ee ha people wi h disabili ies a e ea ed ai ly in I ish socie y
dec eased om 44 pe cen o 36 pe cen be ween 2011 and 2017, al hough
explaining his dec ease is no s aigh o wa d. I could e lec a pe cei ed
de e io a ion in how disabled people a e ea ed, o g ea e awa eness o he ex a
suppo s needed o ‘ ai ’ ou comes, o a combina ion o bo h. Ei he way, people
in I eland gene ally appea o suppo measu es ha would help disabled people.
In he mos ecen su ey (2017), he majo i y (almos 80 pe cen ) endo sed
inc eased wel a e paymen s o disabled people and suppo ed p io i ising hem
6 | Expe imen al es s o public suppo o disabili y policy
in social housing alloca ion and on hospi al wai ing lis s.6 A simila p opo ion
ag eed ha he e a e some ci cums ances whe e i is igh o ea disabled people
mo e a ou ably han o he s.
Su eys like hese allow o socio-demog aphic p edic o s o a i udes o be
iden i ied. The NDA (2017) su ey shows ha knowing someone wi h a disabili y
p edic s many posi i e a i udes owa ds disabled people, including epo ed
com o wi h ha ing a wo k colleague o neighbou wi h a disabili y and belie ha
disabled people should ha e child en i hey wish. The impo ance o amilia i y
and con ac wi h disabled people is eplica ed in mul iple in e na ional s udies and
e iew pape s, ac oss mul iple measu es (Bu ke e al., 2013; Ju e al., 2013). Fo
example, a US su ey o a i udes o people wi h in ellec ual disabili ies ound ha
mo e equen con ac wi h someone who has an in ellec ual disabili y p edic ed
g ea e suppo o he igh s o people wi h in ellec ual disabili ies and
pe cep ions o hei capabili ies (McConkey e al., 2021). Simila ly, a Swiss su ey
o p e-se ice eache s showed a mo e posi i e a i udes o disabled people
among hose who epo ed highe le els o p io con ac wi h disabled people
(Kunz e al., 2021).
The ocus o su eys should no solely be on he a i ude o he gene al public;
esea ch wi h samples o people wi h disabili ies is i al o unde s anding hei
li ed expe ience. The Qua e ly Na ional Household Su ey (QNHS) in 2014
ea u ed an equali y module which p obed expe iences o disc imina ion in I eland
based on p o ec ed cha ac e is ics, including disabili y. Analyses o his module
ha e shown ha disabled people in I eland a e signi ican ly mo e likely o epo
expe iencing disc imina ion in wo kplaces, while seeking wo k, and in accessing
public and p i a e se ices han non-disabled people (McGinni y e al., 2017). The
size o he di e ence is la ge, wi h disabled people mo e han wice as likely o
epo disc imina ion while seeking employmen and in accessing public se ices
(15.5 pe cen and 7.2 pe cen , espec i ely, s. 6.7 pe cen and 2.8 pe cen
among people wi h no disabili y). Fu he analyses sugges ha disabled people
a e mo e ad e sely a ec ed by disc imina ion when hey expe ience i . Almos
one-in- i e people wi h a disabili y who expe ienced disc imina ion epo ed ha
i had a ‘ e y se ious’ e ec on hem, compa ed o one-in- en o hose wi hou
disabili ies (Banks e al., 2018).
Mo e ecen esea ch in he UK ound ha 72 pe cen o disabled people had
expe ienced nega i e a i udes o beha iou s in he las i e yea s (Moss and
F ounks, 2022; see also Dixon e al., 2018). These include assump ions abou hei
disabili y and wha hey can do, impa ience, dismissal o hei condi ion o need
o accommoda ions, accusa ion o aking o being lazy, exclusion and being
pa onised, along wi h many o he s. These a i udes and beha iou s we e mos
_______________________________________________________________________________
6 Al hough he p io i isa ion o disabled people is no cu en ly a policy in housing o heal hca e, no is i a goal o he
NDA.
Backg ound o he s udy | 7
commonly expe ienced om he gene al public, bu also om employe s,
educa o s, public se ice s a and e en iends and amily membe s.
Unsu p isingly, almos 90 pe cen o hose who expe ienced hese beha iou s said
i had a nega i e e ec on hem, wi h highe p opo ions o 18-34 yea olds and
women epo ing a nega i e e ec .
I is di icul o gauge how likely expe iences in he UK a e o gene alise o I eland,
as he e a e ew simila measu es eco ded in bo h coun ies. Whe e compa isons
can be made, a i udes owa ds disabled people in I eland appea mo e posi i e,
a leas in employmen se ings. Whe eas people in I eland epo being gene ally
com o able wi h he idea o wo king wi h a disabled pe son ( a ing a sco e o 8.9
ou o 10 o a physically disabled pe son and 8.2 o a pe son wi h men al heal h
di icul ies; NDA, 2017), he B i ish Social A i udes su ey (Can and Benne , 2022)
epo s ha ‘ he public suppo equal chances o disabled people in he wo kplace
bu would no necessa ily wan o wo k wi h a disabled pe son’ (p. 2). Howe e ,
analysis o he QNHS Equali y Module ci ed abo e has indica ed ha disabled
wo ke s a e 65 pe cen mo e likely o ace wo kplace disc imina ion han non-
disabled wo ke s, wi h la ge di e ences when seeking employmen (McGinni y e
al., 2017; Banks e al., 2018). Mo eo e , he e a e mul iple in e na ional s udies
showing di e ences in epo ed a i udes owa ds employing people wi h
disabili ies and objec i ely measu ed hi ing p ac ices. Re iews show ha
employe s end o epo posi i e a i udes owa ds he p ospec o hi ing
someone wi h a disabili y (B edgaa d and Salado-Rasmussen, 2021; Bu ke e al.,
2013; Ju e al., 2013), ye ield expe imen s e eal high le els o likely
disc imina ion. Audi expe imen s, in which esea che s apply o job pos ings wi h
ic i ious CVs ha a y sys ema ically by ea u es o in e es (e.g. disabili y
disclosu e), ha e been un in he US, Canada and No way. Findings show ha
people who disclose hei disabili y on a job applica ion a e be ween 26 pe cen
o 50 pe cen less likely o be called o in e iew, despi e ha ing o he wise
equi alen applica ions (Ame i e al., 2018; Bellema e e al., 2018; Bjø nshagen,
2021). These indings poin owa ds he impo ance o exploi ing expe imen al
me hods o gain be e insigh in o a i udes ha may be concealed in s anda d
su eys. The nex sec ion de ails some such me hods o iden i ying concealed
nega i e a i udes among he gene al public.
1.4 HIDDEN ATTITUDES AND METHODS TO ELICIT THEM
Su eys o en s uggle o accu a ely cap u e a i udinal da a abou con o e sial o
sensi i e issues. This sec ion conside s e idence o po en ial bias in su ey
esponses on a ange o opics and me hods o add ess his. Es ima es o a i udes
abou sensi i e opics, such as immig a ion, acism and abo ion, a y depending
on he le el o anonymi y esponden s a e p o ided (C eigh on and Jamal, 2015;
Glynn, 2013; Kulinski e al., 1997; Rosen eld e al., 2016). Repo ed p e alence o
sensi i e beha iou s, such as plagia ism, shopli ing and sexual ac i i y, also a ies
(Chuang e al., 2021; Cou s e al., 2011; K umpal, 2013). Compa isons o ime-use

8 | Expe imen al es s o public suppo o disabili y policy
dia ies wi h su ey esponses show ha people also some imes o e - epo
posi i e a ibu es, such as exe cising and a ending eligious se ices (B enne ,
2011; B enne and De Lama e , 2016). One explana ion o he di e ence be ween
wha people epo when asked di ec ly and wha hey migh exp ess when o e ed
anonymi y is pe cei ed p essu e o gi e a esponse ha is belie ed o be socially
desi able and hus p esen hemsel es posi i ely o o he s. A la ge body o
esea ch sugges s ha his ype o mis epo ing o ce ain a i udes can lead o
sys ema ic e o s. This su ey bias is known as ‘social desi abili y bias’ (K umpal,
2013).
One o he mos s aigh o wa d ways o mi iga e social desi abili y bias is h ough
su ey mode. Su eys adminis e ed in-pe son o o e he phone wi h an
in e iewe end o p oduce lowe es ima es o sensi i e a i udes and beha iou s
han su eys ha a e sel -adminis e ed o conduc ed online (K umpal, 2013;
Paulhus, 2002). Hence one way o measu e sensi i e opinions o beha iou s mo e
accu a ely is o adminis e su eys online in a way ha assu es esponden s o hei
anonymi y.
A numbe o inno a i e echniques ha e been used o u he add ess social
desi abili y bias, each d awing upon he expe imen al me hod. In a su ey
expe imen , some pa icipan s see ques ion o ma s in one way (e.g. a di ec
ques ion) and o he pa icipan s see he same ques ion in ano he o ma (e.g. one
ha equi es an indi ec esponse and o e s g ea e anonymi y). C ucially, he
ques ion o ma ha each pa icipan sees is decided a andom, meaning ha any
agg ega e di e ences can be con iden ly a ibu ed o he di e ence in ques ion
o ma and no o unde lying di e ences be ween pa icipan s. Fo example, i
esponse o ma s ha p o ide g ea e anonymi y show a highe p e alence o
ce ain a i udes (e.g. acism), hen esea che s can be easonably su e ha some
people conceal hei ue a i ude when asked ia s anda d su ey ques ions.
Expe imen al echniques ha e been widely used o measu e a i udes in o he
ields (e.g. acism, sa e sex p ac ices), bu ew s udies ha e es ed social desi abili y
e ec s in su eys abou issues ele an o people wi h disabili ies. This aises
ques ions abou he accu acy o esponses o di ec ques ions as epo ed in
Sec ion 1.3, and sugges s he e is a need o de elop o he me hods o assess
a i udes owa ds disabili y in I eland.
Whe e in e na ional s udies ha e assessed hidden a i udes o disabili y issues,
hey indica e some e idence o social desi abili y bias. Speci ically in ela ion o
a i udes owa ds inclusi e educa ion, Lüke and G osche (2018) used an
expe imen al design whe eby he same su ey was p esen ed o ou di e en
g oups as coming om ou di e en o ganisa ions: a uni e si y, a g oup ha
opposed inclusion as i migh lowe gene al academic s anda ds, a g oup ha
opposed inclusion as i h ea ened suppo o child en wi h disabili ies cu en ly
a ending special schools and a g oup which suppo ed inclusion. They ound
Backg ound o he s udy | 9
e idence o s onge suppo o inclusi e educa ion in su eys pu po ing o be
un by he p o-inclusion g oup, implying ha su eys o disabili y issues may
o e es ima e suppo i esponden s a e awa e o he pu pose o he su ey.
O he expe imen s ha e sough o measu e social desi abili y di ec ly, al hough
wi h po en ially un eliable o po en ially con using me hods. Fo example, he
Implici Associa ion Tes (IAT) es s how quickly and accu a ely people can
associa e posi i e and nega i e wo ds wi h di e en g oups o people (e.g. people
wi h disabili ies and people wi hou disabili ies). Di e ences in speed o p ecision
in he ca ego isa ion a e aken o e lec a di e ence in ‘implici associa ions’
be ween he wo g oups. I people a e slowe o associa e a disabled pe son wi h
a posi i e cha ac e is ic han hey a e a non-disabled pe son, i is hough o e eal
an implici and pe haps unconscious nega i e a i ude owa ds disabili y. One US
s udy which used di e en ypes o IATs showed mo e nega i e implici
associa ions o disabled people han we e epo ed in s anda d su ey measu es
o a i udes (Thomas e al., 2014). The IAT me hod is, howe e , p ima ily ocused
on unconscious bias a he han delibe a e concealing o nega i e a i udes and
esul s canno be easily compa ed o su eys o he gene al popula ion. The IAT
me hod i sel is a om uni e sally accep ed (see o example Mi chell and
Te lock, 2017; Ba els and Schoen ade, 2022). I has poo me hodological igou ,
wi h he same people gene a ing di e en sco es on he same e sion o he es
when aken again (G eenwald and Lai, 2020).
Os apczuk and Musch (2011) used wo app oaches o es ima e socially desi able
esponding o a su ey i em abou eeling ‘uneasy’ in he p esence o people wi h
disabili ies. One app oach asked esponden s o epo wha hey belie ed was
‘mos people’s’ a i ude a he han hei own. Resul s showed a la ge disc epancy
be ween his es ima e o ‘mos people’s’ a i ude and he esponse when
pa icipan s we e asked abou hei own a i udes (55 pe cen s. 8 pe cen o
nega i e a i udes owa ds physical disabili y and 79 pe cen s. 27 pe cen o
men al disabili y). The au ho s acknowledge, howe e , ha he di e ence likely
e lec s esponden s’ o e es ima ion o gene al nega i e sen imen s a he han a
p ojec ion o hei own iews on o o he s.
The second app oach hey employed was he ‘Randomised Response Technique’,
in which pa icipan s a e ins uc ed o answe he sensi i e i em ei he u h ully
o o gi e a speci ic esponse based on an i ele an ule o a known p obabili y
(e.g. o answe ‘yes’ i hei mo he was bo n in Feb ua y, Ma ch o Ap il,
ega dless o he ques ion con en ). As esea che s ha e no insigh in o he
pa icipan ’s expe imen al condi ion, he app oach o e s a laye o anonymi y.
The e is hus no way o knowing whe he an indi idual is answe ing he ques ion
o gi ing a di ec ed esponse, bu he p opo ions o answe s o he ques ion a
he o e all sample le el can be es ima ed. Responses eco ded by Os apczuk and
Musch (2011) we e no s a is ically di e en om he di ec ques ion condi ion
10 | Expe imen al es s o public suppo o disabili y policy
(11 pe cen epo ed unease a ound a physically disabled pe son and 24 pe cen
o men al disabili y). The s udy showed high le els o esponden s no answe ing
as ins uc ed by he andomisa ion de ice, which he au ho s in e as ‘chea ing’
and hence some e idence o social desi abili y bias. Howe e , o he s ha e
obse ed ha he Randomised Response Technique can lead o high non- esponse
a es due di icul y unde s anding he ins uc ions (Rosen eld e al., 2016).
We could loca e no o he s udies ha ha e sough o measu e social desi abili y in
su eys abou disabili y issues, ye he e a e o he echniques ha could be
exploi ed. The ‘lis expe imen ’ o i em coun echnique has been equen ly used
o gauge social desi abili y in bo h sensi i e a i udes and sensi i e beha iou s in
o he domains (see abo e). Like he Randomised Response Technique, lis
expe imen s p o ide esponden s wi h ‘pe manen ’ anonymi y by no asking hem
di ec ly o hei esponse (Chuang e al., 2021).7 Howe e , i is simple o employ
and has lowe non- esponse a es (Rosen eld e al., 2016). As i does no ye appea
o ha e been used in ela ion o a i udes o beha iou s owa ds disabili y, in he
ollowing sec ion we explain i s logic and explo e i s use in ela ion o o he issues.
1.4.1 Lis expe imen s
In a lis expe imen , esponden s a e p o ided wi h a lis o i ems, and hey a e
asked how many o hem hey ag ee wi h. C ucially, hey a e no asked which o
he i ems hey ag ee wi h, simply how many. A con ol g oup, selec ed a andom,
is gi en a lis o h ee i ems co e ing issues o he han he one o in e es , such as
a i udes o educa ion, heal h and he en i onmen . The ea men g oup a e
p esen ed wi h he same lis , bu wi h he addi ion o a ocal i em – in his case a
po en ially sensi i e i em ela ed o disabili y. Because bo h samples a e p esen ed
wi h he same con ol lis and a e andomly assigned, any di e ence be ween he
a e age esponse o he wo g oups is due o he addi ional sensi i e i em. The
echnique has been mos widely used in poli ical science esea ch on o ing and
esea ch on acism and a i udes o immig a ion (K umpal, 2013; Ehle e al., 2021).
The e ha e been wo lis expe imen s ielded in I eland o da e. The i s
in es iga ed a i udes o immig a ion o Black and Muslim g oups (McGinni y e
al., 2020). Resul s showed g ea e suppo o Black han Muslim immig a ion
when esponden s we e asked using s anda d su ey echniques, bu no di e ence
when esponden s we e p o ided g ea e anonymi y; 15 pe cen o indi iduals
we e es ima ed o ha e concealed a nega i e a i ude owa ds Black immig a ion
when asked di ec ly. The second in es iga ed compliance wi h COVID- ela ed
_______________________________________________________________________________
7 Pe manen anonymi y e e s o he ac ha esea che s can ne e iden i y a speci ic indi idual’s esponse because o
how he lis expe imen is designed. This is bo h a s eng h and a limi a ion o he me hod. Fo he esponden , i
means hey a e no asked di ec ly abou hei a i ude and hence a e no mo i a ed o espond in a socially desi able
way. Fo he esea che , his means ha i is no possible o deduce, a an indi idual le el, who was concealing
undesi able a i udes; he me hod elies on compa isons o g oup di e ences be ween di ec esponses and
anonymous esponses o his. No e ha , in ou s udy, all esponden s we e anonymous such ha no pe sonally
iden i iable da a we e collec ed.
Backg ound o he s udy | 11
public heal h beha iou s du ing he pandemic (Timmons e al., 2020). The s udy
showed ha app oxima ely 10 pe cen o pa icipan s o e - epo ed hei
compliance wi h hand-washing and social dis ancing ad ice when asked using
s anda d su ey ques ions compa ed o when asked using a lis .
The ex en o which pa icipan s may conceal hei ue a i ude o beha iou
a ies acco ding o he issue in ques ion and he pa icipan ’s cha ac e is ics.
McGinni y e al. (2020) obse ed concealed nega i e a i udes owa ds Black
immig a ion was highes among hose wi h a uni e si y deg ee ( eplica ing
esea ch in he US; Janus, 2010) and among younge esponden s. In addi ion,
women we e shown o be mo e likely o conceal nega i e a i udes owa ds
Muslim immig a ion han men, bu men we e mo e likely o mask nega i e
a i udes owa ds Black immig a ion (McGinni y e al., 2010). Timmons e al. (2020)
simila ly show gende di e ences in COVID-19 mi iga ion beha iou s, whe e men
o e - epo ed hand-washing when asked di ec ly compa ed o he lis , bu he e
was no such di e ence o women.
Use o he lis expe imen echnique o gain mo e accu a e su ey es ima es has
been alida ed agains objec i e benchma ks. Lis es ima es o o es by a sample
o o e s in an abo ion e e endum in he US we e signi ican ly close o he eal
o e coun han es ima es om s anda d su ey ques ions (al hough some
unde es ima ion emained; Rosen eld e al., 2016). Mo eo e , mul iple
independen me a-analyses show no e idence o publica ion bias wi h lis
expe imen s, meaning ha lis expe imen s ha show social desi abili y bias a e
as likely o be published as hose ha do no , u he s eng hening he e idence
in a ou o he me hod (Blai e al., 2020; Ehle e al., 2021; Li and an den
Noo ga e, 2022; Rosen eld e al., 2016). I is likely o be a mo e eliable way o
es ima e social desi abili y han o he me hods used o da e o disabili y issues
(e.g. he IAT). The echnique speci ically a ge s social desi abili y, has a lowe non-
esponse a e han o he me hods and is less cogni i ely demanding o
pa icipan s han he Randomised Response Technique (Rosen eld e al., 2016).
Tha said, he obus ness o he me hod depends on ce ain design conside a ions.
Con ol lis s should be su icien ly long o a oid pa icipan s suspec ing he ocus
o he expe imen bu sho enough o limi cogni i e demand (Blai e al., 2019).
Lis s should also be designed o p ese e anonymi y; any pa icipan who esponds
wi h he minimum o maximum esponse has e ealed hei esponses. As such,
lis s should be cons uc ed such ha a leas some con ol i ems ha e nega i e
co ela ions o ha one i em is expec ed o gene a e high le els o endo semen
and ano he low le els. The e a e also s a is ical es s ha a e equi ed o
de e mine whe he he p esence o he sensi i e i ems al e s he pa e n o
esponding o he con ol i ems (known simply as a ‘design e ec ’; Blai and Imai,
2012). Expe imen s should also be designed such ha pa icipan s do no suspec
he ocus on he sensi i e i em (Chuang e al., 2021). Fo example, McGinni y e al.
18 | Expe imen al es s o public suppo o disabili y policy
Responses a ied om 0 o 4. Because bo h samples a e p esen ed wi h he same
con ol lis i ems, any di e ence be ween he a e age esponse o he con ol and
ea men is due o he addi ional (sensi i e) i em. A he g oup le el, simply
sub ac ing he a e age esponse o he con ol om he a e age esponse o he
ea men o e s a way o asce ain suppo o he sensi i e i em among hose in
he ea men g oup (i.e. in Figu e 2.1A, p io i ising disabled people o social
housing). An addi ional s ep, which is aken in his expe imen , is o ask he con ol
g oup o di ec ly exp ess hei suppo o sensi i e i ems (e.g. p io i ising disabled
people o social housing) ia s anda d su ey ques ions a e he lis expe imen .
The di e ence be ween di ec ly exp essed suppo on hese ques ions and he
es ima e o anonymously exp essed suppo is in e p e able as a measu e o he
ex en o which suppo o hese issues is o e - o unde -s a ed (see Figu e 2.1A).
The key o he success o he lis expe imen is ha esponden s in he lis
ea men a e ne e asked o a icula e suppo o epo on beha iou ega ding
any speci ic i em in he lis , which gua an ees pe manen anonymi y om he
in e iewe a he indi idual le el.13
FIGURE 2.1A EXAMPLE LIST EXPERIMENT DESIGN (SOCIAL HOUSING)
Sou ce: Au ho s.
No e: Because he con ol i ems a e he same o bo h g oups, any di e ence in a e age esponses can be a ibu ed o suppo o
he sensi i e i em. Fo example, i he a e age esponse o he con ol g oup is 1.5 i ems and he a e age esponse o he lis
g oup is 2.3 i ems, 80 pe cen anonymously endo sed he sensi i e i em in he lis g oup (2.3 – 1.5 = 0.8).
In ou s udy, he so wa e assigned pa icipan s a andom o he ea men g oup
(he ea e he ‘lis ’ g oup; n = 1,248) o he con ol g oup (he ea e , he ‘di ec
ques ion’ g oup; n = 752). We p e- egis e ed a andomisa ion a io o 5:3 in a ou
o he lis ea men , due o s a is ical powe equi emen s o lis expe imen s.
Pa icipan s emained in he same ea men o each lis . All pa icipan s saw ou
lis s o i ems and we e asked how many on each lis hey ag eed wi h o applied o
hem. They selec ed a esponse om a d op-down menu. The i s wo lis s
con ained policy s a emen s and he las wo con ained beha iou s (see
Figu es 2.1a o 2.1d). Fo pa icipan s in he lis ea men , bo h policy lis s and one
_______________________________________________________________________________
13 Unde condi ions o pe manen anonymi y he pe son’s opinion is no eco ded. The su ey in e iewe ne e knows
which o he i ems on he lis he esponden suppo s.

Da a collec ion and expe imen al design | 19
o he beha iou lis s con ained a sensi i e i em abou disabili y issues: wel a e
paymen s, social housing and disabled pa king,14 espec i ely. The con ol g oup
we e asked di ec ly abou hei suppo o p io i ising disabled people o social
housing and inc easing wel a e paymen s o people wi h a disabili y, as well as
whe he hey ha e e e pa ked in a disabled space a he end o he su ey (see
Figu e 2.1a).
The o he beha iou lis , abou equency o exe cise, was a ille lis designed o
p e en pa icipan s om suspec ing he ocus o he s udy was disabili y issues
and al e ing hei esponses as a esul .
FIGURE 2.1B WELFARE PAYMENTS LIST EXPERIMENT
Sou ce: Au ho s.
FIGURE 2.1C ‘FILLER’ LIST EXPERIMENT
Sou ce: Au ho s.
No e: This lis was no o analy ic in e es and did no ha e a co esponding di ec ques ion.
_______________________________________________________________________________
14 While ‘accessible pa king’ is he p e e ed e m o he esea ch eam, ‘disabled pa king’ was used in he su ey as i
was el ha his e m would be mo e amilia o he gene al popula ion.
20 | Expe imen al es s o public suppo o disabili y policy
FIGURE 2.1D PARKING LIST EXPERIMENT
Sou ce: Au ho s.
I ems on lis s we e p esen ed in andomised o de . To minimise he po en ial o
maximum o minimum esponses o he lis s, which would in alida e he
anonymi y lis s a o d, he con ol i ems we e cons uc ed such ha we expec ed
one i em o gene a e high le els o ag eemen , one o gene a e low le els o
ag eemen and one o be suppo ed by app oxima ely hal o pa icipan s,
ollowing bes p ac ice in he design o lis expe imen s (Blai and Imai, 2012).
2.3.2 Policy ade-o s
A e comple ing he igne es ( epo ed sepa a ely in Timmons e al.,
o hcoming), pa icipan s saw ou policy s a emen s and we e asked whe he
hey ag eed wi h each one (‘Yes’, ‘No’ o ‘Don’ Know’). The ou issues we e
selec ed om a wide se o eigh , ou o which conce ned disabili y issues
(wheelchai accessible in as uc u e, cos o li ing suppo s, suppo s o child en
wi h disabili ies and employmen ) and ou conce ned o he issues ( e ugees,
pa en al lea e, he en i onmen and u he educa ion). Each pa icipan saw wo
disabili y and wo non-disabili y issues, selec ed a andom. The ocus o his epo
is on he disabili y- ela ed i ems. These a e p esen ed in Table 2.2.
Da a collec ion and expe imen al design | 21
TABLE 2.2 POLICY STATEMENTS
Ve sion 1
Ve sion 2
Ve sion 3
A. Cos o
Li ing
Con ol: Mo e should
be done o suppo
disabled people in
mee ing he ex a
cos s o li ing ela ed
o ha ing a disabili y.
Budge : Mo e o he
Go e nmen ’s budge should be
alloca ed o helping disabled
people in mee ing he ex a
cos s o li ing ela ed o ha ing
a disabili y.
Tax: A ax inc ease should be
used o pu mo e money
owa ds suppo ing disabled
people in mee ing he ex a
cos s o li ing ela ed o ha ing
a disabili y.
B. Suppo s o
Child en
Con ol: Child en
wi h disabili ies
should ge he
suppo s hey need.
Budge : Mo e o he
Go e nmen ’s budge should be
alloca ed o making su e
child en wi h disabili ies should
ge he suppo s hey need.
Tax: A ax inc ease should be
used o pu mo e money
owa ds making su e child en
wi h disabili ies should ge he
suppo s hey need.
C. Wheelchai
accessible
in as uc u e
Con ol: Local
Councils should
p io i ise building
mo e wheelchai
accessible
in as uc u e.
Pa king: Local Councils should
p io i ise building mo e
wheelchai accessible
in as uc u e, e en i pa king
in as uc u e needs o be
emo ed o do so.
Cycle: Local Councils should
p io i ise building mo e
wheelchai accessible
in as uc u e ins ead o
cycling in as uc u e.
D. Employmen
Con ol: Disabled
people should wo k,
in jobs which hey
a e capable o doing.
Suppo : Disabled people
should be suppo ed in
wo king, in jobs which hey a e
capable o doing.
Incen i e: Disabled people
should be incen i ised o wo k,
in jobs which hey a e capable
o doing.
Sou ces: Au ho s.
The idea o hese policy ques ions was o elici dep h o suppo o p og essi e
disabili y policies, by a ying whe he he po en ial unding mechanisms o ade-
o s o policies we e made explici and he na u e o he policy p oposed. I is well-
es ablished ha people a e o en ‘ a ionally ina en i e’, and hence a e unlikely o
spon aneously conside he wide implica ions o policies whe e hey a e implici
(e.g. Sims, 2003). Fo example, he public may suppo a policy ha p oposes o
inc ease wel a e paymen s o disabled people in p inciple, bu suppo may be
weakened when a en ion is d awn o he cos s o such a policy (see also discussion
in Sec ion 1.5). Fo each issue, we cons uc ed h ee e sions and pa icipan s we e
assigned wi h balanced andomisa ion by he so wa e o see one e sion.
Table 2.2 shows he exac wo ding o each s a emen and hei a ian s. Fo h ee
o he disabili y issues, he e sions a ied by a unding mechanism o ade-o
ha was made explici in he ques ion. Fo he ou h (abou employmen ), he
a ian s p esen ed sligh a ia ions o he na u e o he s a emen . The non-
disabili y issues we e again designed o educe he po en ial o pa icipan s
suspec ing he main ocus o he s udy. Policies we e p esen ed wo pe page in
andomised o de , wi h he cons ain ha wo disabili y issues we e no
p esen ed on he same page.
Pa icipan s inished he s udy by comple ing ques ions abou backg ound
cha ac e is ics, including age, gende , educa ional a ainmen , employmen s a us,
and li ing a ea, he esul s o which a e shown in Table 2.1, as well as whe he hey
hemsel es had a disabili y o whe he hey knew anyone wi h a disabili y.
Lis expe imen esul s | 23
CHAPTER 3
Lis expe imen esul s
In his chap e we i s p esen es s o lis expe imen design assump ions,
including analysis o ully anonymised esponse a es and he p esence o design
e ec s (Blai and Imai, 2012). To compa e lis endo semen o sensi i e i ems o
di ec ques ion esponses, we use he i em-coun echnique wi h Welch’s - es s
o accoun o unequal a iances be ween lis and con ol ea men s (Tsai, 2019).
We use logis ic eg ession models o es o di e ences be ween
sociodemog aphic subg oups. We es o di e ences by gende , age, li ing
loca ion (u ban o u al) and socio-economic indica o s. Socio-economic indica o s
we e educa ional a ainmen and ‘social g ade’, a classi ica ion sys em based on
he occupa ion o he chie income ea ne in he household. We compa e hose in
households whe e he chie income ea ne is in a highe , in e media e o junio
manage ial/p o essional ole (‘ABC1’) wi h ones whe e he chie income ea ne is
in a manual o casual wo k ole o unemployed (‘C2DE’). We use S uden ’s - es s
and es s o p opo ions o ollow-up es s whe e models sugges signi ican
di e ences. We also compa e esponses o di ec ques ions o es ima es om he
mos ecen NDA su ey on disabili y a i udes using es s o p opo ions (NDA,
2017).
3.1 TESTING DESIGN EFFECTS
Lis expe imen s need o be designed such ha he p esence o he sensi i e i em
does no al e how pa icipan s espond o o he i ems in he lis . We es ed o
design e ec s on each lis using he kic package in S a a (Blai and Imai, 2012; Tsai,
2019). Resul s o each lis showed sui able join dis ibu ions and no indica ion o
design e ec s (Tables A.1-A.3 in he Appendix). Response dis ibu ions o lis
expe imen s should also show ew pa icipan s answe ing wi h he minimum o
maximum esponse and he eby e ealing hei opinion. Figu e A.1 in he Appendix
shows no e idence o p oblema ic ceiling o loo e ec s.
Las ly, we es ed o any di e ences be ween he g oups on socio-demog aphic
cha ac e is ics, which should be p e en ed by andomisa ion. Chi-squa e es s
showed ha he e was no di e ence in alloca ion o di e en socio-demog aphic
g oups o he lis o di ec ion condi ion, as signalled by he p alue being much
g ea e han 0.05 (Table 3.1). This implies ha he andomisa ion was e ec i e and
he e we e no signi ican di e ences in hese g oups in e ms o hese socio-
demog aphic cha ac e is ics.

24 | Expe imen al es s o public suppo o disabili y policy
TABLE 3.1 CHI-SQUARE TESTS OF LIST EXPERIMENT RANDOMISATION
χ²
p
Gende
0.19
.660
Age
0.87
.647
Educa ional A ainmen
0.03
.866
Wo king S a us
0.16
.925
Socio-Economic G ade
0.24
.888
Li ing A ea
1.11
.292
Disabili y S a us
2.27
.132
Sou ce: Au ho s’ analysis.
3.2 INCREASED WELFARE PAYMENTS
When asked di ec ly, 77 pe cen o pa icipan s endo sed inc eased wel a e
paymen s o disabled people. This igu e is he same as he le el o suppo in
esponse o he same ques ion, es ima ed om he 2017 NDA su ey on disabili y
a i udes (N = 1,294), (77 pe cen ) Z = 0.00, p = .998 (see Figu e 3.1). Es ima es o
suppo om he lis esponses we e lowe (66 pe cen ), (1,620.69) = 2.70,
p = .007. This di e ence implies ha 14 pe cen (11 pe cen age poin s) o people
who endo se inc easing wel a e paymen s o disabled people may do so only o
p esen hemsel es in a posi i e ligh . They do no suppo inc easing wel a e
paymen s when p o ided anonymi y in he lis expe imen (Figu e 3.1).
Table 3.2a p esen s model coe icien s om logis ic eg essions p edic ing suppo
o inc easing wel a e paymen s o di e en socio-demog aphic g oups o
pa icipan s. The able also shows esul s om a non-linea leas -squa es
es ima ion o suppo wi hin he lis ea men (see Blai and Imai, 2012; Tsai,
2019).
Table 3.2a shows ha , compa ed o hose wi hou deg ees, pa icipan s wi h
highe educa ional a ainmen we e less likely o endo se inc eased wel a e
paymen s o disabled people when asked in he lis condi ion,15 bu he e was no
e idence o a di e ence when bo h g oups we e asked di ec ly.16
_______________________________________________________________________________
15 Di ec - es : (1237.76) = 1.99, p = .047.
16 Tes o p opo ions: Z = 0.74, p = .458.
Lis expe imen esul s | 25
TABLE 3.2A LOGISTIC REGRESSION MODELS PREDICTING DIRECT RESPONSES AND LIST
RESPONSES TO INCREASING WELFARE PAYMENTS
Inc ease Wel a e
Paymen s
Di ec Lis
Man
(Re : Woman)
0.11
(0.17)
0.01
(0.35)
Aged 45+ yea s
(Re : <45 yea s)
0.12
(0.18)
-0.36
(0.36)
Deg ee
(Re : Less han Deg ee)
0.00
(0.20)
-0.86**
(0.40)
ABC1 Social G ade
(Re : C2DEF)
-0.27
(0.19)
0.38
(0.39)
U ban
(Re : Ru al)
0.00
(0.18)
-0.36
(0.37)
Cons an
1.22
1.27
N
746
2,000
Sou ces: Au ho s’ analysis and NDA (2017).
No e: *p < .10, ** p < .05, *** p < .01. S anda d e o s a e in pa en heses.
The sample size o he Di ec model a ies om he con ol condi ion sample size because subg oups in which each indi idual
esponded in he same way a e excluded. The sample size o he Lis model is he ull sample because he model equi es
esponses om bo h he con ol lis and he ea men lis o es ima e he p opo ion who endo sed he a ge i em wi hin
each subg oup.
To es o educa ional a ainmen di e ences in social desi abili y bias, we
epea ed he i em-coun analysis o bo h g oups. Those wi h deg ees displayed a
social desi abili y bias, wi h a highe pe cen age endo sing an inc ease when asked
di ec ly (75.6 pe cen ) compa ed o when asked in he lis ea men (58.8 pe
cen ), (799.04) = 3.76, p < .001. The di e ence o hose wi h lowe educa ional
a ainmen was in he same di ec ion bu no s a is ically signi ican (78.0 pe cen
s. 71.1 pe cen , espec i ely), (949.51) = 1.30, p = .195. Hence, he esul s imply
ha indi iduals wi h highe educa ional a ainmen a e less likely o endo se an
inc ease in wel a e paymen s o people wi h a disabili y, and ha his di e ence
may no be de ec ed using s anda d su ey me hods.
3.3 HOUSING
Suppo o p io i ising disabled people o social housing showed a la ge decline
in 2022 compa ed o 2017, when pa icipan s we e asked di ec ly (61.4 pe cen in
2022 s. 78.0 pe cen in 2017), Z = 8.04, p < .001 (see Figu e 3.1). The lis condi ion
showed highe le els o suppo compa ed o he di ec condi ion (70.7 pe cen ),
(1,619.65) = 1.96, p = .050, bu s ill ma ginally lowe han in 2017,
(1,247.13) = 1.68, p = .094. No e ha , con a y o ou p edic ions, he di ec ion
o he di e ence be ween he di ec and lis ea men implies ha p io i ising
disabled people on social housing wai ing lis s is judged as less socially desi able
han opposing i .
Table 3.2b shows ha lowe social g ade was a p edic o o di ec suppo o
housing p io i isa ion. Analysis o esponses wi hin hese g oups showed ha while
26 | Expe imen al es s o public suppo o disabili y policy
66.5 pe cen o hose in he C2DE (skilled and unskilled manual wo ke s, non-
employed) social g ades suppo ed p io i ising disabled people o social housing,
jus 56.9 pe cen o hose in he ABC1 (p o essional, manage ial, adminis a i e)
g ades did.17 Howe e , he lis model in Table 3.2b shows no di e ence be ween
social g ades in he lis ea men .18
We in es iga ed he ela ionship be ween social g ade and social desi abili y
u he by epea ing he i em-coun analysis by subg oups. Those in he C2DE social
g ades did no display a lis e ec , wi h 72.0 pe cen suppo ing in he lis
condi ion, (768.20) = 0.78, p = .437, bu hose in he ABC1 g ades showed highe
suppo in he lis condi ion (69.6 pe cen ), (848.47) = 1.98, p = .047. Hence, he
esul s imply ha s anda d su ey measu es may indica e a di e ence in suppo
o housing p io i isa ion depending on social g ade ha does no exis when
esponden s a e o e ed g ea e anonymi y. Impo an ly, social desi abili y may
lead o lowe suppo o disabled people among g oups in highe social g ades.
One possibili y is ha , du ing a housing c isis whe e he e a e high le els o
compe i ion o limi ed social housing, some o hose in highe social g ades
pe cei e p io i ising any g oup o be socially undesi able.
TABLE 3.2B LOGISTIC REGRESSION MODELS PREDICTING DIRECT RESPONSES AND LIST
RESPONSES TO PRIORITISING DISABLED PEOPLE FOR SOCIAL HOUSING
Housing P io i y
Di ec
Lis
Man
(Re : Woman)
0.20
(0.15)
0.25
(0.44)
Aged 45+ yea s
(Re : <45 yea s)
-0.03
(0.15)
0.08
(0.44)
Deg ee
(Re : Less han Deg ee)
0.21
(0.17)
-0.05
(0.48)
ABC1 Social G ade
(Re : C2DEF)
-0.52***
(0.17)
-0.09
(0.47)
U ban
(Re : Ru al)
0.13
(0.16)
0.18
(0.45)
Cons an
0.49
0.73
N
751
2,000
Sou ces: Au ho s’ analysis.
No e: *p < .10, ** p < .05, *** p < .01. S anda d e o s a e in pa en heses.
The sample size o he Di ec model a ies om he con ol condi ion sample size because subg oups in which each indi idual
esponded in he same way a e excluded. The sample size o he Lis model is he ull sample because he model equi es
esponses om bo h he con ol lis and he ea men lis o es ima e he p opo ion who endo sed he a ge i em wi hin
each subg oup.
3.4 PARKING
In he 2017 NDA su ey, jus 2 pe cen o esponden s epo ed ha hey judge
pa king in a disabled pa king space wi hou a pe mi is accep able. As we we e
in e es ed in beha iou a he han a i udes o his issue, in he 2022 su ey
_______________________________________________________________________________
17 Tes o p opo ions: Z = 2.69, p = .007.
18 T- es : (1246) = 0.26, p = .798.
Lis expe imen esul s | 27
esponden s we e asked whe he hey had pa ked in a disabled pa king space
wi hou a pe mi . Fo he analysis, we include only hose who la e in he su ey
epo ed ha hey d i e (n = 1,674). When asked di ec ly in ou s udy, 4 pe cen
o d i e s admi ha ing done so in he pas and his di e ence is s a is ically
signi ican , Z = 2.57, p = .010 (see Figu e 3.1). The es ima e om esponses o he
lis i ems (4.8 pe cen ) is no s a is ically di e en o he p opo ion who admi ed
pa king in a disabled space wi hou a pe mi o di ec ques ion, (1162.94) = 0.22,
p = .830. The numbe o people who epo ed pa king in a disabled space wi hou
a pe mi is oo low o pe mi analysis o di e ences by socio-demog aphic
subg oups.
FIGURE 3.1 PERCENTAGE OF PARTICIPANTS ENDORSING SENSITIVE ITEMS IN THE THREE
LIST EXPERIMENTS
Sou ces: Au ho s’ analysis and NDA (2017).
No e: E o ba s a e he s anda d e o o he p opo ion. *The pa king ques ion con ains only hose who epo ed d i ing (n = 1,674).
The compa ison wi h NDA (2017) is o a ques ion abou he accep abili y o pa king in a disabled space wi hou a pe mi a he
han a beha iou ques ion.
3.5 FAMILIARITY WITH DISABILITY
We we e in e es ed in he ex en o which con ac wi h disabled people o
expe ience wi h disabili y is associa ed wi h a i udes o he policy suppo . We
p e- egis e ed explo a o y analyses o he associa ion be ween knowing someone
wi h a disabili y o ha ing a disabili y onesel and suppo o disabili y policy.19 We
an u he logis ic eg ession models p edic ing policy endo semen using an
o dinal a iable o amilia i y. Pa icipan s we e asked a he end o he s udy
whe he hey o any o a lis o people hey knew had a disabili y o long las ing
condi ion ha a ec s hei abili y o ca y ou day- o-day ac i i ies.20 They we e
_______________________________________________________________________________
19 Fo his analysis also, he numbe o people who epo ed pa king in a disabled spo wi hou a pe mi is oo low o
subg oup compa isons.
20 The esponse op ions we e Spouse/Pa ne , Child, Pa en , B o he /Sis e /O he ela i e, F iend, Neighbou , Colleague/
wo k con ac , No su e/don’ know, None. Pa icipan s could selec as many as applied o hem.
77 78
2
77
61.4
4
66 70.7
4.8
0
10
20
30
40
50
60
70
80
90
Inc ease wel a e paymen s o
disabled people
P io i ise disabled people o
social housing
I ha e pa ked in a disabled spo
wi hou a pe mi *
Policy A i udes Beha iou
2017 2022 - Di ec 2022 - Lis
34 | Expe imen al es s o public suppo o disabili y policy
TABLE 4.2 LOGISTIC REGRESSION MODELS PREDICTING SUPPORT FOR INCREASED CHILD
SUPPORTS
Table 4.2 Logis ic
Reg ession Models
P edic ing Suppo o
Inc eased Child Suppo s
Con ola ( 1) Budge ( 2) Tax ( 3)
Male
(Re : Female)
-0.11
(0.84)
-0.57*
(0.32)
0.11
(0.23)
Age
(Re : 18-39 yea s)
40-59 yea s
-0.59
(0.95)
0.85**
(0.38)
0.06
(0.28)
60+ yea s
0.11
(1.27)
0.70*
(0.41)
0.37
(0.30)
Deg ee
(Re : No Deg ee)
-0.99
(0.94)
0.10
(0.36)
0.09
(0.26)
Social G ade
(Re : DEF)
C1C2
-0.35
(1.19)
-0.09
(0.40)
-0.02
(0.28)
AB
-0.44
(1.31)
-0.34
(0.48)
0.41
(0.35)
U ban
(Re : Ru al)
-
-0.01
(0.35)
-0.30
(0.25)
In e cep
5.12***
(1.33)
1.67***
(0.43)
0.45
(0.35)
N Yes Responses
329
284
212
To al N
334
333
334
Sou ce: Au ho s’ analysis.
No e: *p < .10, ** p < .05, *** p < .01. S anda d e o s a e in pa en heses.
a. All excep o six o he 334 pa icipan s in his model suppo ed he policy, all o whom li ed in an u ban a ea.
Hence, li ing a ea is no included as a p edic o in his model.
4.3 BUILDING WHEELCHAIR ACCESSIBLE INFRASTRUCTURE
Suppo o policy C on building mo e wheelchai accessible in as uc u e also
a ied ac oss condi ions, χ² (4) = 34.82, p < .001. Suppo was s onges when no
ade-o was made explici (Ve sion 1) bu declined when he p oposal equi ed
emo ing pa king in as uc u e, (Ve sion 2) Z = 2.36, p = .018, o o i o be buil
a he expense o cycling in as uc u e, (Ve sion 3) Z = 5.15, p < .001. Suppo was
s onge i pa king spaces we e o be emo ed han i cycling in as uc u e would
no be buil , (Ve sion 2 s. Ve sion 3) Z = 2.85, p = .004.

Policy ade-o esul s | 35
FIGURE 4.3 SUPPORT FOR BUILDING WHEELCHAIR ACCESSIBLE INFRASTRUCTURE
Sou ce: Au ho s’ analysis.
Ac oss all models o suppo (Table 4.3), men we e less in a ou han women
(71 pe cen s. 81 pe cen ) and olde pa icipan s we e mo e suppo i e o
wheelchai accessible in as uc u e han younge pa icipan s (83 pe cen o o e
60s, 77 pe cen o 40-59 yea olds and 71 pe cen o unde 40s). Highe socio-
economic g ades we e less suppo i e han he lowes g ade g oups when ade-
o s we e made explici (71 pe cen o he AB g oup and 69 pe cen o he C1C2
g oup s. 81 pe cen o he DEF g oup), whe eas he e was no di e ence when he
ade-o was no explici (86 pe cen o he AB g oup and 82 pe cen o he C1C2
g oup and 86 pe cen o he DEF g oup).
84.15 76.88
66.96
4.88 12.61
18.45
10.98 10.51 14.58
0
10
20
30
40
50
60
70
80
90
100
No Explici T ade-O ( 1) By emo ing pa king
in as uc u e ( 2)
Ins ead o cycling in as uc u e
( 3)
C. Local councils should build mo e wheelchai in as uc u e
Yes No Don' Know
36 | Expe imen al es s o public suppo o disabili y policy
TABLE 4.3 LOGISTIC REGRESSION MODEL PREDICTING SUPPORT FOR BUILDING
WHEELCHAIR ACCESSIBLE INFRASTRUCTURE
Wheelchai accessible
in as uc u e
Con ol ( 1) Pa king ( 2) Cycle ( 3)
Male
(Re : Female)
-0.84***
(0.32)
-0.34
(0.27)
-0.51**
(0.25)
Age
(Re : 18-39 yea s)
40-59 yea s
0.10
(0.36)
0.47
(0.32)
0.47*
(0.28)
60+ yea s
1.09**
(0.51)
0.32
(0.35)
0.82**
(0.33)
Deg ee
(Re : No Deg ee)
-0.40
(0.35)
0.44
(0.30)
-0.44
(0.27)
Social G ade
(Re : DEF)
C1C2
-0.17
(0.40)
-0.83**
(0.38)
-0.38
(0.33)
AB
0.30
(0.50)
-0.57
(0.44)
-0.42
(0.39)
U ban
(Re : Ru al)
0.16
(0.33)
0.06
(0.27)
0.44*
(0.26)
In e cep
1.97***
(0.52)
1.54***
(0.42)
0.87**
(0.38)
N Yes Responses
276
256
225
To al N 327 332 336
Sou ce: Au ho s’ analysis.
No e: *p < .10, ** p < .05, *** p < .01. S anda d e o s a e in pa en heses.
4.4 EMPLOYMENT
Unlike he p e ious policies, he e sions o he inal policy ame (policy D) did
no a y by whe he po en ial cos s o unding mechanisms we e made explici .
Ins ead, he policies, which aimed a encou aging disabled people o wo k in jobs
hey a e capable o doing a ied depending on how ha encou agemen was
ph ased. Ve sion 1 assessed a gene al belie ha people wi h disabili ies should
wo k (‘Disabled people should wo k, in jobs which hey a e capable o doing’),
whe eas Ve sion 2 desc ibed disabled people being suppo ed o wo k (‘Disabled
people should be suppo ed in wo king, in jobs which hey a e capable o doing’),
and Ve sion 3 implied ha wo king should be mo i a ed by inancial incen i es
(‘Disabled people should be incen i ised o wo k, in jobs which hey a e capable o
doing’). Resul s showed an o e all e ec o policy e sion, χ² (4) = 35.78, p < .001.
When amed wi h gene al suppo in Ve sion 2, mo e people endo sed he policy
han when asked abou hei gene al belie in Ve sion 1, Z = 4.41, p < .001, o when
suppo was implied o be pu ely inancial in Ve sion 3, Z = 5.75, p < .001. The e
was no di e ence be ween Ve sion 1 and Ve sion 3, Z = 1.53, p = .126.
Policy ade-o esul s | 37
FIGURE 4.4 BELIEFS THAT PEOPLE WITH DISABILITIES SHOULD WORK
Sou ce: Au ho s’ analysis.
Socio-demog aphic di e ences in endo sing policies aimed a encou aging
disabled people o wo k showed simila di e ences by age, social g ade and
disabili y s a us ac oss each e sion o he s a emen (See Table 4.4). Those aged
o e 60 mo e s ongly endo sed he policies han younge esponden s (96 pe
cen s. 88 pe cen o 40-59 yea olds and 87 pe cen o unde 40s) and hose in
he AB and C1C2 g oups mo e s ongly endo sed he policies han hose in he DEF
g oup (92 pe cen o bo h g oups s. 83 pe cen , espec i ely). Younge
esponden s and hose in he DEF g oup show esponse pa e ns mos simila o
hose wi h a disabili y, whe e 82 pe cen endo sed he policies compa ed o 92
pe cen o hose wi hou a disabili y.
88.1
97.02
83.99
2.08 0.6 5.14
9.82 2.38 10.88
0
10
20
30
40
50
60
70
80
90
100
Con ol ( 1) Suppo ed ( 2) Incen i ised ( 3)
D. Should wo k in jobs hey a e capable o
Yes No Don' Know
38 | Expe imen al es s o public suppo o disabili y policy
TABLE 4.4 LOGISTIC REGRESSION MODELS PREDICTING POLICY SUPPORT FROM SOCIO-
DEMOGRAPHIC CHARACTERISTICS
Employmen
Con ol ( 1)
Suppo ( 2)
Incen i e ( 3)
Male
(Re : Female)
-0.62*
(0.35)
-0.26
(0.67)
0.03
(0.33)
Age
(Re : 18-39 yea s)
40-59 yea s
0.11
(0.38)
1.14
(0.853)
0.12
(0.35)
60+ yea s
1.31**
(0.54)
1.59
(1.10)
1.85***
(0.58)
Deg ee
(Re : No Deg ee)
0.28
(0.41)
0.05
(0.71)
0.23
(0.37)
Social G ade
(Re : DEF)
C1C2
0.42
(0.40)
1.14
(0.71)
1.39***
(0.37)
AB
0.41
(0.53)
1.58
(1.16)
0.79*
(0.48)
U ban
(Re : Ru al)
0.50
(0.35)
0.10
(0.69)
0.08
(0.33)
In e cep
1.31**
(0.48)
2.11***
(0.74)
0.48
(0.37)
N Yes Responses
296
326
278
To al N
333
335
330
Sou ce: Au ho s’ analysis.
No e: *p < .10, ** p < .05, *** p < .01. S anda d e o s a e in pa en heses.
4.5 FAMILIARITY WITH DISABILITY
As be o e, we conduc ed explo a o y analyses o he associa ion be ween
amilia i y wi h disabili y and suppo . Table 4.5 p esen s logis ic eg ession models
using he same amilia i y a iable as be o e on whe he he pa icipan endo sed
he policy. The models include con ols o socio-demog aphic cha ac e is ics and
he e sion o he policy he pa icipan saw. Ou ocus on his sec ion is on
di e ences be ween hose who know no one wi h a disabili y, hose whose
pa ne , child o pa en has a disabili y and hose who ha e a disabili y hemsel es,
as he numbe o pa icipan s in he o he g oups a e oo low o eliable es ima es
bu a e shown in Table 4.5 o comple eness ( iend/neighbou /colleague ns = 60
o 67;22 b o he /sis e /o he ela i e ns = 56 o 70).
Resul s show ha pa icipan s who hemsel es ha e a disabili y we e mo e likely
o endo se mo e cos -o -li ing suppo s (policy A) han hose who know no one
wi h a disabili y (Figu e 4.5). Tes s o coe icien s showed he e was no di e ence
be ween hose wi h a disabili y hemsel es and hose whose pa ne , child o
pa en has a disabili y, χ² = 1.46, p = .226. The pa e n is simila on policy D abou
whe he disabled people should wo k. Those wi h a disabili y and hose whose
pa ne , child o pa en has a disabili y we e less suppo i e o s a emen s ha
_______________________________________________________________________________
22 Numbe o obse a ions pe policy a ies because o andomisa ion o see wo o he ou policies.
Policy ade-o esul s | 39
disabled people ‘should wo k’ han hose who know no one wi h a disabili y, bu
he e was no s a is ically signi ican di e ence be ween hese wo g oups, χ² =
0.15, p = .670. This may sugges a gap be ween disabled people’s a i udes owa ds
and expe iences o seeking employmen and non-disabled people’s a i udes
owa ds disabled people seeking employmen .
TABLE 4.5 LOGISTIC REGRESSION MODELS PREDICTING POLICY SUPPORT FROM
FAMILIARITY WITH DISABILITY
Sou ce: Au ho s’ analysis.
No e: *p < .10; **p < .05; ***p < .01
On policy C abou wheelchai accessible in as uc u e, hose ela ed o someone
wi h a disabili y we e mo e suppo i e han hose who know no one wi h a
disabili y and he e was no di e ence be ween he pa ne /child/pa en g oup and
hose wi h a disabili y hemsel es, χ² = 0.33, p = .567. The e we e no di e ences
on policy B abou suppo ing child en wi h disabili ies, e lec ing he high le el o
endo semen ac oss all socio-demog aphic subg oups o his policy.
Looking ac oss he policies (Figu e 4.5), he e was no e idence o a di e ence in
opinion be ween hose wi h a disabili y hemsel es and hose closes o someone
wi h a disabili y (i.e. a pa ne , child o pa en has a disabili y).
A. Cos o Li ing B. Suppo o
Child en
C. Wheelchai
accessible
in as uc u e
D. Employmen
Familia i y wi h Disabili y
(Re : Knows No One)
F iend/Neighbou /Colleague
-0.01
(0.34)
0.01
(0.36)
0.08
(0.33)
0.33
(0.63)
B o he /Sis e /O he Rela i e
-0.28
(0.34)
0.17
(0.36)
0.91**
(0.42)
0.39
(0.56)
Spouse/Pa ne /Child/Pa en
0.29
(0.23)
0.05
(0.29)
0.40*
(0.24)
-0.70**
(0.32)
Has a Disabili y
0.64***
(0.23)
0.33
(0.26)
0.24
(0.22)
-0.83***
(0.26)
Socio-Demog aphic Con ols
Yes Yes Yes Yes
Cons an
2.18
(0.31)
3.89
(0.49)
1.73
(0.29)
2.79
(0.43)
N
997
1,003
997
1,001

40 | Expe imen al es s o public suppo o disabili y policy
FIGURE 4.5 SUPPORT FOR POLICIES BY FAMILIARITY WITH DISABILITY
Sou ce: Au ho s’ analysis.
68.1
81.3
71.9
92.1
76.3
82.7 81.1
85.8
78.7
85.6
79.8 82.0
0
10
20
30
40
50
60
70
80
90
100
A. Mo e suppo o
ex a cos o li ing
B. Mo e suppo o
child en wi h disabili ies
C. Local councils should
build mo e wheelchai
in as uc u e
D. Should wo k in jobs
hey a e capable o
% pa icipan s
Knows No One Pa ne /Child/Pa en Has a Disabili y
Discussion and implica ions | 41
CHAPTER 5
Discussion and implica ions
The majo i y o people in I eland suppo mos policies ha aim o enable people
wi h disabili ies o pa icipa e ully in socie y, e en when p o ided comple e
anonymi y in answe ing he ques ion o when unding mechanisms o policies a e
made explici . This suppo is obse ed ac oss cos -o -li ing, access o housing, and
day- o-day accessibili y policies, and is consis en ac oss socio-demog aphic
subg oups. Howe e , he deg ee o suppo can a y depending on how su eys
a e implemen ed and he le el o de ail p o ided in su ey ques ions. In his
chap e , we summa ise he main esul s om bo h s ages o he s udy, make no e
o some limi a ions and highligh he policy implica ions om he indings.
5.1 SUPPORT FOR SOCIAL WELFARE
The addi ional cos o ha ing a disabili y is es ima ed o be €9,482 o €11,734 pe
yea (Depa men o Social P o ec ion, 2021). Ou esul s sugges ha , al hough
hey may no be awa e o his igu e, mos people in I eland acknowledge he
inancial challenges o li ing wi h a disabili y and a e suppo i e o s onge
go e nmen assis ance. When asked using s anda d su ey echniques, 77 pe cen
o people ag ee ha wel a e paymen s o disabled people should be inc eased
and mos (91 pe cen ) ag ee ha mo e should be done o suppo disabled people
o mee hei ex a cos o li ing. Howe e , suppo a ies as a unc ion o su ey
anonymi y and whe he unding mechanisms a e included in he su ey i em.
When endo semen o inc eased wel a e paymen s is elici ed using a lis
echnique, whe eby esponden s gi e hei iews unde condi ions o pe manen
anonymi y, suppo is 14 pe cen lowe (11 pe cen age poin s) a 66 pe cen . This
di e ence means ha one-in-se en who would ha e suppo ed he policy i asked
di ec ly do no suppo he policy when o e ed g ea e anonymi y. The di e ence
in suppo o ex a cos -o -li ing assis ance is simila (15 pe cen lowe ) when he
policy is p oposed o be unded h ough budge a y ealloca ion compa ed o when
no de ail is gi en on he unding mechanism. The al e na i e o budge
ealloca ion, a ax inc ease, leads o a mo e d as ic di e ence in suppo (54 pe
cen lowe han when no unding mechanism is speci ied), wi h only 42 pe cen in
a ou o mo e suppo in his g oup.
This a ia ion in suppo o inancial assis ance o disabled people depending on
how ques ions a e asked is g ea e among hose wi h highe educa ional
a ainmen . Resul s om he lis expe imen showed ha , when p o ided
pe manen anonymi y, suppo among hose educa ed o deg ee le el o abo e
was 22 pe cen lowe . Suppo among hose educa ed below deg ee le el was jus
9 pe cen lowe and was no s a is ically signi ican ly di e en om suppo when
asked di ec ly. Simila ly, s a is ical models showed ha he di e ence in suppo
42 | Expe imen al es s o public suppo o disabili y policy
o ex a cos -o -li ing assis ance h ough budge ealloca ion compa ed o when
no mechanism was speci ied was mo e p onounced among hose educa ed o
deg ee le el o abo e (20 pe cen lowe ) han hose wi hou deg ees (9 pe cen ).
Impo an ly, s anda d su ey echniques may o e s a e suppo among hose wi h
highe educa ional quali ica ions o socially desi able measu es (Janus, 2010;
McGinni y e al., 2020). In his s udy, he di ec ques ion on inc eased wel a e
paymen s ailed o de ec di e ences by educa ional a ainmen , whe eas esul s
om he lis expe imen ound hose wi h highe educa ional a ainmen o be less
suppo i e o inc eased wel a e paymen s. The con ol condi ion o he policy
s a emen expe imen , which would be a mo e ypical o mula ion, sugges ed
hose wi h highe educa ional a ainmen a e mo e suppo i e o assis ance wi h
he cos o li ing han hei pee s wi h lowe educa ional a ainmen , whe eas
when p esen ed he budge alloca ion ade-o he g oups did no di e . These
di e ences by educa ional a ainmen may e lec a g ea e endency o hose on
highe incomes o oppose edis ibu i e iscal policies in gene al (Mülle and
Regan, 2021).
Child en wi h disabili ies a e especially elian on addi ional suppo in o de o
achie e ull inclusion (e.g. Ca oll e al., 2022a), and almos e e yone (98 pe cen )
ac oss all socio-demog aphic subg oups endo ses such suppo s when unding
mechanisms a e no speci ied. As wi h gene al cos -o -li ing suppo , howe e ,
suppo is lowe when unding mechanisms a e included in su ey i ems, wi h
la ge di e ences o ax inc eases (35 pe cen lowe ) han o budge ealloca ion
(13 pe cen lowe ). I is wo h no ing ha he p opo ion in a ou o a ax inc ease
o help child en wi h disabili ies (63.5 pe cen ) is much la ge han he le el o
suppo o ax inc eases o o he social issues ha ha e high le els o in-p inciple
suppo , such as clima e change (40-47 pe cen ; Leahy, 2021; Timmons and Lunn,
2022) and a Uni ed I eland (22 pe cen ; Sheahan, 2021).
5.2 ACCESSIBILITY AND INFRASTRUCTURE
Day- o-day accessibili y issues u he con ibu e o challenges o he igh s o
disabled people, including, bu no limi ed o, access o sui able pa king and
insu icien in as uc u e. P e ious su eys show ha a e y small mino i y (2 pe
cen ) judge pa king in a disabled pa king space wi hou a pe mi o be accep able
(NDA, 2017). Howe e , ou indings show ha a small bu signi ican ly highe
p opo ion o d i e s epo ha ing done so in he pas (4 pe cen ). When p o ided
pe manen anonymi y, he igu e is 4.8 pe cen (20 pe cen highe ), al hough he
di e ence is no s a is ically signi ican . No e ha , in o de o de ec a s a is ically
signi ican di e ence o beha iou s wi h e y low incidence and he e o e a lowe
absolu e bias-pe cen age, a sample in excess o 10,000 would ha e been equi ed
(Blai e al., 2020). None heless, a ue incidence a e o e en 4 pe cen implies
ha one-in-25 d i e s ha e pa ked in disabled pa king spaces wi hou a pe mi . I
hese d i e s a e ‘ epea o ende s’, his may lead o disp opo iona ely high
encoun e s o such beha iou among disabled people. This es ima e does no
Discussion and implica ions | 43
include o he s who may no pa k di ec ly in disabled pa king spaces bu may
impede access o hem in o he ways. We a e ca e ul o no e also, howe e , ha
he e may be legi ima e use o disabled pa king spaces included in his es ima e
(e.g. aking a ela i e o an appoin men and o ge ing he pe mi o equi ing a
pe mi bu no mee ing echnical equi emen s).23
Tu ning o in as uc u e, he majo i y (84 pe cen ) suppo he p oposi ion ha
local councils should build mo e wheelchai accessible in as uc u e, wi h suppo
highes among olde people. As wi h he cos -o -li ing suppo s, he p opo ion
who suppo he policy is lowe when su ey ques ions highligh a po en ial cos
o ade-o . Fo his policy, he po en ial ade-o was one o al e na i e
p io i isa ion, whe e wheelchai accessible in as uc u e came wi h he emo al
o d i ing in as uc u e (pa king spaces) o ins ead o ac i e anspo
in as uc u e (cycle lanes). Suppo is 7 pe cen lowe in he condi ion whe e
pa king in as uc u e would be emo ed and e en lowe (20 pe cen lowe ) in he
condi ion whe e cycling in as uc u e would be de-p io i ised. No e howe e ha
emo ing pa king in as uc u e and no building cycling in as uc u e a e no
necessa y ade-o s o implemen ing wheelchai in as uc u e (e.g. some may
belie e he in as uc u e should be ins alled bu wi hou cos o pa king o cycling
in as uc u e), bu a e me ely illus a i e o he change in suppo when ade-
o s a e speci ied.
5.3 OTHER SOCIAL ISSUES
O e ing pe manen anonymi y when asked abou social wel a e inc eases al e ed
esponses in line wi h he hypo hesis ha some suppo o disabili y policy is due
o social desi abili y bias. Howe e , he lis expe imen abou housing p io i y
sugges ed he opposi e. Responden s we e mo e suppo i e o p io i ising disabled
people o social housing han when hey we e asked di ec ly abou hei opinion.
This inding may sugges ha people iew he socially accep able esponse o be
one whe e no one g oup is ‘p io i ised’ du ing a pe iod o housing sho ages and
a o dabili y challenges. When p o ided mo e anonymi y, hey e eal p e e ences
held owa ds some g oups (e.g. disabled people). Al e na i ely, o in addi ion, he
lis echnique may be de ec ing nega i e sen imen owa ds o he g oups ha
compe e o limi ed social housing supply (e.g. e ugees). I is no possible o
iden i y wha sen imen d i es he e e se-lis e ec obse ed he e, bu his
sugges ion is in line wi h majo i y suppo o disabili y policies and o he e idence
o concealed nega i e a i udes owa ds immig a ion (McGinni y e al., 2020).
5.4 FAMILIARITY WITH DISABILITY
The esul s also e eal s iking di e ences be ween hose mos amilia wi h
disabled people and disabili y issues (i.e. people wi h disabili ies hemsel es and
hei pa ne s/child en/pa en s) and hose who epo no knowing anyone wi h a
_______________________________________________________________________________
23 We hank he ad iso y boa d o he s udy o making us awa e o bo h o hese limi a ions.
50 | Expe imen al es s o public suppo o disabili y policy
B enne , P.S. (2011). ‘Excep ional Beha iou o Excep ional Iden i y: O e epo ing
o Chu ch A endance in he US’, Public Opinion Qua e ly 75(1): 19-41.
B enne , P.S. and J.D. DeLama e (2016). ‘Social Desi abili y Bias in Sel - epo s o
Physical Ac i i y: Is Exe cise Iden i y he Culp i ?’, Social Indica o s
Resea ch 117(2): 489-504.
Bu ke, J., J. Bezyak, R.T. F ase , J. Pe e, N. Di chman and F. Chan (2013). ‘Employe s’
a i udes owa ds hi ing and e aining people wi h disabili ies: A e iew o
he li e a u e’, The Aus alian Jou nal o Rehabili a ion Counselling, 19(1),
21-38.
Can , J. and M. Benne (2022). ‘Disabled People a Wo k’, in Bu , S., E. Cle y and J.
Cu ice (eds.) B i ish Social A i udes: The 39 h Repo . London: Na ional
Cen e o Social Resea ch.
Ca oll, E., S. McCoy and G. Mihu (2022). ‘Explo ing Cumula i e Disad an age in
Ea ly School Lea ing and Planned Pos -School Pa hways among Those
Iden i ied wi h Special Educa ional Needs in I ish P ima y Schools’, B i ish
Educa ional Resea ch Jou nal 00: 1-18. h ps://doi.o g/10.1002/be j.3815.
Ca oll, E., K. Ye and S. McCoy (2022). ‘Educa ionally main ained inequali y? The
ole o isk ac o s and esilience a 9, 13 and 17 in disabled young people’s
pos -school pa hways a 20’, I ish Educa ional S udies, 41:3, 573-594, DOI:
10.1080/03323315.2022.2093257.
Chen, C., B. MacInnis, M. Wal man and J.A. K osnick (2021). ‘Public opinion on
clima e change in he USA: o wha ex en can i be nudged by
ques ionnai e design ea u es?’, Clima ic Change, 167(3), 1-18.
h ps://doi.o g/10.1007/s10584-021-03194-x.
Chuang, E., P. Dupas, E. Huille y and J. Seban (2021). ‘Sex, lies, and measu emen :
Consis ency es s o indi ec esponse su ey me hods’, Jou nal o
De elopmen Economics, 148, 102582. DOI:
h ps://doi.o g/10.1016/j.jde eco.2020.102582.
Cou s, E., B. Jann, I. K umpal and A. Nähe (2011). ‘Plagia ism in S uden Pape s:
P e alence Es ima es Using Special Techniques o Sensi i e Ques ions’,
Jah büche ü Na ionalökonomie und S a is ik, Vol. 231, no. 5-6, pp. 749-
760. h ps://doi.o g/10.1515/jbns -2011-5-612.
C eigh on, M.J. and A. Jamal (2015). ‘Does Islam play a ole in an i-immig an
sen imen ? An expe imen al app oach’, Soc Sci Res.53, 89-103. DOI:
10.1016/j.ss esea ch.2015.04.001.
C eigh on, M.J., A. Jamal and N. Malancu (2015). ‘Has opposi ion o immig a ion
inc eased in he Uni ed S a es a e he Economic C isis? An Expe imen al
App oach’, In e na ional Mig a ion Re iew, Vol. 49 (3), 727-756.
h ps://doi.o g/10.1111/im e.12091.
CSO (2017). ‘Census o Popula ion 2016 – P o ile 9 Heal h, Disabili y and Ca e s’,
e ie ed om h ps://www.cso.ie/en/ eleasesandpublica ions/ep/p-
cp9hdc/p8hdc/p9d/.
Cullinan, J., B. Gannon and S. Lyons (2011). ‘Es ima ing he ex a cos o li ing o
people wi h disabili ies’, Heal h economics, 20(5), 582-599.
h ps://doi.o g/10.1002/hec.1619.

Re e ences | 51
Depa men o Social P o ec ion (2021). The Cos o Disabili y in I eland. Re ie ed
om h ps://www.go .ie/en/publica ion/1d84e- he-cos -o -disabili y-in-
i eland- esea ch- epo /.
Dixon, S., C. Smi h and A. Touche (2018). The disabili y pe cep ion gap. SCOPE.
D uckman, J.N. (2001). ‘The implica ions o aming e ec s o ci izen
compe ence’, Poli ical beha io , 23(3), 225-256.
h ps://doi.o g/10.1023/A:1015006907312.
Ehle , I., F. Wol e and J. Junke mann (2021). ‘Sensi i e Ques ions in Su eys: A
Comp ehensi e Me a-Analysis o Expe imen al Su ey S udies on he
Pe o mance o he I em Coun Technique’, Public Opinion Qua e ly, Vol.
85, Issue 1, Sp ing 2021, pp 6-27, h ps://doi.o g/10.1093/poq/n ab002.
Eme son, E., R. S ancli e, C. Ha on, G. Llewellyn, T. King, V. To sika, Z. Ai ken and
A. Ka anagh (2021). ‘The impac o disabili y on employmen and inancial
secu i y ollowing he ou b eak o he 2020 COVID-19 pandemic in he UK’,
Jou nal o Public Heal h, 43(3), 472-478. DOI:
h ps://doi.o g/10.1093/pubmed/ daa270.
Eu o ound (2022). People wi h disabili ies and he COVID-19 pandemic: Findings
om he Li ing, wo king and COVID-19 e-su ey, COVID-19 se ies.
Luxembou g: Publica ions O ice o he Eu opean Union.
E ans, J.B.S.T. (2008). ‘Dual-p ocessing accoun s o easoning, judgmen , and social
cogni ion’, Annual Re iew o Psychology, 59, 255-278. DOI:
h ps://doi.o g/10.1146/annu e .psych.59.103006.093629.
Feldman, L. and P.S. Ha (2018). ‘Clima e change as a pola izing cue: F aming
e ec s on public suppo o low-ca bon ene gy policies’, Global
En i onmen al Change, 51, 54-66. DOI:
h ps://doi.o g/10.1016/j.gloen cha.2018.05.004.
Glynn, A.N. (2013). ‘Wha Can We Lea n wi h S a is ical T u h Se um? Design and
Analysis o he Lis Expe imen ’, Public Opinion Qua e ly, Vol. 77, pp.
159-172.
Gollus , S.E., J. Niede deppe and C.L. Ba y (2013). ‘F aming he consequences o
childhood obesi y o inc ease public suppo o obesi y p e en ion
policy’, Ame ican jou nal o public heal h, 103(11), e96-e102.
h ps://doi.o g/10.2105/AJPH.2013.301271.
G eenwald, A.G. and C.K. Lai (2020). ‘Implici social cogni ion’, Annual Re iew o
Psychology, 71, 419-445. DOI: h ps://doi.o g/10.1146/annu e -psych-
010419-050837.
G o i, R., H. Russell, É. Fahey and B. Maî e (2018). Disc imina ion and inequali y
in housing in I eland. I ish Human Righ s and Equali y Commission.
Hu wi z, J. and M. Pe ley (2005). ‘Playing he ace ca d in he pos -Willie Ho on
e a: The impac o acialized code wo ds on suppo o puni i e c ime
policy’, Public Opinion Qua e ly, 69(1), 99-112.
h ps://doi.o g/10.1093/poq/n i004.
52 | Expe imen al es s o public suppo o disabili y policy
Hu on, B. (July 6, 2020). ‘Fixed-cha ge ines o illegally pa king in disabled spaces
soa ’, I ish Times. h ps://www.i ish imes.com/news/social-a ai s/ ixed-
cha ge- ines- o -illegally-pa king-in-disabled-spaces-soa -1.3555499.
I ish Wheelchai Associa ion (2020). Bes P ac ice Access Guidelines.
Janus (2010). ‘The In luence o Social Desi abili y P essu es on Exp essed
Immig a ion A i udes’, Social Science Qua e ly, 91(4), pp. 925-946.
Ju, S., E. Robe s and D. Zhang (2013). ‘Employe a i udes owa d wo ke s wi h
disabili ies: A e iew o esea ch in he pas decade’, Jou nal o Voca ional
Rehabili a ion, 38(2), 113-123.
Kahneman, D. (2011). Thinking, Fas and Slow. New Yo k: Fa a , S auss, Gi oux.
Kelly, E. and B. Maî e (2021). Iden i ica ion O Skills Gaps Among Pe sons Wi h
Disabili ies And Thei Employmen P ospec s, ESRI Su ey and S a is ical
Repo Se ies 107, Dublin: ESRI, h ps://doi.o g/10.26504/sus a 107.
K osnick, J.A. (1991). ‘Response s a egies o coping wi h he cogni i e demands
o a i ude measu es in su eys’, Applied cogni i e psychology, 5(3), 213-
236. DOI: h ps://doi.o g/10.1002/acp.2350050305.
K osnick, J.A. and S. P esse (2010). ‘Ques ion and Ques ionnai e Design’ in W igh ,
J. D. and P.V. Ma sden (Eds). Handbook o Su ey Resea ch. Eme ald G oup
Publishing.
K umpal I. (2013). ‘De e minan s o social desi abili y bias in sensi i e su eys: a
li e a u e e iew’, Qual Quan 47, 2025-47.
Kuklinski, J.H., M.D. Cobb and M. Gilens (1997). ‘Racial a i udes and he “New
Sou h”‘, Jou nal o Poli ics, Vol. 59, pp. 323-49.
Kunz, A., R. Lude and W. Kassis (2021). ‘Belie s and a i udes owa d inclusion o
s uden eache s and hei con ac wi h people wi h disabili ies’, F on ie s
in Educa ion 6.
Leahy, P. (Oc 8, 2021). ‘Poll shows high deg ee o public esis ance o many
po en ial clima e ac ion measu es’, I ish Times.
h ps://www.i ish imes.com/news/i eland/i ish-news/poll-shows-high-
deg ee-o -public- esis ance- o-many-po en ial-clima e-ac ion-measu es-
1.4694287.
Li, J. and W. Van den Noo ga e (2022). ‘A Me a-analysis o he ela i e
e ec i eness o he i em coun echnique compa ed o di ec
ques ioning’, Sociological Me hods & Resea ch, 51(2), 760-799.
Lopez de Leon, F.L. and R. Rizzi (2014). ‘A es o he a ional igno ance hypo hesis:
E idence om a na u al expe imen in B azil’, Ame ican Economic Jou nal:
Economic Policy, 6(4), 380-98. DOI: 10.1257/pol.6.4.380.
Lüke, T. and M. G osche (2018). ‘Wha do I hink abou inclusi e educa ion? I
depends on who is asking. Expe imen al e idence o a social desi abili y
bias in a i udes owa ds inclusion’, In e na ional Jou nal o Inclusi e
Educa ion, 22:1, 38-53, DOI: 10.1080/13603116.2017.1348548.
Re e ences | 53
Ma inelli, C. (2007). ‘Ra ional igno ance and o ing beha io ’, In e na ional
Jou nal o Game Theo y, 35(3), 315-335. h ps://doi.o g/10.1007/s00182-
006-0051-4.
McCa e y, E.J. and J. Ba on (2004). ‘F aming and axa ion: E alua ion o ax
policies in ol ing household composi ion’, Jou nal o Economic
Psychology, 25(6), 679-705. h ps://doi.o g/10.1016/j.joep.2003.07.001.
McConkey, R., P. Sla e , A. Smi h, L. Dubois and A. Shella d (2021). ‘Pe cep ions o
he igh s and capabili ies o people wi h in ellec ual disabili y in he
Uni ed S a es’, J Appl Res In ellec Disabil. 34: 537-545.
h ps://doi.o g/10.1111/ja .12819.
McGinni y, F., D. Wa son and G. Kings on (2012). Analysing he Expe ience o
Disc imina ion in I eland: E idence om he QNHS Equali y Module 2010,
Dublin: The Equali y Au ho i y and he ESRI,
h ps://www.es i.ie/publica ions/analysing- he-expe ience-o -
disc imina ion-in-i eland-e idence- om- he-qnhs-equali y.
McGinni y, F., M. C eigh on and E. Fahey (2020). Hidden e sus e ealed a i udes:
A lis expe imen on suppo o mino i ies in I eland. I ish Human Righ s
and Equali y Commission.
McGinni y, F., R. G o i, O. Kenny and H. Russell (2017). Who expe iences
disc imina ion in I eland? E idence om he QNHS Equali y Modules. I ish
Human Righ s and Equali y Commission.
Mi chell, G. and P.E. Te lock (2017). ‘Popula i y as a poo p oxy o u ili y: The case
o implici p ejudice’, in Psychological science unde sc u iny: Recen
challenges and p oposed solu ions, 164-195.
Moss, C. and A. F ounks (2022). A i udes and disabili y: The expe iences o
disabled people in 2022. Scope.
Mu, R., Y. Li and T. Cui (2021). ‘Policy na a i e, policy unde s anding and policy
suppo in en ion: a su ey expe imen on ene gy conse a ion’, Policy
S udies, 1-21. h ps://doi.o g/10.1080/01442872.2021.1954609.
Mülle , S. and A. Regan (2021). ‘A e I ish o e s mo ing o he le ?’, I ish Poli ical
S udies, 36(4), 535-555.
h ps://doi.o g/10.1080/07907184.2021.1973737.
NDA (2017). Na ional Su ey o Public A i udes o Disabili y in I eland. Re ie ed
om h ps://nda.ie/uploads/publica ions/2017-Main-Repo -Na ional-
su ey-o -public-a i udes.pd .
O’B ien, C. (2021). ‘Hund eds o special-needs child en unable o access school
places’. The I ish Times.
h ps://www.i ish imes.com/news/educa ion/hund eds-o -special-needs-
child en-unable- o-access-school-places-1.4609933.
Os apczuk, M. and J. Musch (2011). ‘Es ima ing he p e alence o nega i e
a i udes owa ds people wi h disabili y: a compa ison o di ec
ques ioning, p ojec i e ques ioning and andomised esponse’, Disabili y
and Rehabili a ion, 33:5, 399-411, DOI: 10.3109/09638288.2010.492067.
54 | Expe imen al es s o public suppo o disabili y policy
Paulhus, D.L. (2002). ‘Socially Desi able Responding: The E olu ion o a Cons uc ’,
in The Role o Cons uc s in Psychological and Educa ional Measu emen .
h ps://doi.o g/10.4324/9781410607454-10.
Po umbescu, G., N. Bellé, M. Cucciniello and G. Nasi (2017). ‘T ansla ing policy
anspa ency in o policy unde s anding and policy suppo : E idence om
a su ey expe imen ’, Public Adminis a ion, 95(4), 990-1008.
h ps://doi.o g/10.1111/padm.12347.
Pullen, E., E.A. Ekl, E. Felix, C. Tu ne , B.L. Pe y and B.A. Pescosolido (2022).
‘Labeling, causal a ibu ions, and social ne wo k ies o people wi h
men al illness’, Social Science & Medicine, 293 DOI:
h ps://doi.o g/10.1016/j.socscimed.2021.114646.
Reynolds, J.P., K. S au z, M. Pilling, S. an de Linden and T.M. Ma eau (2020).
‘Communica ing he e ec i eness and ine ec i eness o go e nmen
policies and hei impac on public suppo : a sys ema ic e iew wi h me a-
analysis’, Royal Socie y Open Science, 7(1), 190522.
h ps://doi.o g/10.1098/ sos.190522.
Ring, E. ( o hcoming). Special Educa ion in an Independen I eland, 1922-2022.
Rosen eld, B., K. Imai and J.N. Shapi o (2016). ‘An empi ical alida ion s udy o
popula su ey me hodologies o sensi i e ques ions’, Ame ican Jou nal
o Poli ical Science, 60(3), 783-802.
DOI: h ps://doi.o g/10.1111/ajps.12205.
Russell, H., I. P i alko, F. McGinni y and S. En igh (2021). Moni o ing adequa e
housing in I eland. Economic and Social Resea ch Ins i u e.
Schwa z, N. and F. S ack (1985). ‘Cogni i e and a ec i e p ocesses in judgmen s
o subjec i e well-being: A p elimina y model’, in B ands a e H. and E.
Ki chle (Eds), Economic psychology (pp. 439-447). Linz, Aus ia: Taune .
Schwa z, N., F. S ack, H.J. Hipple and G. Bishop (1991). ‘The impac o
adminis a ion mode on esponse e ec s in su ey measu emen ’, Applied
Cogni i e Psychology. 5(3), 193-212. DOI:
h ps://doi.o g/10.1002/acp.2350050304.
Sheahan, F. (May 1, 2021). ‘Majo i y a ou a uni ed I eland, bu jus 22pc would
pay o i ’, I ish Independen . h ps://www.independen .ie/i ish-
news/cen ena ies/cen ena ypoll/majo i y- a ou -a-uni ed-i eland-bu -
jus -22pc-would-pay- o -i -40375875.h ml.
Sims, C.A. (2003). ‘Implica ions o a ional ina en ion’, Jou nal o mone a y
Economics, 50(3), 665-690. h ps://doi.o g/10.1016/S0304-
3932(03)00029-1.
Taylo , S.E. and S.T. Fiske (1978). ‘Salience, a en ion, and a ibu ion: Top o he
head phenomena’, in Ad ances in expe imen al social psychology (Vol. 11,
pp. 249-288). Academic P ess.
Thomas, A., E.D. Vaughn, A. Doyle and R. Bubb (2014). ‘Implici Associa ion Tes s
o A i udes Towa d Pe sons Wi h Disabili ies’, The Jou nal o Expe imen al
Educa ion. 82(2), 184-204.
Re e ences | 55
Timmons, S., F. McGinni y, C. Bel on, M. Ba jako a and P. Lunn (2020). ‘I depends
on how you ask: measu ing bias in popula ion su eys o compliance wi h
COVID-19 public heal h guidance’, Jou nal o Epidemiology and Communi y
Heal h. doi.o g/10.1136/jech-2020-215256.
Timmons, S. and P. Lunn (2022). Public unde s anding o clima e change and
suppo o mi iga ion. Economic and Social Resea ch Ins i u e (ESRI)
Resea ch Se ies Repo No. 135. h ps://doi.o g/10.26504/ s135.
Timmons, S., C.A. Bel on, D. A. Robe son, M. Ba jako á, C. La in, H. Julienne and
P.D. Lunn (2022). ‘Is i iskie o mee 100 people ou doo s o 14 people
indoo s? Compa ing public and expe pe cep ions o COVID-19 isk’,
Jou nal o Expe imen al Psychology: Applied.
h ps://doi.o g/10.1037/xap0000399.
Timmons, S., E. Ca oll and F. McGinni y ( o hcoming). ‘Social judgemen s o
men al and physical disabili ies: E idence om a igne e expe imen in
I eland’. mimeo.
Timoney, A. (2022). Long Co id: De ini ion, incidence and heal hca e issues.
T i Handoyo R., A. Ali, K. Scio and A. Hassio is (2021). ‘A i udes o key
p o essionals owa ds people wi h in ellec ual disabili ies and hei
inclusion in socie y: A quali a i e s udy in an Indonesian con ex ’,
T anscul u al Psychia y. 58(3), 379-391.
DOI:10.1177/1363461520909601.
Tsai, C.L. (2019). ‘S a is ical analysis o he i em-coun echnique using S a a’, The
S a a Jou nal, 19(2), 390-434.
Tu ne , M.A. (2005). Disc imina ion agains pe sons wi h disabili ies: Ba ie s a
e e y s ep. Diane Publishing.
T e sky, A. and D. Kahneman (1980). The aming o decisions and he a ionali y
o choice. S an o d Uni Ca Dep o Psychology.
Uni ed Na ions Con en ion on he Righ s o Pe sons wi h Disabili ies (2006).
Decembe 13, 2006. A ailable a :
h ps://www.ohch .o g/en/h bodies/c pd/pages/con en ion igh spe son
swi hdisabili ies.aspx.
Wa son, D., G. Kings on and F. McGinni y (2013). Disabili y in he I ish Labou
Ma ke : E idence om he QNHS Equali y Module 2010. Dublin: Equali y
Au ho i y/ESRI.

Appendix | 57
Appendix
TABLE A.1 JOINT DISTRIBUTIONS AND DESIGN EFFECT TEST FOR WELFARE PAYMENT LIST
Coe . Robus SE Z p λ p
P (R=0,S=1)
0.01
0.01
1.31
.90
P (R, S=0)
0.00
1.00
P (R=0,S=0)
0.02
0.00
5.15
1.00
P (R, S=1)
0.00
1.00
P (R=1,S=1)
0.30
0.02
13.61
1.00
P (R=1,S=0)
0.19
0.01
14.49
1.00
P (R=2,S=1)
0.26
0.02
14.52
1.00
P (R=2,S=0)
0.11
0.02
4.79
1.00
P (R=3,S=1)
0.10
0.01
11.47
1.00
P (R=3,S=0)
0.02
0.01
1.15
.88
Sou ce: Au ho s’ analysis.
TABLE A.2 JOINT DISTRIBUTIONS AND DESIGN EFFECT TEST FOR SOCIAL HOUSING LIST
Coe .
Robus
SE
Z p λ p
P (R=0,S=1) 0.07 0.01 5.70 1.00 P (R, S=0) 0.00 1.00
P (R=0,S=0)
0.04
0.01
7.29
1.00
P (R, S=1)
0.00
1.00
P (R=1,S=1)
0.23 0.02 10.71 1.00
P (R=1,S=0) 0.12 0.02 6.95 1.00
P (R=2,S=1)
0.27
0.02
13.90
1.00
P (R=2,S=0)
0.12 0.02 5.22 1.00
P (R=3,S=1) 0.13 0.01 13.93 1.00
P (R=3,S=0)
0.02 0.02 1.04 .85
Sou ce: Au ho s’ analysis.
TABLE A.3 JOINT DISTRIBUTIONS AND DESIGN EFFECT TEST FOR CAR PARKING LIST
Coe .
Robus SE
Z
p
λ
p
P (R=0,S=1)
0.07
0.01
5.70
1.00
P (R, S=0)
0.00
1.00
P (R=0,S=0)
0.04
0.01
7.29
1.00
P (R, S=1)
0.00
1.00
P (R=1,S=1)
0.23
0.02
10.71
1.00
P (R=1,S=0)
0.12
0.02
6.95
1.00
P (R=2,S=1)
0.27
0.02
13.90
1.00
P (R=2,S=0)
0.12
0.02
5.22
1.00
P (R=3,S=1)
0.13
0.01
13.93
1.00
P (R=3,S=0)
0.02
0.02
1.04
.85
Sou ce: Au ho s’ analysis.
Figu e A.1 shows he dis ibu ion o esponses o each o he lis s. A small mino i y
o pa icipan s in he wel a e and housing lis s (2.1 pe cen and 4.1 pe cen ,
espec i ely) e ealed a socially undesi able belie (i.e. a esponse o ‘0’). The
undesi able esponse (‘4’) o he pa king lis was 2.1 pe cen . Hence he
58 | Expe imen al es s o public suppo o disabili y policy
anonymi y o socially undesi able esponding was su icien ly low o a oid aising
design conce ns.
FIGURE A.1 DISTRIBUTION OF LIST EXPERIMENT RESPONSES
Sou ce: Au ho s’ analysis.
No e: A esponse o ‘4’ is no possible o he di ec condi ion.
Cos o Li ing by Educa ion In e ac ion
Table A.4 shows ha suppo o he policy declined o bo h unding mechanisms
ela i e o he con ol among hose wi h no deg ee, bu he decline in suppo was
e en la ge among hose wi h a deg ee, pa icula ly o he budge ealloca ion
policy. Tes s o coe icien s showed ha suppo in he budge condi ion was
weake among hose wi h deg ees han wi hou , χ² = 4.43, p = .035, bu he e was
no di e ence in he ax condi ion, χ² = 1.49, p = .222.
2.08 4.1 2.1
0
10
20
30
40
50
60
70
80
012340123401234
Wel a e Housing Pa king
%
Response
Di ec Lis
Appendix | 59
TABLE A.4 LOGISTIC REGRESSION OF COST-OF-LIVING SUPPORT: EDUCATION
INTERACTION
Con ol ( 1)
Male
(Re : Female)
0.35
(0.40)
Age
(Re : 18-39 yea s)
40-59 yea s
0.23
(0.44)
60+ yea s
0.95
(0.59)
Educa ion x S a emen In e ac ion
(Re : No Deg ee- Con ol ( 1))
No Deg ee – Budge ( 2)
-0.65*
(0.90)
No Deg ee – Tax ( 3)
-2.52***
(0.28)
Deg ee – Con ol ( 1)
0.97*
(0.44)
Deg ee – Budge ( 2)
-1.07*
(0.51)
Deg ee – Tax ( 3)
-0.60
(0.49)
Social G ade
(Re : DE)
C1C2
-0.36
(0.55)
AB
-0.64
(0.65)
U ban
(Re : Ru al)
-0.29
(0.41)
In e cep
2.05***
(0.63)
N Yes Responses
301
To al N
329
Sou ce: Au ho s’ analysis.
No e: *p < .10, ** p < .05, *** p < .01. S anda d e o s a e in pa en heses.