Kno ek, Edwa d S.; Mi chell, James; Pedemon e, Ma hieu; Shi o , Taylo
Wo king Pape
The e ec s o in e es a e inc eases on consume s'
in la ion expec a ions: The oles o in o medness and
compliance
IDB Wo king Pape Se ies, No. IDB-WP-1641
P o ided in Coope a ion wi h:
In e -Ame ican De elopmen Bank (IDB), Washing on, DC
Sugges ed Ci a ion: Kno ek, Edwa d S.; Mi chell, James; Pedemon e, Ma hieu; Shi o , Taylo
(2024) : The e ec s o in e es a e inc eases on consume s' in la ion expec a ions: The oles
o in o medness and compliance, IDB Wo king Pape Se ies, No. IDB-WP-1641, In e -Ame ican
De elopmen Bank (IDB), Washing on, DC,
h ps://doi.o g/10.18235/0013184
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/309115
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/3.0/igo/
The E ec s o In e es Ra e Inc eases
on Consume s’ In la ion Expec a ions:
The Roles o In o medness and Compliance
Edwa d S. Kno ek II
James Mi chell
Ma hieu Pedemon e
Taylo Shi o
WORKING PAPER No IDB-WP-1641
In e -
A
me ican De elopmen Bank
Depa men o Resea ch and Chie Economis
Sep embe 2024
* Fede al Rese e Bank o Cle eland
** In e -Ame ican De elopmen Bank
The E ec s o In e es Ra e Inc eases
on Consume s’ In la ion Expec a ions:
The Roles o In o medness and Compliance
Edwa d S. Kno ek II*
James Mi chell*
Ma hieu Pedemon e**
Taylo Shi o *
In e -
A
me ican De elopmen Bank
Depa men o Resea ch and Chie Economis
Sep embe 2024
Ca aloging-in-Publica ion da a p o ided by he
In e -Ame ican De elopmen Bank
Felipe He e a Lib a y
The e ec s o in e es a e inc eases on consume s’ in la ion expec a ions: he
oles o in o medness and compliance / Edwa d S. Kno ek II, James Mi chell,
Ma hieu Pedemon e, Taylo Shi o .
p. cm. — (IDB Wo king Pape Se ies; 1641)
Includes bibliog aphical e e ences.
1. Mone a y policy-Ma hema ical models-Uni ed S a es. 2. In la ion (Finance)-
Ma hema ical models-Uni ed S a es. 3. Communica ion policy-Ma hema ical
models-Uni ed S a es. 4. Su eys-Uni ed S a es. I. Kno ek, Edwa d S. II.
Mi chell, James. III. Pedemon e, Ma hieu. IV. Shi o , Taylo . V. In e -Ame ican
De elopmen Bank. Depa men o Resea ch and Chie Economis . VI. Se ies.
IDB-WP-1641
h p://www.iadb.o g
Copy igh © 2024 In e -Ame ican De elopmen Bank ("IDB"). This wo k is subjec o a C ea i e
Commons license CC BY 3.0 IGO (h ps://c ea i ecommons.o g/licenses/by/3.0/igo/legalcode). The
e ms and condi ions indica ed in he URL link mus be me and he espec i e ecogni ion mus be
g an ed o he IDB.
Fu he o sec ion 8 o he abo e license, any media ion ela ing o dispu es a ising unde such license
shall be conduc ed in acco dance wi h he WIPO Media ion Rules. Any dispu e ela ed o he use o
he wo ks o he IDB ha canno be se led amicably shall be submi ed o a bi a ion pu suan o he
Uni ed Na ions Commission on In e na ional T ade Law (UNCITRAL) ules. The use o he IDB's name
o any pu pose o he han o a ibu ion, and he use o IDB's logo shall be subjec o a sepa a e
w i en license ag eemen be ween he IDB and he use and is no au ho ized as pa o his license.
No e ha he URL link includes e ms and condi ions ha a e an in eg al pa o his license.
The opinions exp essed in his wo k a e hose o he au ho s and do no necessa ily e lec he iews o
he In e -Ame ican De elopmen Bank, i s Boa d o Di ec o s, o he coun ies hey ep esen .
Abs ac ∗
We s udy how mone a y policy communica ions associa ed wi h inc easing he
ede al unds a e causally a ec consume s’ in la ion expec a ions in eal ime. In
a la ge-scale, mul i-wa e andomized con olled ial (RCT), we ind weak
e idence ha communica ing hese policy changes lowe s consume s’ medium-
e m in la ion expec a ions on a e age. Howe e , in o ma ion di e s
sys ema ically ac oss demog aphic g oups in e ms o ex an e in o medness abou
mone a y policy and ex pos compliance wi h he in o ma ion ea men . Mone a y
policy communica ions ha e a much s onge e ec on he subse o consume s
who had no p e iously hea d news abou mone a y policy and who ake su icien
ime o ead he ea men . Ou indings show ha , in an in la iona y en i onmen ,
hese consume s expec ha aising in e es a es will lowe in la ion. Mo e
gene ally, ou esul s emphasize he impo ance o measu ing bo h esponden s’
in o ma ion se s and hei compliance wi h ea men when using RCTs in empi ical
mac oeconomics o be e unde s and he eal-wo ld implica ions o mone a y
policy communica ions.
JEL classi ica ions: E31, E52, E58
Keywo ds: Expec a ions o ma ion, Policy communica ion, Mone a y policy,
In la ion, Su eys
∗ Kno ek: Fede al Rese e Bank o Cle eland, Email: Edwa d.kno ek@cle . b.o g; Mi chell: Fede al Rese e Bank
o Cle eland, Email: jam[email p o ec ed] g; Pedemon e: In e -Ame ican De elopmen Bank, Email:
ma hieu.pedemon [email protected]; Shi o : Fede al Rese e Bank o Cle eland, Email: aylo .shi o @cle . b.o g.
We hank Oli Coibion, F ancesco D’Acun o, Rica do Reis, Michael Webe , and Mi ko Wiede hol o help ul
commen s, as well as con e ence and semina pa icipan s a he Fede al Rese e Bank o Cle eland, he Ox o d
Communica ion Wo kshop, he Bank o Canada, he Cle eland Fed’s Cen al Bank Communica ions: Theo y and
P ac ice con e ence, he Second Pa is Con e ence on he Mac oeconomics o Expec a ions, ISF 2024, Venice Summe
Ins i u e, and he CEBRA Annual Mee ing 2024. The RCT was egis e ed as AER RCT ID AEARCTR-0009172
(Kno ek 2022). The iews exp essed he e a e solely hose o he au ho s and do no necessa ily e lec he iews o
he Fede al Rese e Bank o Cle eland o he Boa d o Go e no s o he Fede al Rese e Sys em.
2
1. In oduc ion
Mone a y policymake s on he Fede al Open Ma ke Commi ee (FOMC) aised he ede al unds
a e a a ela i ely apid pace du ing 2022 ia a sequence o a e inc eases ha began on Ma ch 16,
2022. The p ima y eason o his apid igh ening o policy was moun ing conce n ha in la ion
was p o ing long-li ed, wi h he isk o in la ion expec a ions becoming unancho ed. This led
FOMC pa icipan s o alk openly and in ad ance o FOMC mee ings abou he pi o o a igh e
mone a y policy s ance and he need o highe in e es a es. While hese communica ions we e
e lec ed in ising ma ke -based expec a ions o he u u e ede al unds a e, i is no clea whe he
he consume s comp ising he “gene al public” we e paying a en ion o , i hey we e, how hey
we e eac ing o hese communica ions.1
This pape es ima es he causal e ec s o communica ing in e es a e inc eases on
consume s’ in la ion expec a ions using i e wa es o a andomized con olled ial (RCT)
conduc ed ia an online su ey. We ocus on communica ing o consume s in e y simple e ms
he ede al unds a e inc eases o 2022 and assessing he impac on hei in la ion expec a ions.
We do so by conduc ing in eal ime a specially designed se o RCTs immedia ely ollowing he
Ma ch, May, June, July, and Sep embe 2022 FOMC mee ings. Each o hese mee ings esul ed in
inc eases o he ede al unds a e a ge . Ou RCT-based es ima es o he a e age causal e ec o
communica ing hese hikes on households’ in la ion expec a ions ha e a wide ange. We ind ha
p o iding consume s wi h in o ma ion abou he la es in e es a e hike educed hei expec a ions
o in la ion o e he nex i e yea s, on a e age, be ween 0.17 and 2.18 pe cen age poin s,
depending on he p ecise in o ma ion ansmi ed.
While hese ea men e ec s a e s a is ically signi ican on a e age, we aise wo
di icul ies abou hei in e p e a ion ha , as explained below, much o he wide RCT-based
li e a u e in empi ical mac oeconomics also aces. Fi s , ou RCTs ea consume s wi h
in o ma ion ha is al eady publicly a ailable and, mo eo e , hey do so in an en i onmen whe e
mone a y policy news was salien gi en high in la ion eadings. We p o ide no el e idence ha
consume s we e indeed mo e in o med abou mone a y policy in 2022 han hey had been in p io
yea s.2 As a esul , some ea ed indi iduals likely al eady had he ea men in hei ex an e
1 See h ps://www.cmeg oup.com/insigh s/economic- esea ch/2022/ ed- a e-hikes-expec a ions-and- eali y.h ml.
2 This inding ha consume s we e mo e awa e o mone a y policy in 2022 han in p io yea s complemen s o he
wo k ha has documen ed ising awa eness o and a en ion paid o in la ion in pa icula du ing his episode, such as
P äu i (2024) and B acha and Tang (2024).
3
in o ma ion se s, a phenomenon also desc ibed in Webe e al. (2024) and Mackowiak and
Wiede hol (2024). Second, while ou RCTs explici ly communica e in o ma ion o he ea men
g oups, i is ha d o know how much o his in o ma ion is eally ead and/o abso bed by
esponden s wi hin he su ey. We make he analogy o medicine and a physician p esc ibing a
pill o a pa ien : he pa ien is only uly ea ed a e swallowing he pill, no by ha ing been
p esc ibed o e en gi en he pill.3 Volun a y compliance wi h he ea men can only be con i med
ex pos .4
Me hodologically, we in oduce o empi ical mac oeconomic s udies ha use RCTs he
impo ance o ex an e in o medness and ex pos compliance. We p opose a no el, easy, and
accessible way o con ol o bo h he ex an e in o ma ion se o esponden s and how complian
hey a e wi h he in o ma ion ea men ex pos in a manne ha is easible in mos online su eys.
In ou applica ion, we cap u e in o medness by asking households whe he hey had ecen ly hea d
mone a y news, and we measu e olun a y compliance— a he han a emp ing o en o ce
compliance— h ough he ime ha esponden s choose o spend eading he in o ma ion ea men
in he online su ey. In his way, we p o e bially do no o ce pa ien s o ake he pill, bu we
obse e h ough a wo-way mi o whe he hey choose o ake he pill.
Using econome ic me hods amilia o mic oeconomis s when unde aking causal
in e ence wi h non-expe imen al da a, we ob ain complie a e age ea men e ec s by
upweigh ing (downweigh ing) con ol g oup esponden s based on hei p edic ed p obabili y o
complying (no complying) wi h he ea men and excluding ea men g oup non-complie s. A e
accoun ing o compliance wi h he ea men and con olling o consume s’ ex an e in o medness
in ou eg essions, we ind ha mone a y policy communica ions abou inc eases in in e es a es
ha e a s a is ically signi ican and economically meaning ul nega i e impac on medium- e m
in la ion expec a ions o he consume s who we e p e iously unin o med abou ecen policy
ac ions and complian wi h he ea men s. To he ex en ha he in o med indi iduals p o e bially
swallowed he eal-wo ld mone a y policy pill upon hea ing mone a y policy news p io o aking
ou su ey, hen hey oo may ha e educed hei in la ion expec a ions a ha ime. Howe e ,
3 In he 1999 mo ie “The Ma ix,” when Mo pheus gi es Neo he choice be ween aking he blue pill and s aying
bliss ully unawa e o aking he ed pill and joining he mo emen o unde mine he ma ix, he wa ches as Neo akes
he ed pill and washes i down wi h a glass o wa e .
4 An al e na i e in e p e a ion is ha he e is a key in ensi e ma gin o ea men s ha can a y ac oss esponden s
and may no be comple ely andom. Fus e e al. (2022) ind ha people wi h less unce ain p io belie s a e mo e
likely o spend ime eading he in o ma ion ea men .
4
RCTs a e unable o adequa ely cap u e hei beha io s, because he in o ma ion ea men s a e
al eady pa o hei in o ma ion se , as no ed in Webe e al. (2024).5
Ou empi ical esul s con as wi h he indings in And e e al. (2022), who epo ed s ong
disag eemen among consume s’ esponses o hypo he ical si ua ions in ol ing mone a y policy
communica ions based on su ey esponses om a low-in la ion pe iod in 2019. Ou indings,
coming om a high-in la ion pe iod when consume s may espond di e en ly o in o ma ion
shocks, sugges ha consume s had a be e unde s anding o he objec i es o mone a y policy
when gi en he igh in o ma ion— hey unde s ood he basic mechanism ha highe in e es a es
would educe in la ion. In a heo e ical a ional ina en ion model, Mackowiak and Wiede hol
(2024) show ha indi iduals ha e mo e incen i e o pay a en ion o he mac oeconomy and o
comply wi h in o ma ion ea men s when in la ion is high. We explici ly measu e bo h ma gins
and show empi ically ha he ea men e ec , consis en wi h hei model, is highe o he
unin o med and o hose consume s who pay a en ion du ing he su ey.
By obse ing a he han en o cing compliance, we can analyze consume s’ choice o
comply wi h he ea men . We documen sys ema ic demog aphic di e ences ac oss indi iduals
who a e bo h less in o med abou mone a y policy and mo e complian wi h he ea men , and we
show ha hese di e ences a e ele an when in e p e ing a e age and he e ogeneous e ec s o
in o ma ion ea men s in he mac oeconomics RCT li e a u e. Fo example, Coibion,
Go odnichenko, and Webe (2022) ind ha communica ion di ec ly by he FOMC is mo e
e ec i e in mo ing household expec a ions han indi ec communica ion ia he media. They also
ind conside able he e ogenei y ac oss esponden s, wi h emale esponden s’ in la ion
expec a ions eac ing mo e s ongly o mone a y policy in o ma ion ea men s. In ou RCT, we
ind ha women a e mo e likely o spend a longe ime eading he in o ma ion ea men s han
men and o hus be conside ed as ha ing complied wi h he ea men . This inding means ha he
gende di e en ial in Coibion, Go odnichenko, and Webe (2022) may simply e lec women
paying mo e a en ion o he RCT ea men han men. We ind suppo o his conjec u e in ou
5 A he same ime, we do no belie e ha an e en s udy a ound he FOMC mee ings in ou sample is app op ia e.
Inc eases in he ede al unds a e du ing ou sample we e eleg aphed in ad ance o he mee ings—subjec o some
unce ain y abou hei size—and hence i is unclea when he “e en s” anspi ed. Fu he mo e, we p o ide ample
e idence ha he e a e sys ema ic di e ences ac oss indi iduals who a e and who a e no in o med abou mone a y
policy. Thus, compa ing indi iduals who had no hea d abou mone a y policy be o e an FOMC mee ing and hose
who had hea d abou mone a y policy a e he mee ing likely con ounds a numbe o ac o s beyond he ac ual
in o ma ion p o ided by he mee ing—i.e., selec ion in o in o medness is no andom.
5
s udy: a e con olling o in o medness and compliance, he gende gap is no longe s a is ically
signi ican .
Ou analysis also allows us o conside issues ela ed o mone a y policy awa eness mo e
b oadly. Ou indings sugges ha cen al bank communica ions could be augmen ed o each
consume s who ypically do no hea much mone a y policy news. A he same ime, ou
compliance esul s show ha e en when po ions o he public a e p esen ed di ec ly wi h
mone a y policy in o ma ion, he e is no gua an ee ha he in o ma ion will be p ocessed. Ou
ocus on mone a y policy awa eness and compliance hus do e ails wi h he wo k o D’Acun o,
Fus e , and Webe (2021), who ind ha he salience o emale and mino i y ep esen a ion on he
FOMC a ec s how Fed in o ma ion in luences consume s’ expec a ions, pa icula ly o selec ed
demog aphic g oups, po en ially o e ing a pa hway o enhance eal-wo ld compliance o engage
wi h mone a y policy news.
Rela ed Li e a u e
RCTs ha e gained p ominence in empi ical mac oeconomics o unde s and expec a ions
o ma ion; e.g., see A man ie e al. (2016), Binde and Rod igue (2018), and Coibion,
Go odnichenko, and Webe (2022). Using RCTs, Haldane and McMahon (2018) and Bhola e al.
(2019) ind simple ela able communica ions by he cen al bank o be mo e e ec i e in
in luencing households’ expec a ions, hus mo i a ing ou ocus on in o ma ion ea men s ha a e
“sho and swee .” To es whe he he in o ma ion ea men has addi ional powe o a ec
expec a ions i he a ionale o he policy change is also communica ed, addi ional ea men
g oups in ou su ey a e gi en some na a i e o “ igne e” (see And e e al., 2022, a ound he a e
inc ease; e.g., by explaining ha he FOMC is aising a es o educe in la ion. Ou pape hus
e isi s he ques ion o whe he i is bes o communica e a ge s o ins umen s, bu i does so o
he Uni ed S a es, o ien ing he in o ma ion ea men s a ound he ac ual ede al unds a e
decisions made by he FOMC h ough 2022 a he han hypo he icals. In ou case, we p o ide
simple ea men s and compa e ha ea men o o he s ha p o ide addi ional in o ma ion.
Ou esul s build on a body o li e a u e ha s udies communica ion as a cen al policy o
le e o mode n cen al banking. Much o his wo k has ocused on whe he and how cen al banks
can communica e o consume s and i ms (e.g., Blinde e al., 2008, Blinde e al., 2024, and
Binde , 2017). E ec i e communica ion is impo an when cen al banks wan o shape he
12
analysis, omi ing hem has li le impac on ou esul s.8 We do no use demog aphic con ols in
any o ou eg ession models unless o he wise speci ied.
Ou sample is well- andomized and balanced, wi h only mino di e ences be ween he
demog aphic make-up o any o he ea men g oups and he con ol g oups (see Appendix Table
B2). A daily quo a is used o ensu e ha he app op ia e quan i y o esponses is collec ed e en i
esponden s d op ou o he su ey be o e comple ing i . Responden s who ail he su ey’s
ReCAPTCHA check o who a e o he wise lagged by Qual ics as likely bo s o spamme s do no
coun owa d he quo a, no a e hey included in ou da a se .
We d op o o he wise al e as ew esponses om ou sample as possible. The only
esponden s who a e ou igh emo ed om ou sample a e hose whose o al su ey comple ion
imes a e ei he oo sho o oo long. We d op esponden s wi h o al su ey comple ion imes in
he 1s pe cen ile (N=512 consume s) as well as hose who ook mo e han an hou o comple e he
su ey (N=220), since he quali y o esponses om hese esponden s is ypically poo , owing o
ushing h ough he su ey o simply o ge ing abou i . The median esponden ook 14.5 minu es
o comple e he su ey, and jus unde 90 pe cen comple ed i in less han 25 minu es. Fo he July
and Sep embe su ey wa es, a ime was enabled o cap u e he ime ha esponden s spen
eading he sc een wi h an in o ma ion ea men , i ele an . Impo an ly, esponden s choose how
long o spend on he in o ma ion ea men page in ou su ey; we do no o ce hem o spend a
speci ic amoun o ime on his page.
To deal wi h ou lie s, we winso ize esponses a he 2nd and 98 h pe cen iles o all poin
expec a ions and use Hube - obus eg essions. Twel e pe cen o ou sample epo ed ha hey
we e expec ing de la ion o e he nex yea ; among hese esponden s, 59 pe cen an icipa ed
de la ion be ween ze o and 10 pe cen .
The aw median (ac oss consume s and o e ime) esponse o he ques ion on p io ( ha
is, p e- ea men ) yea -ahead in la ion expec a ions was 8 pe cen , wi h he 25 h pe cen ile a 4
pe cen and 75 h pe cen ile a 20 pe cen . The median pos e io ( ha is, pos - ea men ) i e-yea
a e age in la ion expec a ion was 5 pe cen , wi h he 25 h pe cen ile a 2 pe cen and 75 h pe cen ile
a 15 pe cen . A e applying Hube weigh s, he median, 25 h pe cen ile, and 75 h pe cen ile p io
8 These weigh s a e no used in he p opensi y sco e weigh ed eg essions in Sec ion 5. Ra he han eweigh ing o
ma ch popula ion dis ibu ions, hese eg essions eweigh he con ol g oup o ma ch dis ibu ions wi hin he
ea men g oup complie s/non-complie s.
13
expec a ion a e 7 pe cen , 4 pe cen , and 11 pe cen , espec i ely, and 5 pe cen , 2 pe cen , and 10
pe cen o he pos e io .
3. T ea men E ec s o Di e en Communica ion Tools
To es ima e he a e age ea men e ec , 𝛽𝛽𝑗𝑗, o ea men s j=1,2,3,4, we un he ollowing
eg ession:
𝜋𝜋𝑖𝑖,𝑡𝑡
5𝑦𝑦− 𝜋𝜋𝑖𝑖,𝑡𝑡
1𝑦𝑦=𝛼𝛼+∑𝛽𝛽𝑗𝑗×𝐼𝐼�1 𝑖𝑖𝑖𝑖 𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑗𝑗= 1�
4
𝑗𝑗=1 +𝜀𝜀𝑖𝑖𝑡𝑡, (1)
whe e 𝜋𝜋𝑖𝑖,𝑡𝑡
5𝑦𝑦 is he pos e io , i e-yea in la ion expec a ion o indi idual 𝑖𝑖 in wa e 𝑡𝑡, 𝜋𝜋𝑖𝑖,𝑡𝑡
1𝑦𝑦 is he
p io , 12-mon h in la ion expec a ion, and 𝐼𝐼�1 𝑖𝑖𝑖𝑖 𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑗𝑗= 1� is a 0-1 dummy ha akes a
alue o 1 i esponden 𝑖𝑖 ecei ed ea men 𝑗𝑗. I ea men 𝑗𝑗 is e ec i e in changing he pos e io
in la ion expec a ion ela i e o he p io in la ion expec a ion, hen 𝛽𝛽𝑗𝑗 will be di e en om ze o,
implying ha he ea men induces a di e en esponse on a e age o indi iduals who ecei e
he ea men ela i e o he con ol g oup. No e ha 𝛽𝛽𝑗𝑗 measu es he a e age ea men e ec ;
changes o he dis ibu ion o esponses ha do no a ec he a e age will no be cap u ed in ou
eg ession. Consequen ly, a alue o 𝛽𝛽𝑗𝑗< 0 signi ies ha ea men 𝑗𝑗 lowe s esponden s’ in la ion
expec a ions ela i e o hei p io on a e age. To help il e ou lie esponses, we apply Hube
weigh s ob ained om a simila eg ession o i e-yea in la ion expec a ions on one-yea
expec a ions, ea men indica o s, and hei in e ac ions.
Table 1 shows he es ima ion esul s o each wa e sepa a ely, as well as a pooled e sion
ha includes a wa e ixed e ec . The ime ixed e ec s a e impo an , since hey con ol o he
common in o ma ion ha he ea ed and con ol g oups had a he ime. Thanks o he di e en
wa es, in ha speci ica ion we can con ol and alk abou a gene al e ec o e he ull-sample
pe iod.
Table 1 epo s a nega i e a e age ea men e ec o each ea men , implying ha he
a e age esponden eac s o he in o ma ion ea men by educing hei in la ion expec a ions.
This nega i e e ec is p esen e en when consume s only ecei e in o ma ion abou he new
ede al unds a e (T ea men 1). This sugges s ha consume s may ha e some unde s anding o
he mechanism behind mone a y policy ac ions. This is pa icula ly ele an o T ea men 1, as i
does no include in o ma ion abou ei he he policy objec i e o in la ion. The e ec o T ea men
2, howe e , seems o be smalle in magni ude. This could be explained by an “in o ma ion e ec ”
14
(see Nakamu a and S einsson, 2018) ha can con ound communica ions abou in e es a es: when
consume s a e old ha he FOMC wan s o educe in la ion wi h hei ac ions, he e is implici ly
an acknowledgmen o an in la iona y p oblem, which migh educe he size o he ea men
e ec .9
Table 1. Pos e io Minus P io on T ea men s
(1)
(2)
(3)
(4)
(5)
(6)
Ma ch
May
June
July
Sep
Pooled
T ea men 1
-1.93***
-0.17
-0.38*
-1.18***
-0.36
-1.57***
(0.33)
(0.19)
(0.21)
(0.28)
(0.32)
(0.13)
T ea men 2
-0.24
-0.18
-0.45**
-0.84***
-1.59***
-0.35***
(0.20)
(0.18)
(0.21)
(0.29)
(0.36)
(0.09)
T ea men 3
-2.18***
-2.14***
(0.32)
(0.31)
T ea men 4
-0.72***
-0.66***
(0.21)
(0.19)
Placebo
1.09***
0.36*
(0.27)
(0.18)
Obse a ions
7879
5204
4899
5994
5872
29833
No es: Columns 1-5 con ain no con ols o ixed e ec s. Column 6 shows a pooled eg ession wi h ea men
pe iod ixed e ec s.
*
p < 0.10,
**
p < 0.05,
***
p < 0.01
The e a e some e ec s in Table 1 ha a e ha d o a ionalize. Fo example, he placebo
seems o inc ease in la ion expec a ions, depending on he speci ica ion. We e-e alua e hese
esul s below, once we ha e accoun ed o consume s’ ex an e in o medness abou mone a y
policy and hei ex pos compliance wi h he in o ma ion ea men .
4. In o med and Unin o med Responden s: Expanding he Reach o Mone a y
Policy
Table 1 showed ha he in o ma ion ea men was e ec i e in educing he in la ion expec a ions
o he a e age esponden . Bu since his in o ma ion ea men is con eying public in o ma ion
a he ime o he expe imen , in p inciple ou ea men should only a ec esponden s who a e
unin o med abou ecen mone a y policy decisions. In o med esponden s should al eady know
he in o ma ion communica ed in he ea men and hence should be una ec ed by i . In his
9 In Appendix C, we complemen Table 1 by showing ha he ea men e ec s a e la ge o hose consume s wi h
highe p io expec a ions o in la ion.
15
sec ion, we explo e whe he he e a e di e ences be ween he esponden s who ecei e he
ea men and a e likely unin o med ( he “local a e age ea men e ec ”) ela i e o he e ec on
all o he ea ed ( he a e age ea men e ec ).
We s a by iden i ying he unin o med g oup and explo ing i s demog aphic, beha io al,
and socio-economic cha ac e is ics. We do so by exploi ing he ac ha ou su ey includes a p e-
ea men ques ion ha asks esponden s i mo gage a es had changed ecen ly (and, i so, how)
and whe he hey had hea d news abou mone a y policy. As ou ea men s p o ide in o ma ion
abou mone a y policy decisions and changes in in e es a es, hese ques ions help iden i y which
esponden s we e likely o ha e al eady known he in o ma ion in he ea men . We ind ha 82.1
pe cen o esponden s who had hea d news abou mone a y policy we e also awa e ha in e es
a es had isen ecen ly, while only 45.7 pe cen o esponden s who had no hea d mone a y policy
news we e awa e o such an inc ease, and only 29.7 pe cen co ec ly desc ibed he change. Since
he ques ion on mone a y policy news is indeed co ela ed wi h in o medness abou in e es a es,
we use as ou indica o o in o medness whe he esponden s had hea d news.10 Figu e 1 shows
he e olu ion o he answe s o his ques ion o e ime.
Figu e 1 e eals ha he sha e o esponden s who indica e ha hey ha e hea d news abou
mone a y policy has been ising since 2020. While only be ween 25 and 30 pe cen o esponden s
hea d news abou mone a y policy in Oc obe 2020, a ound 45 pe cen o esponden s had hea d
news o mone a y policy in Oc obe 2022. This is consis en wi h he ise in in la ion and he
consequen inc eased public discussion o mone a y policy h ough 2022. In addi ion, we also see
om Figu e 1 ha esponden s a e mo e likely o ha e hea d news abou mone a y policy
immedia ely a e FOMC mee ings, as ep esen ed by he e ical dashed lines in Figu e 1.11
10 Using he ques ion on in e es a es also in oduces addi ional complica ions. Fo example, since he possible ini ial
esponses we e “No,” “No Su e,” and “Yes,” we would need o conside he cases o uly unin o med (i.e., “No”
esponden s) and li le-in o med (i.e., “No su e” esponden s), unde he enuous assump ion ha hese labels a e
accu a e. Using hese as al e na i e indica o s o in o medness does no signi ican ly change ou esul s, howe e .
11 Using a s a i ied andom sample o he US public wo days be o e and wo days a e he FOMC p ess con e ence,
Lamla and Vinog ado (2019) also ind ha mone a y policy announcemen s lead o an inc ease in he p opo ion o
people who ha e hea d mone a y policy news.
16
Figu e 1. Pe cen age o Responden s Hea ing News abou Mone a y Policy
We should expec smalle ea men e ec s o esponden s who ha e al eady hea d news
abou mone a y policy. To es his hypo hesis, we e- un he eg essions in Table 1 bu only on
ha subse o esponden s who epo ed ha hey had no hea d news abou mone a y policy. No e
ha by design ( andom) assignmen o a ea men g oup is no co ela ed wi h ha ing hea d news
abou mone a y policy (see Appendix Table B2).12
Table 2 con i ms ha he es ima ed ea men e ec s o esponden s who ha e no hea d
news a e indeed o en la ge han hose epo ed in Table 1, likely because o hese esponden s
he in o ma ion con ained in he ea men is mo e in o ma i e han i is o hose who ha e al eady
hea d some mone a y policy news. In pa icula , in July and Sep embe we obse e s onge
ea men e ec s in Table 2 han in Table 1.
12 Appendix Table E2 p esen s es ima es showing how he p obabili y ha esponden s had hea d news abou mone a y
policy a ies wi h demog aphic ac o s and as a unc ion o he ime since he mos ecen FOMC mee ing.
17
Table 2. Pos e io Minus P io on T ea men s, Hea d News = No
(1)
(2)
(3)
(4)
(5)
(6)
Ma ch
May
June
July
Sep
Pooled
T ea men 1
-1.24**
-0.20
-1.34*
-2.03***
-0.93**
-1.34***
(0.62)
(0.59)
(0.70)
(0.45)
(0.47)
(0.26)
T ea men 2
0.26
-0.01
0.73
-1.30***
-2.81***
-0.70***
(0.46)
(0.56)
(0.58)
(0.43)
(0.61)
(0.23)
T ea men 3
-1.26**
-1.40***
(0.55)
(0.50)
T ea men 4
-0.75
-0.91**
(0.49)
(0.44)
Placebo
-0.61
0.01
(0.45)
(0.41)
Obse a ions
4508
3075
2669
3503
3337
17091
No es: Columns 1-5 con ain no con ols o ixed e ec s. Column 6 shows a pooled eg ession wi h
ea men pe iod ixed e ec s. * p < 0.10, ** p < 0.05, *** p < 0.01
This di e en ial e ec speaks o a gene al limi a ion o his ype o RCT. The ea men
e ec o communica ing in o ma ion depends on how “in o ma i e” he in o ma ion p o ided
ac ually is. Responden s who a e al eady ully in o med ha e no need o upda e hei p io s pos -
ea men . In he nex sec ion, we p opose a way o dis inguishing be ween he in o ma ional
con en o he (speci ic) ea men using a measu e o whe he he esponden complied wi h he
ea men .
5. Who Reads he T ea men ? Measu ing Compliance
Assigning a esponden o a ea men g oup does no gua an ee ha he esponden complies
wi h—and hus ac ually ecei es— he ull ex en o he in o ma ion con ained wi hin he
ea men . A pa allel can be made wi h medical RCTs, in which pa ien s may be andomly assigned
he ea men , bu hey may no comply, o example, by no swallowing he pill p esc ibed by he
physician. In he con ex o ou su ey o consume s, he e could be many easons o non-
compliance. Consume s migh be ina en i e wi hin he su ey i sel : hey could be dis ac ed, o
example, and con inue o he nex ques ion wi hou ha ing p ocessed he ea men . Al e na i ely,
hey migh no be willing o pay a en ion o he ea men a all, and hey simply skip o he nex
ques ion as as as possible wi hou diges ing he in o ma ion. While demog aphic cha ac e is ics
and ques ion-based assessmen s o eading and/o nume ical li e acy can help con ol o a ying
18
le els o a en i eness and unde s anding o he ea men , hey canno assess who eads—and
hence complies wi h— he ea men and who does no . Because non-compliance may be sel -
selec ed, i is “nonigno able,” since i unde mines he andom alloca ion in o he ea men g oup
equi ed o unbiased es ima ion o he local a e age ea men e ec ( he complie a e age causal
e ec ); see Imbens and Ang is (1994) and Ang is , Imbens, and Rubin (1996). The e o e, he
es ima es in Tables 1 and 2 o e unbiased es ima es o he e ec o assignmen , he so-called
“in en ion- o- ea ,” no o he ea men i sel . They e lec he e icacy o bo h he ea men and
he compliance.
In his sec ion, we p opose a no el ool o measu e whe he esponden s ead ( ha is,
comply wi h) he ea men , which we hen use o es ima e he a e age ea men e ec ee om
any con ounding e ec s o non-compliance. This is acili a ed by he ac ha in he July and
Sep embe wa es we measu ed how much ime esponden s chose o spend on he ea men
page.13 We use his measu e as a p oxy o whe he esponden s ead and p ocessed he ea men ;
we hen sepa a e he sample in o a en i e and ina en i e esponden s. Because he ea men s a y
in leng h and con en , and ac oss wa es, sepa a e ules a e calcula ed o each ea men in each
wa e. We use a ule whe eby a esponden is conside ed o ha e “ ead he ea men ” i hey ook
a leas hal o he a e age amoun o ime spen by esponden s assigned o hei ea men in hei
wa e o ead he ea men . This cu o is 4.1 and 5.4 seconds o T ea men s 1 and 2 in July,
espec i ely, and 4.3, 5.9, and 2.5 seconds o T ea men 1, T ea men 2, and he placebo in
Sep embe , espec i ely.14 Table 3 shows how demog aphic and o he cha ac e is ics p edic
a en i eness o he ea men .
While we canno know om eading ime alone whe he a esponden uly ead, p ocessed,
and co ec ly unde s ood he in o ma ion in he ea men , we can make easonable assump ions
ha hose wi h ea men eading imes below hese cu o s a e no likely o ha e done so. E en
hough ou ea men s p o ide succinc , “ wee -s yle” snippe s o in o ma ion, i is un easonable
o imagine ha a esponden could ead he en i e y o he ea men in ewe han 4 o 5 seconds.
13 This ype o in o ma ion is easy o implemen in online su eys. Fus e e al. (2022) use a simila coun e o ack
he ele ance o in o ma ion. They measu e ime spen epo ing he pos e io . They ind a posi i e ela ionship
be ween unce ain p io s and he ime spen epo ing he pos e io .
14 Ou esul s a e no sensi i e o using di e en cu o s. Using an al e na i e s anda d o he 25 h, 50 h, o 75 h pe cen ile
eading ime p oduces simila esul s, wi h s onge ea men e ec s associa ed wi h longe eading ime
equi emen s. In gene al, ou “hal - he-a e age” ule equi es a ea men eading ime sligh ly below he median. Fo
a plo o he empi ical dis ibu ion o eading imes, see Appendix Figu e D1. Fo ea men e ec s by he eading
ime pe cen ile used as he cu o poin , see Appendix Figu e D3.
19
Ou measu e o compliance, he e o e, is a conse a i e one. The e may e y well be esponden s
we conside complian who in ac we e no complian , bu he e could only be e y ew
esponden s who we could be inco ec ly conside ing non-complian , gi en ou eading cu o
imes.
Table 3. Likelihood o Reading he T ea men
Full Sample
No News Only
(1)
(2)
(3)
(4)
Logi Odds
Ra io
OLS
Logi Odds
Ra io
OLS
Male
0.78***
(0.05)
-0.05***
(0.01)
0.80***
(0.07)
-0.05***
(0.02)
Nonwhi e
0.67***
(0.05)
-0.08***
(0.02)
0.64***
(0.07)
-0.09***
(0.02)
Hispanic
0.63***
(0.07)
-0.09***
(0.02)
0.64***
(0.09)
-0.09***
(0.03)
P ima y Shoppe
0.89
(0.10)
-0.02
(0.02)
1.09
(0.14)
0.02
(0.02)
Nume ical Li e acy
1.83***
(0.16)
0.11***
(0.02)
1.50***
(0.17)
0.08***
(0.02)
Hea d News
0.94
(0.06)
-0.01
(0.01)
Age:
36-50
1.89***
(0.17)
0.13***
(0.02)
2.17***
(0.24)
0.16***
(0.02)
51-65
4.71***
(0.45)
0.35***
(0.02)
4.48***
(0.53)
0.34***
(0.02)
66+
13.38***
(1.56)
0.53***
(0.02)
10.18***
(1.59)
0.49***
(0.03)
Income:
$35,000-$49,999
1.16
(0.11)
0.03
(0.02)
1.23*
(0.13)
0.04*
(0.02)
$50,000-$99,999
0.92
(0.07)
-0.02
(0.02)
0.94
(0.09)
-0.01
(0.02)
$100,000 o mo e
0.77**
(0.08)
-0.05**
(0.02)
0.65***
(0.09)
-0.09***
(0.03)
Educa ion:
Some College
1.63***
(0.13)
0.10***
(0.02)
1.46***
(0.14)
0.08***
(0.02)
Bachelo 's Deg ee
1.57***
(0.14)
0.09***
(0.02)
1.73***
(0.20)
0.11***
(0.02)
Ad anced Deg ee
1.47***
(0.17)
0.07***
(0.02)
1.83***
(0.30)
0.12***
(0.03)
Poli ical Pa y:
Democ a
0.89
(0.07)
-0.02
(0.01)
0.86
(0.08)
-0.03
(0.02)
Republican
1.08
(0.09)
0.02
(0.02)
1.05
(0.11)
0.01
(0.02)
Cons an
0.63**
(0.11
0.39***
(0.04)
0.56***
(0.12)
0.37***
(0.04)
Obse a ions
9421
9421
5347
5347
No es: S anda d e o s a e epo ed in pa en heses. Columns 1 and 3 epo esul s om a logi model p edic ing
compliance as a unc ion o he lis ed a iables as odds a ios, while columns 2 and 4 epo esul s om OLS
eg essions o an o he wise iden ical model.
*
p < 0.10,
**
p < 0.05,
***
p < 0.01
We see om Table 3 ha women end o pay mo e a en ion o he ea men , e en
condi ional on no ha ing al eady hea d news. In addi ion, whi e, non-Hispanic esponden s a e
mo e likely o ead he ea men . Olde and mo e educa ed esponden s end o pay mo e a en ion,
bu income plays a nega i e ole, i any. Poli ical a ilia ion does no seem o a ec compliance.
20
In o al, we can explain a good p opo ion o compliance ha does no come om idiosync a ic
cha ac e is ics ( o example, how i ed he esponden s a e). This esul is use ul because i will
allow us o cha ac e ize po en ial candida es who a e likely o be a en i e in he con ol g oup,
and he e o e o see i he e ec o he ea men is s onge o consume s who pay a en ion.
We ind a close ela ionship be ween he p edic ed p obabili ies o compliance and he
Hube weigh s used in Tables 1 and 2: esponden s wi h Hube weigh s close o 1 a e p edic ed o
be much mo e likely o ead he ea men han hose downweigh ed in he Hube - obus
eg essions (see Appendix Figu e D2). In o he wo ds, hose who a e p edic ed o be mo e likely
o comply wi h hei assigned ea men a e a less likely o p o ide ou lie esponses, and ice-
e sa. Accoun ing o compliance he e o e has an added bene i in ha doing so downweigh s
ou lie esponses on he basis o he esponden s’ beha io ins ead o using measu es de i ed om
he s a is ical p ope ies o all (o a subse o ) esponses, as in he case o Hube weigh s.
While ha ing hea d news does no p edic assignmen o a ea men g oup, eading he
ea men does; esponden s could only pass he eading ime cu o i hey had a ea men o ead
in he i s place. This p oblem is compounded by he p edic abili y o compliance demons a ed
in Table 3: selec ion in o compliance o non-compliance is p edic able by esponden
demog aphics. The e o e, simply excluding non-complie s om he sample and e unning he
eg essions in Tables 1 and 2 would no longe lea e us wi h a andomly assigned ea men g oup.
This implies ha he a e age ea men e ec es ima es p o ided in Tables 1 and 2 a e likely o
unde s a e he local a e age ea men e ec ( o complie s), because hose es ima es do no
accoun o he pa ial endogenei y o ea men selec ion; ha is, esponden s mus be andomly
assigned a ea men g oup o be ea ed, bu hey may o may no comply wi h being ea ed in a
non- andom ashion. This unde es ima ion is unde s ood by no ing (see Imbens and Ang is , 1994,
and Ang is , Imbens, and Rubin, 1996) ha , unde he “exclusion es ic ion” ha ea men does
no a ec compliance, he local a e age ea men e ec is he a io o he es ima ed in en - o- ea
e ec (as shown in Tables 1 and 2) and he es ima ed p opo ion o complie s (as modeled in Table
3).
Gi en he e idence om Table 3 ha we ha e co a ia es ha explain compliance, we use
hese o calcula e p opensi y sco es ha , in u n, a e used o e-es ima e he ea men e ec .
Following Jo and S ua (2009), we use a wo-s ep p ocess o es ima e complie and non-complie
ea men e ec s. In he i s s ep, simila o Follmann (2000), we es ima e p opensi y sco es (𝑝𝑝)
21
o esponden s in he con ol g oup o a (gi en) ea men pe iod by i ing a logi model o
compliance wi hin he ea men g oup using he co a ia es seen in Table 3. In he case o
addi ional sample es ic ions ( ha is, excluding esponden s who hea d news abou mone a y
policy), he logi model is es ima ed on a subsample wi h he same es ic ions, and i ed alues
a e only calcula ed o esponden s in he con ol g oup belonging o he same g oup ( ha is, who
also did no hea news abou mone a y policy). Because o andomiza ion, he co a ia es used in
he ea men g oup should also explain compliance in he con ol g oup, o whom we canno
measu e compliance di ec ly, o cou se, as hey did no ecei e a ea men and hus we e no
imed.
In he second s ep, we e u n o he eg ession model in Tables 1 and 2, bu now use he
es ima ed p opensi y sco es o eweigh esponden s. To es ima e he a e age causal e ec among
complie s (CACE), esponden s in he con ol g oup a e assigned weigh s 𝑝𝑝𝑖𝑖/(1 − 𝑝𝑝𝑖𝑖), while
ea men g oup complie s and non-complie s a e assigned weigh s equal o one and ze o,
espec i ely ( hus equi alen ly excluding ea men g oup non-complie s). To es ima e he a e age
causal e ec among non-complie s (NACE), esponden s in he con ol g oup a e assigned weigh s
(1 − 𝑝𝑝𝑖𝑖)/𝑝𝑝𝑖𝑖, while ea men g oup complie s and non-complie s a e assigned weigh s o ze o and
one, espec i ely ( he in e se o he CACE weigh ing egime).15 In ui i ely, hese eg essions
es ima e complie (non-complie ) ea men e ec s by gi ing he mos likely and leas likely
complie s (non-complie s) wi hin he con ol g oup he la ges and smalles weigh s, espec i ely,
e ec i ely eweigh ing he con ol g oup as a whole o ma ch he cha ac e is ics o he ea men
g oup complie s (non-complie s).
Table 4 epo s CACE and NACE es ima es ob ained ia he wo-s ep p ocedu e ou lined
abo e, o he July and Sep embe wa es and b oken down by p io news exposu e. To accoun
o he es ima ion unce ain y om using es ima ed p opensi y sco es in he second-s age weigh ed
eg ession, we boo s ap each s ep o he wo-s ep p ocess. Table 4 epo s he means and s anda d
de ia ions o he dis ibu ion o each coe icien ’s d aws, wi h p-sco es calcula ed using hese
means and s anda d de ia ions.
15 The eason ha ea men g oup complie s and non-complie s ge weigh s o 1 and 0, espec i ely (and ice- e sa
when es ima ing he NACE), a he han p edic ed sco es, is ha we know wi h ce ain y whe he hey passed he
eading cu o ime o no . Using p edic ed sco es o e he obse ed compliance would disca d his in o ma ion.
28
While ou ea men s and analysis ocus on he communica ion o mone a y policy
decisions, consume s may be mo e awa e o o he , longe - e m in e es a es—which po en ially
inco po a e o wa d-looking expec a ions o mone a y policy—because hese a es a e mo e
ele an o he eal-wo ld mone a y ansmission mechanism o bo owe s and sa e s. In
Appendix G, we e-es ima e he ea men e ec s in Table 4 while con olling o wo addi ional
measu es o b oade awa eness abou in e es a es ha ou su ey asks abou p e- ea men :
whe he he esponden epo ed ha he in e es a es ha people pay o bo ow money in gene al
had changed ecen ly; and whe he he esponden indica ed ha bo owing a es in gene al had
gone up by mo e han 1 pe cen age poin ecen ly.20 As in Table 4, we con inue o ind nega i e
and s a is ically signi ican ea men e ec s o all esponden s who complied wi h he ea men
and o hose esponden s who complied and we e unin o med abou mone a y policy. We ind no
e idence ha p io awa eness abou bo owing a es a enua es he ea men e ec s, as seen by
he s a is ically insigni ican in e ac ion e ms. Thus, ou esul s a e no dependen on p o iding
in o ma ion o consume s who we e comple ely ina en i e o he economic en i onmen ; he
mone a y policy ea men s also mo ed he in la ion expec a ions o hose who al eady had some
sense o wha was happening o in e es a es. These esul s p o ide sugges i e e idence ha i is
he combina ion o communica ing he ac o ( ha is, he FOMC) and he ac ion ( ha is, he
decision o aise he policy a e) in ou ea men s ha has deli e ed he la ge nega i e impac on
in la ion expec a ions, and no a change in in e es a es alone. Fu u e wo k can e ine he ole o
hese wo ma gins u he .
Ou esul s highligh he impo ance o con olling o bo h in o medness and compliance
so ha RCTs ha e a highe deg ee o ex e nal alidi y and can be e help economis s unde s and
he eal-wo ld e ec s o mone a y policy communica ions. In o medness is no he same o
e e ybody (see Appendix Table E2), wi h some agen s acc uing in o ma ion abou policy ahead
o o he s in he “ eal-wo ld” a he han in he “labo a o y” o he RCT. In o med consume s
upda e hei in la ion expec a ions ou side o he RCT. This means ha he small and s a is ically
insigni ican ea men s e ec s we ind o hem (Table 4, Columns (2) and (5)) do no mean ha
he mone a y policy communica ion has been ine ec i e, jus ha his g oup o consume s may
20 Appendix G documen s ha he e is a high co ela ion be ween in o medness abou mone a y policy and awa eness
o inc eases in bo owing a es mo e gene ally, bu no a pe ec co ela ion be ween he wo di e en concep s, which
o ms he basis o hese eg ession esul s.
29
ha e al eady upda ed hei expec a ions. The unin o med do ha e an unde s anding o he main
mechanism o mone a y policy, so new, a ge ed communica ions o each his g oup in he eal-
wo ld should also be expec ed o weaken hei esponsi eness o ea men s wi hin he RCT.
7. Conclusion
This pape epo s and analyzes esul s om a specially designed mul i-wa e RCT o es whe he
and how communica ions abou ac ual inc eases in he ede al unds a e in 2022 causally a ec ed
consume s’ in la ion expec a ions. We ind ha simple communica ions abou inc eases in he
ede al unds a e educed consume s’ medium- e m in la ion expec a ions, bu mos no ably only
o hose who we e p e iously unawa e o bu willing o pay a en ion o ( ha is, in ou language,
“complian ” wi h) he in o ma ion communica ed in he RCT. Ou esul s hus highligh he
impo ance o measu ing he ex an e in o medness and he ex pos compliance o esponden s in
applied mac oeconomics RCTs.
Ou esul s he e o e sugges ha he FOMC’s policy ac ions o 2022 likely helped o
di ec ly ( e)ancho medium- e m in la ion expec a ions and con ibu e o he disin la ion p ocess
o some consume s. In u n, he ac ha eal- ime and eal-li e mone a y policy communica ions
o he policy ac ion alone (e en absen any in o ma ion on i s in en ) a e ound o lowe longe -
e m in la ion expec a ions sugges s ha consume s did ha e some common unde s anding o he
mone a y policy ansmission mechanism and o he FOMC’s in en ion ha he a e hikes should
d i e in la ion down.
Ou inding ha communica ing he mone a y policy changes o 2022 had small e ec s on
in la ion expec a ions on a e age, bu la ge e ec s on he p e iously unin o med and complian ,
ein o ces he g owing consensus in he li e a u e ( o example, Coibion and Go odnichenko, 2015,
and And e e al., 2022) ha he e is conside able he e ogenei y ac oss consume s in e ms o how
hey o m and upda e hei in la ion expec a ions. Du ing he high in la ion o 2022, we ind ha
no e e yone was paying a en ion o mone a y policy no was e e yone equally a en i e o he
in o ma ion ea men adminis e ed in ou i e RCTs, e en when p esen ed wi h e y sho , wee -
like mone a y policy communica ions designed o be easy o ead. By iden i ying g oups o
consume s who end o epo being less in o med abou mone a y policy news and o be mo e
likely o pay a en ion o news when i is sha ed wi h hem, and by con olling o hose g oups in
ou eg essions, we ind e idence ha he e is scope o inc ease he impac o mone a y policy
30
communica ions by a ge ing speci ic g oups o he gene al public, no ably women. Mo e
gene ally, ou esul s sugges ha i is impo an when in e p e ing he he e ogeneous ea men
e ec s commonly ound in RCTs in mac oeconomics o unpack bo h he complian om he non-
complian and hose o whom he in o ma ional ea men is news om hose o whom he
in o ma ion is al eady known. O he wise, as we illus a e, i is possible o misclassi y he d i e s
o he he e ogenei y and misdiagnose he bes ways o inc easing he e ec i eness o mone a y
policy communica ions.
31
Re e ences
And e, Pe e , Ca lo Pizzinelli, Ch is ophe Ro h, and Johannes Wohl a . 2022.
“Subjec i e Models o he Mac oeconomy: E idence om Expe s and Rep esen a i e
Samples.” Re iew o Economic S udies 89 (6): 2958-91.
h ps://doi.o g/10.1093/ es ud/ dac008.
Ang is , Joshua D., Guido W. Imbens, and Donald B. Rubin. 1996. “Iden i ica ion o Causal
E ec s Using Ins umen al Va iables.” Jou nal o he Ame ican S a is ical Associa ion 91:
444–55. h ps://doi.o g/10.1080/01621459.1996.10476902.
A man ie , Oli ie , Sco Nelson, Gio gio Topa, Wilbe an de Klaauw, and Basi Za a . 2016.
“The P ice Is Righ : Upda ing In la ion Expec a ions in a Randomized P ice In o ma ion
Expe imen .” Re iew o Economics and S a is ics 98 (3): 503–23.
h ps://doi.o g/10.1162/REST_a_00499.
Bhola , Da id, Nida B ough on, Janna Te Mee , and E yk Walczak. 2019. “Enhancing Cen al
Bank Communica ions Using Simple and Rela able In o ma ion.” Jou nal o Mone a y
Economics 108: 1-15. h ps://doi.o g/10.1016/j.jmoneco.2019.08.007.
Binde , Ca ola. 2017. “Fed Speak on Main S ee : Cen al Bank Communica ion and Household
Expec a ions.” Jou nal o Mac oeconomics 52 (C): 238-51.
h ps://doi.o g/10.1016/j.jmac o.2017.05.003.
Binde , Ca ola and Alex Rod igue. 2018. “Household In o medness and Long-Run In la ion
Expec a ions: Expe imen al E idence.” Sou he n Economic Jou nal 85: 580-98.
h ps://doi.o g/10.1002/soej.12306.
Blinde , Alan. S., Michael Eh mann, Ma cel F a zsche , Jakob De Haan, and Da id-Jan Jansen.
2008. “Cen al Bank Communica ion and Mone a y Policy: A Su ey o Theo y and
E idence.” Jou nal o Economic Li e a u e 46 (4): 910–45.
h ps://doi.o g/10.1257/jel.46.4.910.
Blinde , Alan S., Michael Eh mann, Jakob de Haan, and Da id-Jan Jansen. 2024. “Cen al Bank
Communica ion wi h he Gene al Public: P omise o False Hope?” Jou nal o Economic
Li e a u e 62(2): 425-57. h ps://doi.o g/10.1257/jel.20231683.
B acha, Ana and Jenny Tang. 2024. “In la ion Le els and (In)A en ion.” Re iew o Economic
S udies. h ps:// doi.o g/10.1093/ es ud/ dae063.
32
Coibion, Oli ie , and Yu iy Go odnichenko. 2015. “In o ma ion Rigidi y and he Expec a ions
Fo ma ion P ocess: A Simple F amewo k and New Fac s.” Ame ican Economic Re iew
105 (8): 2644-78. h ps://doi.o g/10.1257/ae .20110306.
Coibion, Oli ie , Yu iy Go odnichenko, and Sa en Kuma . 2018. “How Do Fi ms Fo m Thei
Expec a ions? New Su ey E idence.” Ame ican Economic Re iew 108 (9): 2671–2713.
h ps://doi.o g/10.1257/ae .20151299.
Coibion, Oli ie , Yu iy Go odnichenko, Sa en Kuma , and Ma hieu Pedemon e. 2020. “In la ion
expec a ions as a policy ool?” Jou nal o In e na ional Economics, 124: 103297.
h ps://doi.o g/10.1016/j.jin eco.2020.103297.
Coibion, Oli ie , Dimi is Geo ga akos, Yu iy Go odnichenko, and Michael Webe . 2023a.
“Fo wa d Guidance and Household Expec a ions”. Jou nal o he Eu opean Economic
Associa ion. 21 (5): 2131-71. h ps://doi.o g/10.1093/jeea/j ad003.
Coibion, Oli ie , Yu iy Go odnichenko, Edwa d S. Kno ek II, and Raphael Schoenle. 2023b.
“A e age In la ion Ta ge ing and Household Expec a ions.” Jou nal o Poli ical Economy
Mac oeconomics 1 (2): 403-46. h ps://doi.o g/10.1086/722962.
Coibion, Oli ie , Yu iy Go odnichenko, and Michael Webe . 2022. “Mone a y Policy
Communica ions and Thei E ec s on Household In la ion Expec a ions.” Jou nal o
Poli ical Economy, 130 (6): 1537-84. h ps://doi.o g/10.1086/718982.
D’Acun o, F ancesco, Daniel Hoang, Ma i a Palo ii a, and Michael Webe . 2020. “E ec i e
Policy Communica ion: Ta ge s e sus Ins umen s.” Wo king Pape 2020-148. Uni e si y
o Chicago Becke F iedman Ins i u e o Economics.
h ps://doi.o g/10.2139/ss n.3712658.
D’Acun o, F ancesco, And eas Fus e , and Michael Webe . 2021. “Di e se Policy Commi ees
Can Reach Unde ep esen ed G oups.” Wo king Pape 29275. Na ional Bu eau o
Economic Resea ch. h ps://doi.o g/10.3386/w29275.
Die ich, Alexande M., Edwa d S. Kno ek II, K is ian O e R. My se h, Robe W. Rich, Raphael
S. Schoenle, and Michael Webe . 2022. “G ea e Than he Sum o he Pa s: Agg ega e s.
Agg ega ed In la ion Expec a ions.” Fede al Rese e Bank o Cle eland, Wo king Pape
No. 22-20. h ps://doi.o g/10.26509/ bc-wp-202220.
33
Follmann, Dean A. 2000. “On he E ec o T ea men Among Would-Be T ea men Complie s:
An Analysis o he Mul iple Risk Fac o In e en ion T ial.” Jou nal o he Ame ican
S a is ical Associa ion 95: 1101–09. h ps://doi.o g/10.1080/01621459.2000.10474306.
Fus e , And eas, Rica do Pe ez-T uglia, Mi ko Wiede hol , and Basi Za a . 2022. “Expec a ions
wi h Endogenous In o ma ion Acquisi ion: An Expe imen al In es iga ion.” The Re iew
o Economics and S a is ics 104 (5): 1059–1078. h ps://doi.o g/10.1162/ es _a_00994
Go odnichenko, Yu iy, Tho Pham, and Oleksand Tala e a. 2023. “The Voice o Mone a y
Policy.” Ame ican Economic Re iew 113 (2): 548-84. h ps://doi.o g/10.1257/ae .20220129.
Haaland, Inga , Ch is ophe Ro h, and Johannes Wohl a . 2023. “Designing In o ma ion
P o ision Expe imen s.” Jou nal o Economic Li e a u e 61 (1): 3-40.
h ps://doi.o g/10.1257/jel.20211658.
Haldane, And ew and Michael McMahon. 2018. “Cen al Bank Communica ions and he Gene al
Public.” AEA Pape s and P oceedings 108: 578-83. h ps://doi.o g/10.1257/pandp.20181082.
Hazell, Jona hon, Juan He eño, Emi Nakamu a, and Jón S einsson. 2022. “The Slope o he
Phillips Cu e: E idence om U.S. S a es.” The Qua e ly Jou nal o Economics 137 (3):
1299–1344. h ps://doi.o g/10.1093/qje/qjac010.
Ho mann, Ma hias, Emanuel Moench, and Guido Schul e anken eld. 2023. “A KISS o Cen al
Bank Communica ion in Times o High In la ion.” mimeo
Imbens, Guido W. and Joshua D. Ang is . 1994. “Iden i ica ion and Es ima ion o Local A e age
T ea men E ec s.” Econome ica 62 (2): 467–75. h ps://doi.o g/10.2307/2951620.
Jo, Booil and Elizabe h A. S ua . 2009. “On he Use o P opensi y Sco es in P incipal Causal
E ec Es ima ion.” S a is ics in Medicine 28 (23): 2857-75.
h ps://doi.o g/10.1002/sim.3669.
Kno ek, Edwa d S., II, Raphael S. Schoenle, Alexande M. Die ich, Kei h Kues e , Ge no J.
Mülle , K is ian O e R. My se h, and Michael Webe . 2020. “Consume s and COVID-19:
A Real-Time Su ey.” Economic Commen a y 2020-08. Fede al Rese e Bank o
Cle eland. h ps://doi.o g/10.26509/ bc-ec-202008.
Lamla, Michael J. and Dmi i V. Vinog ado . 2019. “Cen al Bank Announcemen s: Big News o
Li le People?” Jou nal o Mone a y Economics 108: 21-38.
h ps://doi.o g/10.1016/j.jmoneco.2019.08.014.
34
Maćkowiak, Ba osz and Mi ko Wiede hol . 2024. “Ra ional Ina en ion du ing an RCT.”
h ps://cep .o g/publica ions/dp18875.
Maćkowiak, Ba osz, Filip Ma ějka, and Mi ko Wiede hol . 2023. “Ra ional Ina en ion: A
Re iew.” Jou nal o Economic Li e a u e 61 (1): 226-73. h ps://doi.o g/10.1257/jel.20211524.
Nakamu a, Emi and Jón S einsson. 2018. “High-F equency Iden i ica ion o Mone a y Non-
Neu ali y: The In o ma ion E ec .” Qua e ly Jou nal o Economics 133 (3): 1283–
1330. h ps://doi.o g/10.1093/qje/qjy004.
Pedemon e, Ma hieu. 2023. “Fi eside Cha s: Communica ion and Consume s’ Expec a ions in he
G ea Dep ession.” Re iew o Economics and S a is ics. h ps://doi.o g/10.1162/ es _a_01200.
P äu i, Oli e . 2024. “The In la ion A en ion Th eshold and In la ion Su ges.” Unpublished
manusc ip .
Qual ics. 2019. “ESOMAR 28: 28 Ques ions o Help Buye s o Online Samples.” Ap il 2019,
Technical Repo . h ps://www.iup.edu/a l/ iles/qual ics/esoma .pd .
Ro h, Ch is ophe and Johannes Wohl a . 2020. “How Do Expec a ions abou he Mac oeconomy
A ec Pe sonal Expec a ions and Beha io ?” Re iew o Economics and S a is ics 102 (4):
731-48. h ps://doi.o g/10.1162/ es _a_00867.
Sims, Ch is ophe A. 2003. “Implica ions o Ra ional Ina en ion.” Jou nal o Mone a y
Economics 50 (3): 665-90. h ps://doi.o g/10.1016/S0304-3932(03)00029-1.
Webe , Michael, Be na do Candia, Tiziano Ropele, Rod igo Llube as, Se a in F ache, B en H.
Meye , Sa en Kuma , Yu iy Go odnichenko, Dimi is Geo ga akos, Oli ie Coibion, Geo
Kenny, and Jo ge Ponce. 2024. “Tell Me Some hing I Don’ Al eady Know: Lea ning in
Low and High-In la ion Se ings.” Wo king Pape 31485. Na ional Bu eau o Economic
Resea ch. h ps://doi.o g/10.3386/w31485.
35
Online Appendix A. The 5 FOMC S a emen s
Exhibi A1. Ma ch 16, 2022, FOMC S a emen
36
Exhibi A2. May 4, 2022, FOMC S a emen
37
Exhibi A3. June 15, 2022, FOMC S a emen
44
Table C1. Pos e io on P io -x-T ea men
(1)
Ma ch
(2)
May
(3)
June
(4)
July
(5)
Sep
(6)
Pooled
P io
0.96***
0.92***
0.93***
0.65***
0.69***
0.91***
(0.01)
(0.01)
(0.01)
(0.01)
(0.01)
(0.01)
T ea men 1
3.26***
-0.54**
-0.30
0.49**
0.29
2.07***
(0.23)
(0.22)
(0.23)
(0.20)
(0.25)
(0.10)
T ea men 2
0.16
-0.57***
-0.50**
0.55***
0.46*
-0.45***
(0.23)
(0.22)
(0.23)
(0.20)
(0.25)
(0.10)
T ea men 3
3.22***
2.66***
(0.24)
(0.20)
T ea men 4
0.03
-0.54***
(0.23)
(0.20)
Placebo
-1.38***
0.10
(0.24)
(0.19)
T ea men 1 x P io
-0.54***
0.04***
-0.01
-0.18***
-0.10***
-0.39***
(0.02)
(0.01)
(0.01)
(0.02)
(0.02)
(0.01)
T ea men 2 x P io
-0.04***
0.03**
0.01
-0.11***
-0.20***
0.01
(0.01)
(0.02)
(0.01)
(0.02)
(0.02)
(0.01)
T ea men 3 x P io
-0.57***
-0.52***
(0.02)
(0.02)
T ea men 4 x P io
-0.08***
-0.02*
(0.02)
(0.01)
Placebo x P io
0.25***
0.03**
(0.02)
(0.01)
Obse a ions
7879
5204
4899
5994
5872
29833
Adjus ed R2
0.83
0.89
0.89
0.64
0.77
0.83
No es: Columns 1-5 con ain no con ols o ixed e ec s. Column 6 shows a pooled (o e wa es) eg ession wi h
ea men pe iod ixed e ec s.
* p < 0.10, ** p < 0.05, *** p < 0.01
The abundance o s a is ically signi ican coe icien s in Table C1 sugges s ha he
ea men s did a ec esponden s’ pos e io belie s. We ind ha he coe icien on he p io
(𝜃𝜃 in
Eq. C1, he weigh ha esponden s in he con ol g oup assigns o hei p io in o ming hei
pos e io expec a ions) is gene ally high, sugges ing a high co ela ion be ween hose wo a iables
o he con ol g oup. This sugges s ha one-yea in la ion expec a ions a e a good measu e o a
hypo he ical p io i e-yea in la ion expec a ion, which we do no di ec ly obse e. Table C1 also
shows ha , in gene al, consume s pu a lowe weigh on hei p io a e he ea men , e idence
ha hey a e lea ning om he in o ma ion ea men . We obse e bigge upda es o hose wi h
highe ex an e in la ion expec a ions. In he case o he pooled eg ession (column 6), T ea men 1
45
has a signi ican e ec when in e ac ed wi h he p io . The e is he e ogenei y in he di e en
wa es, bu , in gene al, we ind nega i e coe icien s, indica ing ha he ea men did cause
esponden s o e ise down hei p io .
Online Appendix D. T ea men Reading Times and Compliance
Figu e D1. T ea men Reading Times (Pooled Ac oss he July and Sep embe Su eys)
46
Figu e D2. Compliance P obabili ies s. Hube Weigh s
47
Figu e D3. T ea men E ec s on Complian and Unin o med Responden s
by Reading Time Pe cen ile
48
Online Appendix E. In la ion Expec a ions by In o medness and In o medness
by Demog aphics
Figu e E1: In la ion Expec a ions Based on Whe he Consume s Had Hea d
News abou Mone a y Policy
49
Table E2. P obabili y o Ha ing Hea d News abou Mone a y Policy:
Logi and Linea P obabili y Models
(1)
(2)
(3)
(4)
(5)
(6)
Male
2.22***
(0.08)
2.22***
(0.08)
2.22***
(0.08)
0.17***
(0.01)
0.17***
(0.01)
0.17***
(0.01)
Nonwhi e
0.97
(0.05)
0.97
(0.05)
0.97
(0.05)
-0.01
(0.01)
-0.01
(0.01)
-0.01
(0.01)
Hispanic
1.55***
(0.10)
1.56***
(0.10)
1.55***
(0.10)
0.09***
(0.01)
0.09***
(0.01)
0.09***
(0.01)
P ima y Shoppe
1.71***
(0.11)
1.72***
(0.11)
1.75***
(0.11)
0.11***
(0.01)
0.11***
(0.01)
0.11***
(0.01)
Nume ical Li e acy
1.10*
(0.05)
1.10*
(0.05)
1.10**
(0.05)
0.02*
(0.01)
0.02*
(0.01)
0.02*
(0.01)
Age:
36-50
0.92
(0.05)
0.92
(0.05)
0.93
(0.05)
-0.02
(0.01)
-0.02
(0.01)
-0.02
(0.01)
51-65
0.78***
(0.04)
0.78***
(0.04)
0.78***
(0.04)
-0.05***
(0.01)
-0.05***
(0.01)
-0.05***
(0.01)
66+
1.76***
(0.11)
1.76***
(0.11)
1.76***
(0.11)
0.12***
(0.01)
0.12***
(0.01)
0.12***
(0.01)
Income:
$35,000-$49,999
1.27***
(0.06)
1.27***
(0.06)
1.27***
(0.06)
0.05***
(0.01)
0.05***
(0.01)
0.05***
(0.01)
$50,000-$99,999
1.43***
(0.06)
1.43***
(0.06)
1.44***
(0.06)
0.07***
(0.01)
0.07***
(0.01)
0.07***
(0.01)
$100,000 o mo e
1.92***
(0.12)
1.92***
(0.12)
1.93***
(0.12)
0.14***
(0.01)
0.14***
(0.01)
0.14***
(0.01)
Educa ion:
Some College
1.25***
(0.06)
1.25***
(0.06)
1.26***
(0.06)
0.05***
(0.01)
0.05***
(0.01)
0.05***
(0.01)
Bachelo 's Deg ee
2.16***
(0.11)
2.16***
(0.11)
2.18***
(0.11)
0.17***
(0.01)
0.17***
(0.01)
0.18***
(0.01)
Ad anced Deg ee
3.20***
(0.21)
3.20***
(0.21)
3.21***
(0.21)
0.26***
(0.01)
0.26***
(0.01)
0.26***
(0.01)
Poli ical Pa y:
Democ a
1.34***
(0.06)
1.33***
(0.06)
1.34***
(0.06)
0.06***
(0.01)
0.06***
(0.01)
0.06***
(0.01)
Republican
1.25***
(0.06)
1.25***
(0.06)
1.25***
(0.06)
0.04***
(0.01)
0.04***
(0.01)
0.04***
(0.01)
Days Since FOMC
0.97***
(0.01)
-0.01***
(0.00)
Days Since FOMC, Sq.
1.00***
(0.00)
0.00***
(0.00)
Wa e Fixed E ec
No
Yes
Yes
No
Yes
Yes
Obse a ions
33728
33728
33728
33728
33728
33728
No es: * p < 0.10, ** p < 0.05, *** p < 0.01. Columns 1-3 show esul s as odds a ios om a logi model, while
columns 4-6 show esul s om OLS eg ession.
50
Online Appendix F. T ea men E ec s on Addi ional Va iables
Table F1. T ea men E ec s on GDP Expec a ions
ia P opensi y Sco e Weigh ed Reg essions
Complie s Non-complie s
(1)
(2)
(3)
(4)
(5)
(6)
Hea d News =
All
Yes
No
All
Yes
No
July
T ea men 1
-0.55
0.53
-1.66
2.32*
-0.02
3.73**
(0.84)
(1.15)
(1.19)
(1.40)
(2.30)
(1.78)
T ea men 2
-0.49
-1.25
-0.17
3.34**
5.21**
1.78
(0.85)
(1.14)
(1.24)
(1.42)
(2.30)
(1.83)
Obse a ions
4672
1992
2680
4283
1832
2451
Sep embe
T ea men 1
0.82
-0.80
1.79
1.26
-0.40
3.04
(1.10)
(1.47)
(1.54)
(1.60)
(2.47)
(2.16)
T ea men 2
0.44
-3.23**
2.67*
1.81
0.26
3.57*
(1.12)
(1.57)
(1.53)
(1.54)
(2.42)
(2.04)
Placebo
0.38
-2.51*
2.31
0.75
-1.73
3.45
(1.05)
(1.42)
(1.49)
(1.68)
(2.54)
(2.30)
Obse a ions
4338
1900
2438
3978
1832
2146
No es: * p < 0.10, ** p < 0.05, *** p < 0.01. The i s -s age logi eg essions a e exac ly as in Table 3. The dependen
a iable in he second-s age eg essions is he pos - ea men i e-yea a e age GDP g ow h expec a ion minus he
p e- ea men yea -ahead expec a ion, each winso ized a he 2nd and 98 h pe cen iles.
51
Table F2. T ea men E ec s on A e age Income G ow h
ia P opensi y Sco e Weigh ed Reg essions
Complie s Non-complie s
(1)
(2)
(3)
(4)
(5)
(6)
Hea d News =
All
Yes
No
All
Yes
No
July
T ea men 1
0.53
0.60
-0.18
2.32*
1.22
3.26*
(0.77)
(1.12)
(1.07)
(1.33)
(2.22)
(1.68)
T ea men 2
-0.13
-0.22
-0.64
3.17**
3.55
3.13*
(0.76)
(1.00)
(1.13)
(1.37)
(2.26)
(1.75)
Obse a ions
4672
1992
2680
4283
1832
2451
Sep embe
T ea men 1
-0.81
-0.24
-1.70
-1.18
-1.26
-0.90
(0.85)
(1.16)
(1.18)
(1.61)
(2.55)
(2.10)
T ea men 2
-0.05
-1.82
0.63
-0.48
-0.38
-0.31
(0.97)
(1.53)
(1.29)
(1.58)
(2.45)
(2.11)
Placebo
-0.27
-2.11
0.75
-1.54
-2.96
0.05
(0.90)
(1.29)
(1.24)
(1.71)
(2.70)
(2.25)
Obse a ions
4338
1900
2438
3978
1832
2146
No es: * p < 0.10, ** p < 0.05, *** p < 0.01. The i s -s age logi eg essions a e exac ly as in Table 3. The dependen a iable
in he second-s age eg essions is he pos - ea men i e-yea a e age income g ow h expec a ion minus he p e- ea men
yea -ahead expec a ion, each winso ized a he 2
nd
and 98
h
pe cen iles.
52
Online Appendix G. T ea men E ec s Condi ional on In e es Ra e P io s
Table G1. T ea men E ec s wi h Quali a i e P io In e es Ra e Change In e ac ion
Complie s
Non-complie s
(1)
(2)
(3)
(4)
(5)
(6)
Hea d News =
All
No
Yes
All
No
Yes
T ea men 1
-2.93**
-3.05**
-1.41
0.33
1.02
-2.60
(1.21)
(1.26)
(3.08)
(1.35)
(1.49)
(3.32)
T ea men 2
-3.35***
-3.24**
-3.23
0.39
1.34
-3.65
(1.29)
(1.41)
(2.54)
(1.38)
(1.51)
(3.36)
Placebo
0.60
0.42
2.93
-0.57
2.69
-14.19**
(1.73)
(1.88)
(3.91)
(2.11)
(2.01)
(6.98)
Ra es Changed
-1.88**
-1.17
-2.99*
1.97
0.45
0.01
(0.90)
(1.02)
(1.63)
(1.33)
(1.56)
(2.50)
Changed x T ea men 1
1.39
-0.49
1.72
-0.83
-1.30
1.99
(1.34)
(1.53)
(3.17)
(1.81)
(2.20)
(3.74)
Changed x T ea men 2
1.33
0.11
2.26
-1.07
-2.06
2.99
(1.41)
(1.63)
(2.65)
(1.85)
(2.22)
(3.80)
Changed x Placebo
-1.09
-1.98
-2.54
-0.43
-2.34
12.29*
(1.92)
(2.32)
(4.03)
(2.60)
(2.81)
(7.30)
Obse a ions
9010
5118
3892
8261
4597
3664
No es: * p < 0.10, ** p < 0.05, *** p < 0.01. Pooled ac oss he July and Sep embe wa es. T ea men e ec s es ima ed
using he p opensi y sco e me hodology desc ibed in Sec ion 5. The “Ra es changed” dummy =1 i esponden s
answe ed “Yes” o he ques ion, “Thinking abou he in e es a es ha people pay o bo ow money, such as
mo gage a es, would you say ha hose in e es a es in gene al ha e changed ecen ly?” and =0 i hey answe ed
“No” o “No su e.”
53
Table G2. T ea men E ec s wi h Quan i a i e P io In e es Ra e Change In e ac ion
Complie s
Non-complie s
(1)
(2)
(3)
(4)
(5)
(6)
Hea d News =
All
No
Yes
All
No
Yes
T ea men 1
-2.57**
-3.27***
0.59
0.01
1.42
-3.03
(1.01)
(1.14)
(2.09)
(1.28)
(1.42)
(2.67)
T ea men 2
-3.23***
-3.47***
-1.81
-0.39
1.19
-3.88
(1.08)
(1.27)
(1.81)
(1.32)
(1.42)
(2.90)
Placebo
0.84
0.28
3.44
-1.24
2.58
-9.65**
(1.48)
(1.70)
(2.76)
(1.83)
(1.87)
(4.19)
Ra es Up ≥ 1pp
-1.65**
-1.03
-1.89*
1.28
0.77
-0.69
(0.75)
(0.97)
(1.08)
(1.40)
(1.58)
(2.60)
Ra es Up x T ea men 1
1.01
-0.12
-0.52
-0.50
-2.74
3.02
(1.17)
(1.47)
(2.21)
(1.84)
(2.23)
(3.32)
Ra es Up x T ea men 2
1.30
0.53
0.76
0.20
-1.99
3.99
(1.23)
(1.57)
(1.96)
(1.86)
(2.30)
(3.44)
Ra es Up x Placebo
-1.49
-1.86
-3.41
0.82
-2.43
8.69*
(1.70)
(2.21)
(2.95)
(2.48)
(2.90)
(4.81)
Obse a ions
9010
5118
3892
8261
4597
3664
No es: * p < 0.10, ** p < 0.05, *** p < 0.01. Pooled ac oss he July and Sep embe wa es. T ea men e ec s es ima ed
using he p opensi y sco e me hodology desc ibed in Sec ion 5. The “Ra es Up ≥1 pp” dummy =1 i esponden s
said ha gene al bo owing a es had ecen ly inc eased by mo e han 1 pe cen age poin o he ques ion, “Thinking
abou he in e es a es ha people pay o bo ow money, such as mo gage a es, how much would you say ha
in e es a es in gene al ha e changed ecen ly?” and =0 o he wise.
60
would con inue o educe he size o i s balance shee . These ac ions we e pa o an
e o o help b ing in la ion back down owa d i s objec i e.
July Wa e
TLO 2.1 [Con ol G oup] Please p oceed o he nex ques ion.
TLO 2.2 [T ea men 1] On July 27, 2022, he Fede al Open Ma ke Commi ee (FOMC)
aised i s p ima y policy in e es a e ( he ede al unds a e) by h ee-qua e s o a
pe cen age poin , o a a ge ange o 2-1/4 o 2-1/2 pe cen . The FOMC also said ha i
would con inue o educe he size o i s balance shee .
TLO 2.3 [T ea men 2] On July 27, 2022, he Fede al Open Ma ke Commi ee (FOMC)
aised i s p ima y policy in e es a e ( he ede al unds a e) by h ee-qua e s o a
pe cen age poin , o a a ge ange o 2-1/4 o 2-1/2 pe cen , and an icipa ed ha ongoing
inc eases in he a ge ange will be app op ia e. The FOMC also said ha i would
con inue o educe he size o i s balance shee . These ac ions we e pa o an e o o
help b ing in la ion back down owa d i s objec i e.
Sep embe Wa e
TLO 4.1 [Con ol G oup] Please p oceed o he nex ques ion.
TLO 4.2 [T ea men 1] On Sep embe 21, 2022, he Fede al Open Ma ke Commi ee
(FOMC) aised i s p ima y policy in e es a e ( he ede al unds a e) by h ee-qua e s o
a pe cen age poin , o a a ge ange o 3 o 3-1/4 pe cen . The FOMC also said ha i
would con inue o educe he size o i s balance shee .
TLO 4.3 [T ea men 2] On Sep embe 21, 2022, he Fede al Open Ma ke Commi ee
(FOMC) aised i s p ima y policy in e es a e ( he ede al unds a e) by h ee-qua e s o
a pe cen age poin , o a a ge ange o 3 o 3-1/4 pe cen , and an icipa ed ha ongoing
inc eases in he a ge ange will be app op ia e. The FOMC also said ha i would
con inue o educe he size o i s balance shee . Fede al Rese e Chai Je ome Powell
said, “The FOMC is s ongly esol ed o b ing in la ion down o 2 pe cen and we will
keep a i un il he job is done.”
TLO 4.4 [Placebo] F om 2015 o 2021, he popula ion in he Uni ed S a es g ew in a
ange o 3 o 3-1/4 pe cen .
QJH13 O e he nex 5 yea s, do you hink ha he e will be in la ion o de la ion on a e age?
o In la ion
o De la ion (opposi e o in la ion)
61
QJH13a [I QJH13 = In la ion] Wha do you expec he a e age annual a e o in la ion o be
o e he nex 5 yea s? Please gi e you bes guess.
I expec he a e age annual a e o in la ion o be __ pe cen pe yea o e he nex 5 yea s.
QJH13b [I QJH13 = De la ion (opposi e o in la ion)] Wha do you expec he a e age annual
a e o de la ion o be o e he nex 5 yea s? Please gi e you bes guess.
I expec he a e age annual a e o de la ion o be __ pe cen pe yea o e he nex 5 yea s.
QJH14 O e he nex 5 yea s, do you hink ha he e will be an inc ease o dec ease in GDP on
a e age?
o Inc ease
o Dec ease
QJH14a [I QJH14 = Inc ease] Wha do you expec he a e age annual a e o inc ease in GDP
will be o e he nex 5 yea s? Please gi e you bes guess.
I expec he a e age annual a e o inc ease o be __ pe cen pe yea o e he nex 5 yea s.
QJH14b [I QJH14 = Dec ease] Wha do you expec he a e age annual a e o dec ease in
GDP will be o e he nex 5 yea s? Please gi e you bes guess.
I expec he a e age annual a e o dec ease o be __ pe cen pe yea o e he nex 5 yea s.
QJH15 In you iew, will he o al income o all membe s o you household (including you),
a e axes and deduc ions, inc ease o dec ease o e he nex 5 yea s on a e age?
o Inc ease
o Dec ease
QJH15a [I QJH15 = Inc ease] Wha do you expec he a e age annual a e o inc ease in he
o al income o all membe s o you household will be o e he nex 5 yea s? Please gi e you
bes guess.
I expec he a e age annual a e o inc ease in he o al income o all membe s o my household
o be __ pe cen pe yea o e he nex 5 yea s.
QJH15b [I QJH15 = Dec ease] Wha do you expec he a e age annual a e o dec ease in
GDP will be o e he nex 5 yea s? Please gi e you bes guess.
I expec he a e age annual a e o dec ease in he o al income o all membe s o my household
o be __ pe cen pe yea o e he nex 5 yea s.
QJH18a Wha do you hink is he chance ha in la ion will be mo e han 4% in he nex 12
mon hs?
62
Q53. Wha is you ci il s a us?
o Single
o Pa ne (no co-habi ing)
o Pa ne (co-habi ing)
o Ma ied
o Di o ced
o Widowed
Q121 Wha would you say is you poli ical a ilia ion?
o Democ a
o Independen
o Republican
o O he
Q54 How many child en do you ha e? ___
QX2 Imagine he e a e whi e and black balls in a ballo box. You d aw a ball 70 imes. 56 imes,
you ha e d awn a whi e ball, 14 imes a black ball. Gi en his eco d, wha would you say is
he p obabili y o d awing a black ball he nex ime?
The p obabili y is ___ pe cen .
