B anca elli, Caloge o; Inde s , Roman
A icle — Published Ve sion
CPG consump ion in imes o ecession: no el e idence
om ma ched adminis a i e da a
Quan i a i e Ma ke ing and Economics
P o ided in Coope a ion wi h:
Sp inge Na u e
Sugges ed Ci a ion: B anca elli, Caloge o; Inde s , Roman (2025) : CPG consump ion in imes
o ecession: no el e idence om ma ched adminis a i e da a, Quan i a i e Ma ke ing and
Economics, ISSN 1573-711X, Sp inge US, New Yo k, NY, Vol. 23, Iss. 2, pp. 265-289,
h ps://doi.o g/10.1007/s11129-025-09296-5
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/323527
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 p://c ea i ecommons.o g/licenses/by/4.0/
Vol.:(0123456789)
Quan i a i e Ma ke ing and Economics (2025) 23:265–289
h ps://doi.o g/10.1007/s11129-025-09296-5
CPG consump ion in imeso ecession: no el e idence
omma ched adminis a i e da a
Caloge oB anca elli1· RomanInde s 2
Recei ed: 15 May 2023 / Accep ed: 2 Feb ua y 2025 / Published online: 10 Ma ch 2025
© The Au ho (s) 2025
Abs ac
We u ilize a no el da ase ha me ges household adminis a i e income and socio-
demog aphic in o ma ion om ax eco ds wi h scanne da a on CPG consump ion.
Ou analysis e eals signi ican a ia ion in household pe -capi a expendi u es.
Howe e , we ind only a modes economic ela ionship be ween CPG spending and
income, e en amids subs an ial wi hin-household income luc ua ions du ing he
Du ch "double-dip ecession" om 2011 o 2018. This ela ionship emains small
o households wi h low income and low liquidi y and holds ac oss bo h ood and
non- ood expendi u es.
Keywo ds Income e ec s· Consume -packaged goods· Adminis a i e da a
JEL Classi ica ion D12· E21· E32· M30
This pape is he esul o a collabo a ion wi h he AiMa k ounda ion and Cen aal Bu eau oo
de S a is iek (CBS) Ne he lands. We hank bo h o hei con inuous suppo . We a e also g a e ul
o commen s ecei ed om Thomas O e and Michalis Haliassos and pa icipan s o he 2023
Bundesbank In e na ional Household Finance Con e ence. The au ho s bene i ed om inancial
suppo om a long- e m esea ch g an by he Think Fo wa d Ini ia i e (TFI). Du ing he s udy
implemen a ion, B anca elli wo ked o Goe he Uni e si y F ank u and is now wo king o he
Eu opean Cen al Bank (ECB). The iews exp essed he ea e ou own and do no necessa ily e lec
hose o he ECB o he Eu osys em.
* Roman Inde s
[email p o ec ed]
Caloge o B anca elli
Caloge [email p o ec ed]
1 Eu opean Cen al Bank, F ank u AmMain, Ge many
2 Goe he Uni e si y F ank u , F ank u AmMain, Ge many
266
C.B anca elli, R.Inde s
1 In oduc ion
Consump ion o consume packaged goods (CPGs) cons i u es a sizable ac ion
o all excep he iches households’ consump ion and disposable income (on a e -
age 9.95% in ou sample). Analyzing how changes in income a ec household CPG
expendi u e, o example o e he business cycle, is hus o in e es bo h o ma -
ke ing schola s and p ac i ione s as well as o mac oeconomis s and policymake s.
Ye , quan i ying his ela ionship wi h su ey da a p o es challenging due o meas-
u emen e o s in household-speci ic income and weal h componen s and a lack o
g anula i y in expendi u es.
In his pape , we add ess hose challenges by d awing on a no el da a sou ce ha
links indi idual households’ CPG expendi u es om G K’s scanne panel1 wi h
o al disposable income om adminis a i e da a, as collec ed by he Du ch s a is i-
cal o ice (Cen aal Bu eau oo de S a is iek, CBS).As a main esul , we ind only
an economically small ela ionship be ween CPG spending and income as a 10%
change in disposable household income is ela ed o only a 0.55% change in CPG
consump ion.2
Ou empi ical s a egy exploi s he g anula panel s uc u e o ou da ase and
d aws on ins i u ional speci ics o g oce y e ailing in he Ne he lands and p o-
nounced household-speci ic income a ia ions. No ably, he ime ame o ou anal-
ysis co e s he pe iod be ween 2011 and 2018 and hus he “second dip” (2011 o
2013) o he Ne he lands’ double-dip ecession ollowing he inancial c isis. O e
his pe iod, GDP g ow h d opped by 3.6 pe cen age poin s om he ou h qua e in
2011 o he hi d qua e in 2013 and unemploymen ose by 3.3 pe cen age poin s.
One- hi d o income changes in ou da a a e nega i e. We also documen conside -
able emaining he e ogenei y be ween households e en a e s ipping ou (yea -by-
yea ) agg ega e luc ua ions. Thanks o his ins i u ional se up, household-speci ic
income a ia ion can be conside ed as good as andomly assigned. We also no e
ha ou measu e o income, namely household disposable income, is mo e comp e-
hensi e han main ea ne income alone, which is commonly used in he li e a u e
and ypically (sel -) epo ed in household scanne da a. Mo eo e , elying on admin-
is a i e da a educes he isk ha esul s a e downwa ds biased by measu emen
e o s. S ill, o es his mo e o mally, we co obo a e ou inding o only an eco-
nomically small ela ionship be ween household income and CPG expendi u e using
G K sel - epo ed igu es. Ou indings a e obus o a se ies o di e en al e na-
i e speci ica ions. Fi s , ou key inding o only an economically small ela ionship
be ween household income and CPG expendi u e holds o bo h ood and non- ood
1 Resea che s om he U.S. may be mo e amilia wi h Nielsen’s household (o homescan) panel da a.
In many o he coun ies, he panels p o ided by G K o he sis e o ganiza ion Kan a u nish he same
se ice. In he Ne he lands, Nielsen does no ha e a household panel, making G K’s panel he indus y
s anda d.
2 Though in ou main eg ession, he income coe icien is s a is ically signi ican a he 1%-le el, gi en
i s small size con idence in e als a e ela i ely la ge. S ill, in he epo ed case he uppe bounda y
o he 95%-con idence in e al co esponds o only a 0.74% inc ease in consump ion ollowing a 10%
inc ease in income.
267
CPG consump ion in imeso ecession: no el e idence om…
ca ego ies. Second, i also holds o households wi h low income o low sa ings.
Thi d, ou indings s ill apply when conside ing only nega i e income changes,
which accoun o a ound one- hi d o obse a ions in ou sample.
Despi e he obus ness o ou esul s, ou analysis e eals conside able di e -
ences in c oss-sec ional pe -capi a CPG spending pe household, e en hough in he
Ne he lands, he e ailing landscape is pa icula ly homogeneous. As we documen
na ional p icing and asso men s a egies o supe ma ke and discoun e chains,
we can ule ou such supply side explana ions o his a ia ion. Ou indings hus
sugges ha di e ences in household (pe -capi a) CPG expendi u es may be la gely
a esul o di e ences in as e and, consequen ly, households do no show s ong
adjus men s in expendi u es when acing changes in income.
Ou indings a e ele an o ma ke ing p o essionals in he CPG and e ailing
indus y. In addi ion, hey a e o ele ance o policy-o ien ed (mac oeconomic)
wo k. I sugges s ha in coun ies wi h a simila ins i u ional and socioeconomic
backg ound as he Ne he lands, CPG consump ion is pa icula ly s able ac oss he
business cycle, despi e p onounced di e ences in pe -capi a household expendi u es
and o en la ge p ice di e ences be ween p oduc s in he same ca ego y, such as
be ween s o e b ands and na ional b ands.
Ou disagg ega e analysis con ibu es o a la ge s and o li e a u e on business
cycle esea ch in ma ke ing (Dekimpe & Delee snyde , 2018). While his li e a u e
has iden i ied la ge e ec s o some ca ego ies such as consume du ables (e.g.,
Deele snyde e al., 2004), gene ally e ec s a e much less p onounced o necessa y
consume goods (c . Du & Padmanabhan, 2011, which uses agg ega e da a in a
c oss-coun y analysis o cu ency c ises). Relying on wi hin-household a ia ions,
we ind only an economically small ela ionship. Dubé e al. (2018), whose app oach
we ollow, ind a small ela ionship be ween epo ed income and households’ p i-
a e-label sha e in he US. Biswas e al. (2021) ex end his o a join analysis o how
households a y hei budge alloca ions ac oss p oduc ca ego ies as well as s o e
o ma s and b and ypes, which allows hem o de i e mo e nuanced business impli-
ca ions, e.g., ega ding he e ec s o a ecession on non- ood s. ood i ems sold in
di e en s o e o ma s. B anca elli e al. (2022) apply his me hodology o he choice
o s o e b ands in he Ne he lands, con i ming an economically small ela ionship
wi h income changes. The economically small ela ionship be ween s o e b and con-
sump ion and income change o e he business cycle con as s s ongly wi h ea lie
indings using only agg ega e da a (e.g., Lamey e al., 2007).
The e is also sizable li e a u e in mac oeconomics, especially household
inance, ha seeks o measu e he esponse o (households’) agg ega e consump-
ion o changes in income.3 The e, changes in consump ion a e o en measu ed
3 S ill, a s and o he la ge li e a u e s udies he shape o consume demand esponses o income
changes also o di e en consump ion ca ego ies. Typically, ood, which makes up he bulk o CPG
expendi u e, is ound o be a no mal good, so ha spending inc eases wi h income, albei i s sha e in
o e all expendi u e dec eases. While we s ipula e a linea ela ionship be ween log income and log
expendi u e, he li e a u e has also analyzed mo e lexible unc ional o ms (e.g., Banks e al., 1997).
268
C.B anca elli, R.Inde s
by su eys, based on dia ies o in e iews, a he han by eco ding ac ual expen-
di u es.4 In such su eys also hypo he ical ques ions can be asked, which allows
o disen angling esponses o an icipa ed s. non-an icipa ed as well as empo a y
s. pe manen changes in income (see, o ins ance, Jappelli & Pis a e i, 2014 o
Ch is elis e al., 2019). A key inding o his li e a u e is ha households would
consume a su p isingly la ge ac ion o only ansi o y income shocks, con adic -
ing he “pe manen income hypo hesis”. These s udies o en s ess he impo ance
o household liquidi y as a media ing ac o . While esea ch has also documen ed
such excess sensi i i y o consump ion o eal ansi o y income shocks (such as
om he ede al go e nmen shu down in 2013, Bake & Yannelis, 2017, o om
ax eba es in 2018, Mis a & Su ico, 2014), in e es ingly a s udy by Pa ke e al.
(2013) documen ed only small e ec s o expendi u es on ood and non-du able
i ems. No ably, ou da a does no allow us o dis inguish be ween an icipa ed and
non-an icipa ed income changes. S ill, we co obo a e ou inding o only a small
ela ionship wi h ac ual CPG spending by including lags. In con as o his mac-
oeconomic li e a u e, we do no ind ha households wi h low income o sa ings
eac signi ican ly mo e s ongly.
The emainde o his pape is o ganized as ollows: In sec ion “Da a”, we in o-
duce ou da ase . In sec ion “S ylized ac s”, we p esen s ylized ac s abou CPG
consump ion and e ailing. Sec ion “Es ima ion esul s” ep esen s he esul s om
ou empi ical analysis. We conclude in sec ion “Conclusion”. Addi ional esul s a e
p o ided in an appendix.
2 Da a
We use and en ich a no el da ase in oduced by B anca elli e al. (2022) ma ching
household scanne da a om he Du ch G K consume panel wi h adminis a i e da a
on weal h, income, and sociodemog aphic cha ac e is ics, as p o ided by CBS.5 Ma ch-
ing he da ase s was possible as mo e han 5,000 households indi idually consen ed.6
Be ween 2011 and 2018 we a e hus able o obse e on he le el o indi idual house-
holds bo h hei pu chases and ac ual, a he han only epo ed, in o ma ion ega ding
income and weal h. Ou le el o agg ega ion is he household, a he han indi idu-
als bo h o he conside ed expendi u es as well as o income. When using household
scanne da a alone, ypically epo ed income ins ead only ela es o he main ea ne .
In he ollowing subsec ions, we desc ibe each o ou wo main da ase s in de ail.
4 Mac oeconomic esea ch has only ecen ly u ned owa ds he use o household scanne da a, see o
ins ance Kaplan and Schulho e -Wohl (2017) o he measu emen o in la ion. A comp ehensi e e iew
o he use o household scanne da a in a ious s ands o economic li e a u e is p o ided by Dubois
e al. (2022).
5 Household scanne da a om he G K consume panel a e p o ided by he AiMa k ounda ion.
6 Ma ching, based on household cha ac e is ics p o ided di ec ly by G K, ook place in he secu e en i-
onmen o CBS and was unde aken by i s s a is ical s a . All analyses wi h ma ched da a ook place in
he secu e en i onmen o CBS.
269
CPG consump ion in imeso ecession: no el e idence om…
2.1 Household scanne da a
We use household scanne da a om he G K consume panel, which is he
main p o ide o such in o ma ion in he Ne he lands. I co e s household pu -
chases o a wide ange o CPGs, bo h ood and non- ood.7 To egis e pu -
chases, households selec ed o he G K consume panel a e equipped wi h an
elec onic home scanning de ice. Ou uni o obse a ion is hus a single p od-
uc on he ba code le el bough by an indi idual household on a ce ain day a
a speci ic e ail chain. Fo ou analysis, we agg ega e he da a on a household-
yea le el.
Ma ching wi h CBS da a is only possible o hose households ha explici ly con-
sen ed o he ma ching o hei da a.8 This es ic s ou unbalanced panel om o igi-
nally 11,041 o 6,151 dis inc households in 2018. O hese, we obse e 89% o e
he en i e pe iod. Also, no e ha we a e s ill able o d aw on he ull Du ch popula-
ion om CBS and we use his in o ma ion bo h o analyze he ep esen a i eness o
ou sample and o cons uc eg ession weigh s.
We ob ain om G K also households’ sel - epo ed (mean ea ne ) income da a,
which is collec ed annually by G K. We use his o co obo a e ou main indings
based on CBS adminis a i e income da a. Panelis s a e asked o epo a e age
mon hly income ne non- inancial income (including sala y, income om sel -
employmen , pensions, and social bene i s). Responses a e p o ided a he end o a
gi en yea and G K e eshes panel in o ma ion in he i s qua e o he subsequen
yea . We ake his lag in o accoun .
We conduc he ollowing da a selec ion. We neglec households o which we
obse e less han six mon hs o pu chase da a o less han h ee shopping ips on
a e age pe mon h. Nex , we exclude ansac ions wi h un easonably high o low
p ices by emo ing ansac ions wi h p ices ha a e ou imes highe han he
median o less han a qua e o he median o he same p oduc . We iden i y he
same p oduc ia ba codes and calcula e median p ices ac oss households and
mon hs o each yea . O e all, we lose less han 1% o o al expendi u es. To cap u e
ull yea s o ansac ions o each household, he eby a oiding seasonal biases as
ou le el o obse a ion consis s o household-yea s, we exclude panel membe s
ha en e ed he panel a e Janua y o a pa icula yea . We he eby lose 6.42% o
households.
7 This includes esh, ozen, e ige a ed ood, alcoholic and non-alcoholic be e ages, heal h and beau y
p oduc s as well as pe needs, cleaning, de e gen p oduc s, and gene al me chandise a a wide a ie y o
e ail ou le s.
8 In he Eu opean Union in May 2018 he Gene al Da a P o ec ion Regula ion (GDPR) came in o o ce,
which se e ely limi s he c ea ion, s o age, and use o pe sonal da a. Due o he GDPR, G K is obliged o
ask o he consen o each household. In 2018 all households, which we e in he panel a he ime, had
he oppo uni y o olun a ily ag ee o hei household scanne da a being linked o CBS da a.
270
C.B anca elli, R.Inde s
2.2 CBS da a
The Du ch Cen aal Bu eau oo de S a is iek (CBS), also e e ed o as S a is ics
Ne he lands, is a go e nmen al ins i u ion ha collec s s a is ical in o ma ion in he
Ne he lands. I p o ides mic oda a o au ho ized ins i u ions unde s ic condi ions
o con iden iali y. CBS mic oda a da a ha e been used p e iously in he li e a u e,
o ins ance in De Meije e al. (2013), Lamme s e al. (2013), and Raymond e al.
(2010).
CBS da a include, in ou case, households’ sociodemog aphic in o ma ion
and in o ma ion on income and weal h. We use CBS in o ma ion on he numbe
o household membe s (a he same add ess), he numbe o child en, he numbe
o male o emale pe sons, and he numbe o household membe s ecei ing an
income. Changes in hese a iables may a ec he equi ed consump ion baske and
he eby expendi u es. We hus accoun o he composi ion o a household h ough
he inclusion o hese a iables in ou eg ession analysis below.
In he desc ip i e pa s o his pape , we adjus income, weal h, and consump-
ion igu es o household size. We ollow A anasio and Pis a e i (2016) and de ine
household size as
1+0.7(n−1)+0.3k
, whe e
n
is he numbe o adul s and
k
he
numbe o child en in he household.9 Fo ins ance, i a single-pe son household
g ows wi h he addi ion o an adul , he headcoun inc eases om 1 o 1.7.
Ou key explana o y a iable e e s o disposable household income. CBS p o-
ides esea che s wi h highly accu a e income da a om Du ch ax au ho i ies. As
some componen s o income, such as ha om sel -employmen , a e only a ailable
yea ly and as his gene ally applies o ne income a e axes, which is co ec ly cal-
cula ed only a he end o he yea , we use yea ly income o ou analysis. Dispos-
able (household) income is de ined as g oss household income minus deduc ions,
consis ing o ans e s and ax paymen s o all household membe s.10
CBS also p o ides de ailed in o ma ion abou households’ inancial weal h.
Financial asse s consis o deposi and cu en accoun balances and he ma ke
alue o s ocks and bonds as o 1s Janua y o he ollowing yea . Du ch households
need o decla e his in o ma ion in hei annual income ax decla a ion o de i e
he base o he income ax on e u n on asse s.11 As we explain below, he use o
9 We no e ha his does no ully co espond o scales e e ed o as he OECD (modi ied o equi a-
lence) scales (see h ps:// www. oecd. o g/ els/ soc/ OECD- No e- Equi alenc eScal es. pd ). We hank a e e ee
o poin ing his ou . None o he insigh s om ou desc ip i es would, howe e , be a ec ed by changing
he scale.
10 G oss household income comp ises labo income, business income, capi al income, social ans e s,
insu ance bene i s, allowances, and alimony om ex-spouses o all household membe s. Deduc ed ans-
e paymen s comp ise income ans e s, such as alimony o ex-spouses, as well as manda o y social
ans e s o social assis ance and na ional insu ance bene i s. Taxes deduc ed om g oss household
income consis o all income ax and a weal h ax paid wi hin a household.
11 Whene e no such in o ma ion is p o ided in he income ax decla a ion, CBS uses epo ed da a
om he espec i e inancial ins i u ions o impu e hese alues indi idually. As o 2016, his applies also
o o eign inancial asse s ha a e in coun ies ha a e pa o he OECD Common Repo ing S anda d
adop ed by he Eu opean Commission.
271
CPG consump ion in imeso ecession: no el e idence om…
inancial weal h da a allows us o analyze whe he households wi h smalle sa ings
eac mo e s ongly o income changes as o en sugges ed by he li e a u e.
Th ough CBS we also ha e access o he g anula da a on he income o all Du ch
households. We use his bo h o check o he ep esen a i eness o ou sample and
o pe o m a weigh ed analysis. We commen below on he ep esen a i eness and,
in his con ex , also p o ide desc ip i e s a is ics o socio-demog aphic cha ac e is-
ics, including income and inancial weal h, bo h o ou sample and he popula ion
(c . also Table10 in he appendix).
2.3 Rep esen a i eness o ma ched da a
Via CBS, we ha e access o household-le el in o ma ion o he whole Du ch popu-
la ion and can compa e his o ou sample. In Table1, we i s p o ide de ailed in o -
ma ion on ou main explana o y a iable o in e es , disposable income, which we
ha e de la ed and s anda dized o household size.
Table1 shows ha he mean, median, and key pe cen iles o he income a iable
in ou sample a e simila o hose ob ained om he popula ion as a whole. Table10
in he appendix p o ides, in addi ion, key s a is ical desc ip i es o all ele an soci-
odemog aphic a iables, including income and inancial weal h, and o bo h end-
poin s o ou conside ed pe iod om 2011 o 2018. Once mo e, we epo igu es
Table 1 Summa y s a is ics o
disposable income in sample
and popula ion
Table1 shows summa y s a is ics o disposable income pe s and-
a dized household membe . Values a e exp essed in cons an 2018
Eu o. Household size is s anda dized using an OECD scale de ined
as 1 + 0.7(n- 1) + 0.3k, whe e n is he numbe o adul s and k he
numbe o child en in he household
Disposable pe capi a income
2011 2018
Mean—S 23,379.33 24,037.68
Mean- Pop 23,532.35 24,852.26
P10—S 14,736.30 14,683.00
P10—Pop 12,164.27 12,513.91
P25—S 17,660.22 17,795.22
P25—Pop 15,925.80 16,552.00
P50—S 21,993.73 22,571.00
P50—Pop 21,001.24 22,349.00
P75—S 27,269.93 28,410.00
P75—Pop 27,768.59 29,635.88
P90—S 33,986.74 35,008.50
P90—Pop 36,471.79 38,164.17
N—S 3,110 5,451
N—Pop 7,544,704 7,973,137
272
C.B anca elli, R.Inde s
bo h o ou sample and o he whole popula ion. O e all, his con i ms ha along
hese dimensions ou sample is la gely compa able o he whole popula ion.
No ably, when looking a simila i ies along he dis ibu ion, we ind ha he la g-
es di e ences p e ail a he 10 h and 90 h pe cen iles. In pa icula , his may e lec
well-known p oblems in he unde - ep esen a i eness o comme cial panels a he
ails o he income dis ibu ion. To alle ia e conce ns ha his may bias he e ec s
o in e es in ou analysis, we exploi he access o CBS egis y da a o cons uc
sampling weigh s based on he p ocedu e desc ibed in B anca elli e al. (2022). Fo
ou main eg ession and obus ness checks we p o ide sepa a e esul s including
hose weigh s.12
3 S ylized ac s
3.1 CPG consump ion and e ailing
Figu e1 illus a es ha household expendi u es o CPGs ep esen a sizable compo-
nen o o al disposable income, namely be ween 9.5% and 11.5% o e ime.
The p oduc s unde conside a ion a e di ided in o wo main segmen s: ood and
non- ood i ems. The non- ood segmen includes, o example, de e gen s, cleaning
p oduc s, and pe supplies. Food i ems consis o be e ages as well as a ious ypes
o esh and e ige a ed oods. Ou da a allow us o b eak down o al CPG expendi-
u es in o ood and non- ood componen s, which we dis inguish be ween a a ious
poin s in he analysis. O e all, ood consump ion accoun s o 82.1% o o al CPG
expendi u e.
Table 2 depic s he dis ibu ion o CPG ( ood and non- ood) spending le els
pooled ac oss all household-yea obse a ions, whe e, again, household size has
been s anda dized using he OECD scale. Values a e exp essed in cons an 2018
Eu o (using as a de la o he Du ch Ha monized Index o Consume P ices, HICP).
The median CPG consump ion le el pe yea and s anda dized household membe s
is 1,968 eu os, which masks subs an ial he e ogenei y. Rema kably, o al CPG con-
sump ion pe s anda dized household membe a he 90 h pe cen ile is h ee imes
la ge han consump ion a he 10 h pe cen ile.
This he e ogenei y emains e iden e en when examining a ia ions wi hin
ela i ely na ow b acke s o household disposable income (e.g., quin iles o
12 Speci ically, we cons uc yea -speci ic adjus men weigh s using pos -s a i ica ion. We ollow Be h-
lehem and Bi ignandi (2011) and s a by s a i ying ou eg ession samples in o s a a using disposable
income om he CBS household income egis y INHATAB. Fo each yea , we i s spli he popula-
ion in o income deciles o c ea e 10 s a a o equal size. Nex , we compu e o each s a um a co ec-
ion weigh as he a io be ween he popula ion pe cen age o households o e he sample pe cen age
o households. Finally, we mul iply he co ec ion weigh by he in e se p obabili y o each household
being in he sample. Since he ue (and household-speci ic) p obabili y o being sampled is unknown
and since we assume ha households wi hin he s a um a e homogeneous conce ning hei inclusion
p obabili y, we assign an equal inclusion p obabili y o each household wi hin he s a um. We impu e
he s a um-speci ic inclusion p obabili y wi h he a io o households wi hin he s a um o e he o al
numbe o households in he popula ion.
279
CPG consump ion in imeso ecession: no el e idence om…
We in e p e hese esul s o be consis en wi h he iew ha households may
adjus consump ion also wi h some delay o an ini ial income change, o ins ance,
due o a change in pe cep ions o due o e ised expec a ions abou u u e income.
4.3 Po en ially asymme ic esponses
Va ious s udies ha e documen ed asymme ies be ween up- and downswings in he
economy o likewise income inc eases and dec eases. Fo ins ance, Deele snyde
e al. (2004) documen his o e he business cycle o du able goods. Wi h ega ds
o CPG, he e is also some e idence ega ding asymme ic esponses o consum-
e s’ sha e o p i a e labels (Lamey e al., 2007). In mac oeconomics, based on su -
ey ques ions a ious s udies ind ha households eac mo e s ongly o nega i e
changes (Ch is elis e al., 2019). Such asymme ies may de i e om he way house-
holds gain o lose us in he economic clima e. Also, i has been sugges ed ha
consume s may i s use addi ional inancial slack o pay back deb s du ing eco e y
Table 4 Main eg ession analysis
Table4 p esen s he es ima ion esul s o he baseline eg ession. The dependen a iable is he log o
o al CPG expendi u es pe yea . Column 1 p esen s esul s including only he log o disposable income,
nex o household and yea ixed e ec s. Column 2 adds demand side a iables, column 4 includes he
lag o log disposable income. Column 5 adds supply-side con ols o he speci ica ion in column 2.
S anda d e o s, clus e ed a he household le el, a e displayed in pa en heses. Signi icance le els a e
de ined as ollows: * p < 0.10, ** p < 0.05, *** p < 0.01. All models a e es ima ed using a cons an
(1) (2) (3) (4) (5)
Log(CPG) Log(CPG) Log(CPG) Log(CPG) Log(CPG)
Log(HH income) 0.134*** 0.060*** 0.058*** 0.048*** 0.060***
(0.012) (0.010) (0.011) (0.011) (0.010)
L.Log(HH income) 0.047***
(0.009)
Sha e o emales in HH −0.110*** −0.108*** −0.090** −0.109***
(0.036) (0.038) (0.038) (0.036)
Sha e o child en in HH 0.126*** 0.138*** 0.128*** 0.125***
(0.030) (0.034) (0.033) (0.030)
Sha e o income ea ne s in HH 0.007 0.003 −0.003 0.007
(0.018) (0.022) (0.020) (0.018)
HH membe s 0.082*** 0.078*** 0.075*** 0.081***
(0.007) (0.008) (0.007) (0.007)
Household ixed e ec s Yes Yes Yes Yes Yes
Yea ixed e ec s Yes Yes Yes Yes Yes
Supply-side con ols No No No No Yes
Sample weigh s No No Yes No No
Obse a ions 30,543 30,543 30,543 25,309 30,543
R20.900 0.903 0.901 0.917 0.903
Adjus ed R20.880 0.884 0.881 0.898 0.884
280
C.B anca elli, R.Inde s
o ha di e ences a e accoun ed o by loss a e sion o a p ecau iona y sa ings
mo i e (see also below o a speci ic analysis o consume s’ inancial si ua ion).
We de ine an income change by calcula ing he yea -on-yea pe cen age change
o de la ed disposable income o each household. Nex , we di ide he sample by
sepa a ing nega i e a es om hose being posi i e o ze o. O e all, 36% o all
measu ed income changes a e nega i e.
As epo ed in Table5, he coe icien emains economically small also o nega-
i e income changes: The coe icien ’s size in column 2 implies ha a 10% dec ease
in disposable income is associa ed wi h a 0.81% dec ease in CPG consump ion. We
no e also ha while he epo ed coe icien is sligh ly la ge o nega i e income
changes han o posi i e income changes, he espec i e con idence in e als a e
again ela i ely la ge. When we es ima e a single eg ession wi h a dummy deno -
ing nega i e income changes, he espec i e coe icien is no s a is ically signi ican
(e en on a 10%-le el).
Table 5 Nega i e s. posi i e income changes
Table5 p esen s he es ima ion esul s sepa a ed by nega i e (columns 1 h ough 2) and posi i e (col-
umns 3 h ough 4) income changes. The dependen a iable is he log o o al CPG expendi u es pe
yea . Column 1 ( espec i ely 3) p esen s esul s including only he log o disposable income, nex o
household and yea ixed e ec s. Column 2 ( espec i ely 4) adds demand side a iables. S anda d e o s,
clus e ed a he household le el, a e displayed in pa en heses. Signi icance le els a e de ined as ollows:
* p < 0.10, ** p < 0.05, *** p < 0.01. All models a e es ima ed using a cons an
Nega i e income changes Posi i e income changes
(1) (2) (3) (4)
Log(CPG) Log(CPG) Log(CPG) Log(CPG)
Log(HH income) 0.162*** 0.085*** 0.139*** 0.063***
(0.024) (0.023) (0.016) (0.015)
L.Log(HH income)
Sha e o emales in HH −0.128** −0.111**
(0.058) (0.045)
Sha e o child en in HH 0.079 0.142***
(0.069) (0.034)
Sha e o income ea ne s in HH −0.004 0.022
(0.035) (0.022)
HH membe s 0.086*** 0.079***
(0.012) (0.009)
Household ixed e ec s Yes Yes Yes Yes
Yea ixed e ec s Yes Yes Yes Yes
Supply-side con ols No No No No
Obse a ions 10,994 10,994 19,549 19,549
R20.932 0.935 0.910 0.912
Adjus ed R20.890 0.894 0.877 0.881
281
CPG consump ion in imeso ecession: no el e idence om…
4.4 Using G K sel ‑ epo ed income da a
We nex conduc ou main eg ession using now G K sel - epo ed income.
As shown in Table 6, he income coe icien emains small e en wi h he
upda ed measu e: a coe icien o 0.079 indica es ha a 10% change in disposable
income is associa ed wi h a 0.76% change in CPG consump ion. While his aligns
wi h ou p e ious indings, a de ailed compa ison be ween coe icien s de i ed
om adminis a i e e sus epo ed income da a wa an s u he examina ion.
Using G K sel - epo ed income, one migh expec a weake ela ionship wi h
expendi u es o se e al easons: (1) G K da a only cap u es he income o he
household’s main ea ne , (2) i is eco ded in 200-Eu o b acke s, po en ially limi -
ing a iabili y, and (3) i may con ain epo ing inaccu acies, including eluc ance
Table 6 Analysis using G K sel - epo ed income. (G K’s income igu es a e also censo ed a he bo om
and he op, wi h espec i e ca ego ies o below 700 EUR and abo e 4.100 EUR, hough only ca. 6% o
households all in o hese ca ego ies. As we associa e each household wi h he mean o he espec i e
b acke , we lose hose in he highes band. In addi ion, o a ound 7% o household yea s income is miss-
ing, which oge he explains he d op in obse a ions.)
Table 6 p esen s he es ima ion esul s o he baseline eg ession using sel - epo ed income. The
dependen a iable is he log o o al CPG expendi u es pe yea . Column 1 p esen s esul s including
only he log o disposable income, nex o household and yea ixed e ec s. Column 2 adds demand-side
a iables, column 3 includes he lag o log disposable income, and column 5 supply-side con ols o he
speci ica ion in column 2. S anda d e o s, clus e ed a he household le el, a e displayed in pa en heses.
Signi icance le els a e de ined as ollows: * p < 0.10, ** p < 0.05, *** p < 0.01. All models a e es ima ed
using a cons an
(1) (2) (3) (4)
Log(CPG) Log(CPG) Log(CPG) Log(CPG)
Log(HH income) 0.114*** 0.079*** 0.075*** 0.079***
(0.014) (0.013) (0.015) (0.013)
L. Log(HH income) 0.002
(0.015)
Sha e o emales in HH −0.106*** −0.086** −0.105***
(0.037) (0.040) (0.037)
Sha e o child en in HH 0.081*** 0.069** 0.080**
(0.031) (0.035) (0.031)
Sha e o income ea ne s in HH 0.020 0.014 0.019
(0.019) (0.021) (0.019)
HH membe s 0.094*** 0.090*** 0.094***
(0.007) (0.007) (0.007)
Household ixed e ec s Yes Yes Yes Yes
Yea ixed e ec s Yes Yes Yes Yes
Supply-side con ols No No No Yes
Obse a ions 28,328 28,312 23,233 28,312
R20.901 0.905 0.918 0.905
Adjus ed R20.880 0.885 0.899 0.885
282
C.B anca elli, R.Inde s
among panelis s o upda e hei epo ed income. In B anca elli e al. (2022), we
sugges ed ha hese ac o s likely con ibu ed o he weake ela ionship be ween
sel - epo ed income and households’ p i a e label sha es. Howe e , sel - epo ed
income may be e cap u e households’ pe cei ed economic well-being, possi-
bly indica ing a s onge link o consump ion. Ou cu en indings on o al CPG
consump ion suppo his pe spec i e, and we e isi his issue in ou concluding
ema ks.
4.5 Analysis o po en ial he e ogenei y
Ou inding o only a modes ela ionship be ween income measu es and CPG con-
sump ion may conceal e ec he e ogenei y wi hin he sample, sugges ing po en ially
la ge coe icien s o ce ain subg oups. To in es iga e his possibili y, we analyze
Table 7 Analysis o po en ial he e ogenei y
Table7 p esen s he es ima ion esul s o modi ied baseline eg essions. All columns use he same se
o con ol a iables as in baseline Table4. Column (1) epea s he baseline om Table4. In column (2)
he dependen a iable is he log o o al CPG ood expendi u es pe yea . In column (3) he dependen
a iable is he log o o al CPG non- ood expendi u es pe yea . Column 4 p esen s esul s including an
in e ac ion e m o Log(income) and a dummy a iable o low household income. Column 5 p esen s
esul s including an in e ac ion e m o Log(income) and a dummy a iable o low household inancial
weal h. S anda d e o s, clus e ed a he household le el, a e displayed in pa en heses. Signi icance le els
a e de ined as ollows: * p < 0.10, ** p < 0.05, *** p < 0.01. All models a e es ima ed using a cons an
(1) (2) (3) (4) (5)
Log(CPG) Log(Food) Log(Non- ood) Log(CPG) Log(CPG)
Log(HH income) 0.060*** 0.060*** 0.058*** 0.069*** 0.062***
(0.010) (0.010) (0.020) (0.013) (0.011)
Low Income x Log(HH income) −0.024
(0.021)
Low liquidi y x Log(HH income) −0.010
(0.027)
Sha e o emales in HH −0.110*** −0.133*** 0.060 −0.109*** −0.110***
(0.036) (0.036) (0.059) (0.036) (0.036)
Sha e o child en in HH 0.126*** 0.131*** 0.109** 0.128*** 0.126***
(0.030) (0.031) (0.052) (0.030) (0.030)
Sha e o income ea ne s in HH 0.007 −0.000 0.034 0.007 0.007
(0.018) (0.018) (0.034) (0.018) (0.018)
HH membe s 0.082*** 0.086*** 0.066*** 0.081*** 0.082***
(0.007) (0.007) (0.011) (0.007) (0.007)
Household ixed e ec s Yes Yes Yes Yes Yes
Yea ixed e ec s Yes Yes Yes Yes Yes
Supply side con ols No No No No No
Obse a ions 30,543 30,543 30,540 30,543 30,543
R20.903 0.906 0.862 0.903 0.903
Adjus ed R20.884 0.887 0.834 0.884 0.884
283
CPG consump ion in imeso ecession: no el e idence om…
po en ial he e ogenei y and p esen he esul s in Table7. Each column shows he
ou come o ou baseline analysis ( epea ed in column 1) when he eg ession is
applied o subsamples. Column 2 ocuses solely on ood ca ego ies, and column 3
on non- ood ca ego ies. The e ec size emains simila and hus economically mod-
es ac oss bo h ood and non- ood ca ego ies.
As sugges ed by mac oeconomic esea ch (e.g., Bake , 201818; Jappelli & Pis a-
e i, 2010), households unde igh e inancial cons ain s may eac mo e s ongly
o income changes. To examine i his applies o CPG consump ion in he Ne he -
lands, we classi y a household as low-income i i s de la ed s anda dized income
alls wi hin he lowes quin ile. Simila ly, we de ine a low-liquidi y household
based on low inancial asse s, using da a om CBS. Columns 4 and 5 o Table7
show esul s when in oducing an in e ac ion e m o low income o low liquidi y.
The s a is ically insigni ican coe icien s o he in e ac ion e m do no sugges a
s onge esponse among hese g oups. Al hough he con idence in e als a e b oad,
he alues emain modes e en a he uppe bounds o s anda d con idence in e als.
5 Conclusion
In his pape , we combine household scanne da a wi h adminis a i e income da a
and explo e CPG consump ion pa e ns o households in imes o ecession. Ou
key inding is an economically small ela ionship be ween household income and
expendi u es o CPGs.
Ou empi ical esul s es on a se ies o unique ins i u ional ac o s ha consid-
e ably mi iga e es ima ion challenges. Fi s , ou da a co e he second episode o
he Du ch double-dip ecession ollowing he inancial c isis which ended in 2018.
As no ed abo e, his ensu es ha a ound one- hi d o yea ly income changes a e
nega i e. Second, we documen na ional p icing and asso men s a egies in he
Ne he lands which allows o ule ou a se ies o supply side ac o s. We also docu-
men ha in he Ne he lands, households ha e compa able access o discoun e s o
s o e b ands and ha such o e s a e conside ably cheape . Thi d, adminis a i e da a
allows us o minimize measu emen e o o ou wo key a iables so ha we can
ule ou such conce ns as an explana ion o ou es ima ed small coe icien . Access
o adminis a i e da a also enables us o lea n abou households’ inancial sa -
ings, so we can explo e whe he households espond mo e p onouncedly o income
changes when hey a e mo e likely o be inancially cons ained. Disagg ega e con-
sump ion da a in u n allows us o analyze e ec s wi hin ood and non- ood ca -
ego ies. Ye , ou key inding o an economically small ela ionship emains obus
ac oss a ious model speci ica ions and sample spli s ha ca e o hese ac o s. In
sum, we a e con iden ha he (albei small) es ima ed coe icien indeed cap u es
beha io al changes.
Despi e his, we show also ha households’ (pe -capi a) CPG expendi u es a y
widely, e en wi hin income b acke s despi e con olling o common o e - ime
18 F om a ma ke ing science pe spec i e, see, o ins ance, O hun and Palazzolo (2019).
284
C.B anca elli, R.Inde s
a ia ion ia yea ixed e ec s. Households’ limi ed eac ion o nega i e income
shocks may hus be la gely due o consump ion p e e ences, which may no ha e
been su icien ly “ es ed” in he analyzed ecession en i onmen . This suppo s ea -
lie esea ch by B anca elli e al. (2022), which documen s ha in he Ne he lands
also he ela ionship be ween income and households’ consump ion o s o e b ands
is economically small. While no epo ed he e, we ind such a small ela ionship
also o households’ p opensi y o equen discoun e s (c . Biswas e al., 2021 o an
analysis o households’ choice o s o e o ma s in he US).
Casual e idence om ma ke esea ch sugges s ha he ecen su ge in in la ion,
which no only a ec s ( ela i e) p ices bu also eal income and weal h, may ha e
had a la ge impac and may indeed ha e il ed CPG choices. Con as ing his epi-
sode wi h ha o he ecession documen ed in his pape (and wi h i ou inding o
only an economically small adjus men o households) could be an a enue o u u e
esea ch wi h he ma ched da a.19
Finally, ou da a enables a compa ison be ween indings based on adminis a i e
income da a and hose using sel - epo ed income da a. As discussed ea lie , B anca-
elli e al. (2022) ind ha , o households’ p i a e-label sha e, he income coe icien
was signi ican ly a enua ed when elying on sel - epo ed da a. Tha s udy explo ed
po en ial easons o his a enua ion, including he income b acke ing applied by
bo h G K and Nielsen. By le e aging bo h sel - epo ed and adminis a i e income
da a, we e i ied ha indings om Dubé e al. (2018), which used Nielsen sel -
epo ed income, a e no solely due o measu emen e o in sel - epo ed income.
The ela ionship be ween income and households’ p i a e-label sha e emains eco-
nomically modes ega dless o da a ype.
Ou cu en analysis co obo a es his o o al CPG consump ion, showing ha
he ela ionship be ween income and consump ion emains modes bo h wi h admin-
is a i e and sel - epo ed income da a. While he small e ec sizes and ela i ely
b oad con idence in e als p e en us om examining in dep h why one income
measu e migh p oduce a mo e p onounced ela ionship han he o he , we p opose
ha u u e esea ch should add ess his ques ion. Such in es iga ion could enhance
ou unde s anding o he limi a ions associa ed wi h sel - epo ed income da a.
Finally, ou da a allows us o con as indings wi h adminis a i e income da a o
hose wi h sel - epo ed income da a. As discussed abo e, B anca elli e al. (2022)
ind ha o households’ p i a e-label sha e, he coe icien was much a enua ed
when elying on sel - epo ed da a. The pape ied o disen angle a ious easons
o his, including he b acke ing o income epo s, as p ac iced by bo h G K and
Nielsen. Using bo h sel - epo ed and adminis a i e da a allowed us o con i m
19 The ma ching o adminis a i e and household scanne da a o he p esen analysis ended in 2018,
gi en also he ocus on he Ne he lands’ pas ecession. In con as , while CPG supply was se e ely dis-
up ed in he i s phase o he COVID-19 c isis, con a y o o he coun ies income ha was no de i ed
om sel -employmen was la gely una ec ed in he Ne he lands, gi en ex ensi e public suppo p o-
g ams o i ms and employees.
285
CPG consump ion in imeso ecession: no el e idence om…
ha he indings in Dubé e al. (2018), which uses Nielsen sel - epo ed income, a e
no due o measu emen e o s in sel - epo ed income: The ela ionship be ween
income and households’ p i a e-label sha e emains economically small in ei he
case. Ou p esen analysis con i ms his o o al CPG consump ion, as again he
ela ionship emains economically small bo h wi h adminis a i e income da a and
wi h sel - epo ed income. While he small e ec s, oge he wi h ela i ely la ge
con idence in e als, do no allow us o dig deepe in o whe he and why he ela-
ionship wi h one measu e o income is mo e p onounced, we sugges add essing
his issue in u u e wo k. This should allow a be e unde s anding o po en ial limi-
a ions when wo king wi h sel - epo ed income.
Appendix
Sociodemog aphic cha ac e is ics: O e all G K da a s. eg ession da a
We epo key summa y s a is ics o panelis s in G K da a compa ed o only hose
who consen ed o he ma ching wi h CBS.
Table8
Table 8 Summa y s a is ics o panelis s in G K compa ed o consen ing households in 2011
Table8 shows key summa y s a is ics o he yea 2011 o sociodemog aphic a iables ob ained om
G K da a o bo h all G K households and households consen ing o he ma ching wi h CBS. Panel A
shows summa y s a is ics o age, panel B shows summa y s a is ics o household size, panel C ocuses
on social class, and panel D on he income o he household head. The able is based on appendix
Table14 in B anca elli e al. (2022)
Panel A: Age Panel B: Household
size
Panel C: Social class Panel D: Income
class
All Consen All Consen All Consen All Consen
Mean 6.94 7.40 2.49 2.40 4.53 4.61 14.35 14.72
SD 3.02 2.53 1.24 1.20 1.64 1.61 6.25 6.00
P5 1.00 1.00 1.00 1.00 1.00 1.00 3.00 3.00
P25 6.00 6.00 2.00 2.00 4.00 4.00 9.00 11.00
P50 7.00 8.00 2.00 2.00 5.00 5.00 16.00 16.00
P75 9.00 9.00 4.00 3.00 6.00 6.00 20.00 20.00
P90 11.00 11.00 4.00 4.00 6.00 6.00 21.00 21.00
P95 11.00 11.00 5.00 5.00 6.00 6.00 22.00 22.00
286
C.B anca elli, R.Inde s
Na ional p icing
We desc ibe how e aile s in he Ne he lands p ac ice na ional p icing. To
illus a e ha na ional p icing exis ed h oughou he en i e pe iod o Du ch
e aile s, we analyze p ice a ia ion ac oss egions o six majo e aile s in he
Ne he lands.
Fo his pu pose, we emo e p oduc s wi h p ices ha a e ou old highe han
he median o smalle han a qua e o he median p ice. Fu he mo e, we also
dis ega d ba codes a i icially in oduced by G K o classi y unpackaged esh
ood p oduc s, which usually do no ha e a GTIN. F om his se o p oduc s,
we iden i y and keep he 500 mos popula p oduc s pe e aile pe yea and
sepa a ely es ima e a p ice eg ession o each e aile and yea om 2011 o 2018.
In ou eg ession, he dependen a iable is he log o a household shopping ip-
speci ic p ice o any gi en p oduc . The independen a iables comp ise a se o
e aile -speci ic ba code dummies, mon hly ime ixed e ec s, and h ee egional
ixed e ec s.20 Th oughou he analysis, we use No h Ne he lands as he basis.
O e all, we es ima e p ice eg essions o six e aile s and eigh yea s21 and ex ac
he se o egional ixed e ec s o each eg ession.
Summa y s a is ics a e ep esen ed in Table9, showing ha ac oss all e aile s
and yea s, he ixed e ec s a e conside ably small, wi h mo e han 90% being
smalle han 1%.
We inally no e ha e en i a gi en e aile p ac iced egional p ice di e en ia ion
no cap u ed by his app oach, his would no isk con ounding ou in e p e a ion o
he income coe icien as long as p ice di e ences do no change o e ime.
Table 9 Summa y s a is ics o egional ixed e ec s ac oss all yea s and e aile s
Table9 depic s summa y s a is ics o egional ixed e ec s om independen yea ly eg essions o six
majo e aile s based on he household scanne da a o he G K consume panel be ween 2011 and 2018.
The able is based on Table4 in B anca elli e al. (2022)
Mean SD Skew P5 P10 P25 Median P75 P90 P95 N
Regional FE 0.0017 0.0042 1.1826 -0.0031 -0.0022 -0.0009 0.0008 0.0031 0.0078 0.010 132
20 Household scanne da a do no con ain geo in o ma ion abou indi idual shopping ips. The e o e,
we ely on household loca ion (pos code) o cons uc egional ixed e ec s. Regions a e de i ed om
he o icial NUTS 1 codes o he Ne he lands: No h Ne he lands, Eas Ne he lands, Wes Ne he lands,
Sou h Ne he lands.
21 In sum, we es ima e 44 eg ession speci ica ions: Un il 2014, we es ima e he eg ession o six majo
e aile s. Since wo majo e aile s me ged in 2014, we only es ima e he eg ession o i e e aile s om
2015 onwa d.
287
CPG consump ion in imeso ecession: no el e idence om…
Socio‑demog aphics o households in eg ession da a
Table10
Table 10 Summa y s a is ics o key socio-demog aphic a iables
Table10 shows summa y s a is ics o bo h he analysis sample and he popula ion o he ollowing socio-demog aphic a iables o he yea s 2011 and 2018: disposable
income pe s anda dized household membe , inancial asse s pe s anda dized household membe , he numbe o household membe s, he numbe o emales, he numbe
o child en and he numbe o income ea ne s in a household. Columns abb e ia ed wi h S. show he co esponding s a is ics o he sample and columns abb e ia ed
wi h Pop. Show co esponding popula ion igu es No e ha household size o disposable income and inancial asse s is s anda dized using an OECD scale (de ined as
1 + 0.7(n- 1) + 0.3k, whe e n is he numbe o adul s and k he numbe o child en. Eu o alues a e exp essed in cons an 2018 Eu o
Yea Label m—S m—Pop p10—S p10—Pop p25—S p25—Pop p50—S p50—Pop p75—S p75—Pop p90—S p90—Pop
2011 Disposable income
p.c
23,379.33 23,532.35 14,736.30 12,164.27 17,660.22 15,925.80 21,993.73 21,001.24 27,269.93 27,768.59 33,986.74 36,471.79
2018 Disposable income
p.c
24,037.68 24,852.26 14,683.00 12,513.91 17,795.22 16,552.00 22,571.00 22,349.00 28,410.00 29,635.88 35,008.50 38,164.17
2011 Financial asse s p.c 29,624.24 37,910.28 1,060.96 344.06 3,333.90 1,943.71 11,589.33 9,032.25 29,338.02 26,900.76 66,394.65 76,083.19
2018 Financial asse s p.c 27,996.16 36,221.45 715.00 346.15 2,677.65 1,790.77 9,814.00 8,731.18 25,378.06 26,476.00 58,580.00 70,196.00
2011 Household mem-
be s
2.46 2.18 1.00 1.00 2.00 1.00 2.00 2.00 3.00 3.00 4.00 4.00
2018 Household mem-
be s
2.40 2.12 1.00 1.00 1.00 1.00 2.00 2.00 3.00 3.00 4.00 4.00
2011 Numbe o child en 0.49 0.43 0.00 0.00 0.00 0.00 0.00 0.00 1.00 0.00 2.00 2.00
2018 Numbe o child en 0.49 0.39 0.00 0.00 0.00 0.00 0.00 0.00 1.00 0.00 2.00 2.00
2011 Numbe o emales 1.26 1.10 1.00 0.00 1.00 1.00 1.00 1.00 2.00 1.00 2.00 2.00
2018 Numbe o emales 1.25 1.07 1.00 0.00 1.00 1.00 1.00 1.00 2.00 1.00 2.00 2.00
2011 Numbe o inc.
ea ne s
1.94 1.72 1.00 1.00 1.00 1.00 2.00 2.00 2.00 2.00 3.00 3.00
2018 Numbe o inc.
ea ne s
1.92 1.73 1.00 1.00 1.00 1.00 2.00 2.00 2.00 2.00 3.00 3.00
288
C.B anca elli, R.Inde s
Funding Open Access unding enabled and o ganized by P ojek DEAL. One au ho was pa ially unded
by a long- e m esea ch g an om he Think Fo wa d Ini ia i e (TFI, g an 3902030019).
Decla a ions
Compe ing in e es s None o he au ho s has compe ing in e es s.
Open Access This a icle is licensed unde a C ea i e Commons A ibu ion 4.0 In e na ional License,
which pe mi s use, sha ing, adap a ion, dis ibu ion and ep oduc ion in any medium o o ma , as long
as you gi e app op ia e c edi o he o iginal au ho (s) and he sou ce, p o ide a link o he C ea i e
Commons licence, and indica e i changes we e made. The images o o he hi d pa y ma e ial in his
a icle a e included in he a icle’s C ea i e Commons licence, unless indica ed o he wise in a c edi line
o he ma e ial. I ma e ial is no included in he a icle’s C ea i e Commons licence and you in ended
use is no pe mi ed by s a u o y egula ion o exceeds he pe mi ed use, you will need o ob ain pe mis-
sion di ec ly om he copy igh holde . To iew a copy o his licence, isi h p://c ea i ecommons.o g/
licenses/by/4.0/.
Re e ences
Allco , H., Diamond, R., Dubé, J. P., Handbu y, J., Rahko sky, I., & Schnell, M. (2019). Food dese s
and he causes o nu i ional inequali y. The Qua e ly Jou nal o Economics, 134(4), 1793–1844.
A anasio, O. P., & Pis a e i, L. (2016). Consump ion inequali y. Jou nal o Economic Pe spec i es,
30(2), 3–28.
Bake , S. R. (2018). Deb and he esponse o household income shocks: Valida ion and applica ion o
linked inancial accoun da a. Jou nal o Poli ical Economy, 126(4), 1504–1557.
Bake , S. R., & Yannelis, C. (2017). Income changes and consump ion: E idence om he 2013 ede al
go e nmen shu down. Re iew o Economic Dynamics, 23, 99–124.
Banks, J., Blundell, R., & Lewbel, A. (1997). Quad a ic Engel cu es and consume demand. Re iew o
Economics and S a is ics, 79(4), 527–539.
Be hlehem, J., & Bi ignandi, S. (2011). Handbook o web su eys. Wiley.
Biswas, S., Chin agun a, P.K., & Dha , S.K. (2021). Do households’ budge alloca ions a y wi h eco-
nomic ac o s? E idence om Nielsen da a.
B anca elli, C., F i zsche, A., Inde s , R., & O e , T. (2022). Measu ing income and weal h e ec s on
p i a e-label demand wi h ma ched adminis a i e da a. Ma ke ing Science, 41(3), 637–656.
Ch is elis, D., Geo ga akos, D., Jappelli, T., Pis a e i, L., & Van Rooij, M. (2019). Asymme ic con-
sump ion e ec s o ansi o y income shocks. The Economic Jou nal, 129(622), 2322–2341.
De Meije , C., O’Donnell, O., Koopmanschap, M., & Van Doo slae , E. (2013). Heal h expendi u e
g ow h: Looking beyond he a e age h ough decomposi ion o he ull dis ibu ion. Jou nal o
Heal h Economics, 32(1), 88–105.
Deele snyde , B., Dekimpe, M. G., Sa a y, M., & Pa ke , P. M. (2004). Wea he ing igh economic imes:
The sales e olu ion o consume du ables o e he business cycle. Quan i a i e Ma ke ing and Eco-
nomics, 2(4), 347–383.
Dekimpe, M. G., & Delee snyde , B. (2018). Business cycle esea ch in ma ke ing: A e iew and esea ch
agenda. Jou nal o he Academy o Ma ke ing Science, 46(1), 31–58.
Dubé, J. P., Hi sch, G. J., & Rossi, P. E. (2018). Income and weal h e ec s on p i a e-label demand: E i-
dence om he G ea Recession. Ma ke ing Science, 37(1), 22–53.
Dubois, P., G i i h, R., & O’Connell, M. (2022). The use o scanne da a o economics esea ch. Annual
Re iew o Economics, 14, 723–745.
Du , P., & Padmanabhan, V. (2011). C isis and consump ion smoo hing. Ma ke ing Science, 30(3),
491–512.
IPLC (2016). Re aining consume s emp ed by he discoun model (Resea ch Repo ).
IRI (2016). Special epo : P i a e label in wes e n economies (Resea ch Repo ).
Jappelli, T., & Pis a e i, L. (2010). The Consump ion esponse o income changes. Annual Re iew o .
Econonomics, 2, 479–506.
