Chau in, Juan Pablo; Rubião, Ra ael; Talamas Ma cos, Miguel Ángel
Wo king Pape
The unde coun ing o child-mo he bi hs
IDB Wo king Pape Se ies, No. IDB-WP-1665
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: Chau in, Juan Pablo; Rubião, Ra ael; Talamas Ma cos, Miguel Ángel (2025) : The
unde coun ing o child-mo he bi hs, IDB Wo king Pape Se ies, No. IDB-WP-1665, In e -Ame ican
De elopmen Bank (IDB), Washing on, DC,
h ps://doi.o g/10.18235/0013407
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/315931
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 Unde coun ing o Child-Mo he Bi hs
J
uan Pablo Chau in
Ra ael Rubião
Miguel Ángel Talamas Ma cos
WORKING PAPER No IDB-WP-1665
In e -
A
me ican De elopmen Bank
Depa men o Resea ch and Chie Economis
Feb ua y 2025
* In e -
A
me ican De elopmen Bank
** Uni e si y o Cali o nia, Los Angeles
The Unde coun ing o Child-Mo he Bi hs
J
uan Pablo Chau in*
Ra ael Rubião**
Miguel Ángel Talamas Ma cos*
In e -
A
me ican De elopmen Bank
Depa men o Resea ch and Chie Economis
Feb ua y 2025
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
Chau in, Juan Pablo.
The unde coun ing o child-mo he bi hs / Juan Pablo Chau in, Ra ael
Rubião, Miguel Ángel Talamas Ma cos.
p. cm. — (IDB Wo king Pape s Se ies ; 1665)
Includes bibliog aphic e e ences.
1. P egnan eenage s-Census-B azil. 2. P egnan eenage s-Census-Mexico.
3. P egnan eenage s-Census-Uni ed S a es. 4. Census-Cos s-B azil. 5.
Census-Cos s- Mexico. 6. Census-Cos s-Uni ed S a es. 7. Popula ion-
S a is ics-B azil. 8. Popula ion-S a is ics-Mexico. 9. Popula ion-S a is ics-
Uni ed S a es. I. Rubião, Ra ael. II. Talamas, Miguel. III. In e -Ame ican
De elopmen Bank. Depa men o Resea ch and Chie Economis . IV. Ti le.
V. Se ies.
IDB-WP-1665
h p://www.iadb.o g
Copy igh ©2025 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 ∗
Accu a e demog aphic da a a e essen ial o e ec i e policy design, ye p i a e cos s
may de e indi iduals om u h ully epo ing sensi i e in o ma ion. We examine
his ma ke ailu e and i s implica ions in he con ex o child mo he hood. Using
adminis a i e eco ds om B azil, Mexico, and he Uni ed S a es, along wi h census
da a om 59 coun ies, we iden i y sys ema ic pa e ns o unde - epo ing, indica ing
ha child mo he hood is signi ican ly mo e p e alen han p e iously hough . Bi hs
o mo he s aged 10-14 a e o en missing om con empo a y adminis a i e eco ds bu
appea in censuses conduc ed a decade la e , wi h unde -coun ing in bi h egis ies
eaching 20-30% in B azil, Mexico, and he Uni ed S a es. We in oduce a model whe e
epo ing decisions weigh ins umen al bene i s agains age-dependen p i a e cos s,
yielding p edic ions ha align wi h obse ed pa e ns: u h ul epo ing inc eases
ma kedly wi h he mo he ’s age, unde - epo ing o child-mo he bi hs dec eases wi h
he ime elapsed be ween da a collec ion and childbi h, and e ospec i e census es i-
ma es gene ally p o ide mo e accu a e bi h coun s han con empo a y adminis a i e
eco ds o his age g oup, bu no o olde mo he s. Ou indings sugges ha social
cos s, a he han ea o legal epe cussions, a e a p ima y d i e o unde - epo ing.
JEL classi ica ions: J13, D10, J18, D82
Keywo ds: child mo he hood, p i a e in o ma ion, social no ms.
∗
Chau in: In e -Ame ican De elopmen Bank: [email p o ec ed]. Rubião: UCLA Ande -
son: a ael. [email p o ec ed]. Talamas Ma cos: In e -Ame ican De elopmen Bank:
[email p o ec ed]. We hank Gio anna Cha es and Juan Fe e o hei excellen esea ch assis ance,
Li ia Almeida, Michael Ca dona, Sh u hi Nagamani, and Vinícius P incio i o addi ional esea ch suppo ,
and wo anonymous e iewe s o use ul commen s. The opinions exp essed in his publica ion 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 .
1
1 In oduc ion
Accu a e demog aphic and heal h da a a e a aluable public good c i ical o e ec i e esou ce alloca ion,
in o med policy esponses, and ad ocacy ha add esses social p oblems. P oducing his public good elies
on ins i u ional capaci y and, a imes, hea ily on indi iduals u h ully epo ing in o ma ion. Howe e ,
when epo ing incu s p i a e cos s, such as legal exposu e o social s igma, indi iduals may choose o
wi hhold in o ma ion o epo i inaccu a ely, imposing a nega i e ex e nali y on socie y by educing he
o e all a ailabili y and quali y o his public good (Blank e al.,2009;Du lo e al.,2013).
We s udy his asymme ic in o ma ion ma ke ailu e in he con ex o an impo an social issue: child-
mo he bi hs. Ex ensi e esea ch shows ha childbea ing a a young age ha ms he well-being o bo h
mo he and child (e.g., Du lo 2012;Kea ney and Le ine 2015;Lang and Weins ein 2015). This is especially
ue o younge mo he s (Aize e al.,2022), whose p egnancies a e o en unwan ed and a esul o sexual
iolence. Hence, accu a e da a on he p e alence o child mo he hood is essen ial o unde s anding he
magni ude o he issue a hand and keeping ack o i s e olu ion. Howe e , his age g oup, o which
accu a e da a a e essen ial, is also likely he g oup o which he asymme ic in o ma ion p oblem is he
la ges because social o legal epe cussions end o in ensi y he younge he mo he is a childbi h.
To guide ou analysis, we p opose a simple amewo k o he decision o epo a bi h wi h wo epo ing
me hods: adminis a i e eco ds and census in e iews. Mo he s can epo bi hs using ei he o bo h o he
me hods o choose no o epo . Thei decision depends on whe he he bene i s o epo ing ou weigh he
associa ed cos s. Young mo he s, in addi ion o acing ins umen al cos s (e.g., ees, a el, o ime equi ed
o epo ), may also encoun e "social cos s" (such as s igma iza ion o os acism) and "judicial cos s" (i he
sexual encoun e leading o he p egnancy is conside ed a c ime).1These social and judicial cos s dec ease
as he mo he ages and mo e ime passes be ween he bi h and he mo he ’s age a epo ing.
Mo i a ed by he es able p edic ions o ou amewo k, we in es iga e he magni ude and d i e s o
he unde - epo ing o child mo he hood in adminis a i e eco ds and censuses om o e 50 coun ies.
We begin by es ima ing he ex en o unde - epo ing o child-mo he bi hs in adminis a i e eco ds in
h ee la ge coun ies: B azil, Mexico, and he Uni ed S a es. These coun ies ha e signi ican popula ion
sizes and he a ailabili y o comp ehensi e adminis a i e and census da a. By compa ing bi h coun s om
adminis a i e eco ds wi h hose econs uc ed om census da a collec ed up o 15 yea s a e childbi h,
1
E en i di ec penal ies a ec only he a he , he child-mo he may expe ience a di ec disu ili y om his
legal exposu e, ei he because hey sha e an emo ional, inancial, o amilial bond wi h he a he , because
hey a e ulne able o e bal o physical e alia ion om him, o because going h ough his legal p ocess
gene a es s ess and su e ing o he ic im.
2
we can iden i y pa e ns o unde - epo ing ac oss di e en age g oups a he ime o childbi h.
In ou analysis, we expec ha , in he absence o unde coun ing in bi h egis ies, he numbe o bi hs
eco ded in hose egis ies will exceed he bi hs in e ed om subsequen censuses because census coun s
depend on child en li ing wi h hei mo he s, which is no always he case and becomes less common as
he child en g ow olde . This p edic ion holds o mo he s aged 15 o olde a childbi h. Howe e , o
mo he s aged 14 o younge a childbi h, we ind ha in all h ee coun ies, he e a e mo e bi hs based on
he es ima es om he census 10 yea s a e childbi h han in bi h egis ies. These gaps, which ep esen
lowe bounds o unde - epo ing in bi h egis ies, a e subs an ial: 20% in B azil (1999), 22% in Mexico
(2009), and 29% in he Uni ed S a es (2009).
Nex , we explo e how child-mo he bi h epo ing changes wi h he mo he ’s age a he ime o epo ing.
We le e age wo complemen a y sou ces o a ia ion. Fi s , we compa e bi h coun s o he same coho
o mo he s eco e ed om subsequen census ounds. By compa ing epo ing o he same bi hs ac oss
di e en census ounds, we hold he epo ing me hod cons an while a ying only he mo he ’s age a
epo ing. This compa ison e eals ha while bi h coun s o een and young-adul mo he s (aged 15-24 a
deli e y) a e simila ac oss census ounds, coun s o child mo he s a e subs an ially lowe in he baseline
census—when mo he s a e s ill in hei eens—compa ed o a decade la e when hey a e in hei wen ies.
Figu e 1: Es ima ed Bi h Coun s Based on Subsequen Censuses
No es: The sample comp ises 80 pai s o na ional censuses om 59 coun ies, wi h one obse a ion pe census pai , age o he mo he a
bi h and he yea o bi h. We es ima e he numbe o bi hs by iden i ying he age a bi h o he epo ed mo he . Samples exclude
child en o o eign ci izenship.
Figu e 1displays his s iking pa e n using 78 census pai s om 59 coun ies: while bi h coun es ima es
o he same coho o een and young-adul mo he s a e compa able ac oss subsequen censuses (le side),
3
bi hs o mo he s aged 14 o younge ( igh side) show sys ema ic unde - epo ing in he baseline censuses
wo ldwide. This e idence is consis en wi h households no acknowledging he bi h o a mo he 10 o 14
yea s old i he mo he is s ill in he eens a he ime o he census bu acknowledging i once he mo he is
in he wen ies a he ime o he census. This mis epo ing could ake se e al o ms, o example, omi ing
he p esence o he child a he ime o he in e iew o a ibu ing he ma e ni y o he child o ano he
household membe , o which we ind e idence and la e discuss.
The magni ude o unde - epo ing is in e sely ela ed o he mo he ’s age a bi h: baseline censuses
miss app oxima ely 65% o bi hs o mo he s aged 13 o younge , 50% o bi hs o 14-yea -old mo he s, and
22% o bi hs o 15-yea -old mo he s when compa ed o coun s based on subsequen censuses. This pa e n
is ema kably consis en ac oss coun ies and mos p onounced o bi hs occu ing in he yea immedia ely
p eceding he census, wi h he epo ing gap diminishing as he ime be ween bi h and census inc eases.
These indings sugges ha p e ious es ima es o unde - egis a ion o bi hs—based on he possession
o bi h ce i ica es among child en epo ed in su eys and censuses occu ing wi hin a ew yea s om
childbi h—may signi ican ly unde s a e he ue magni ude o he p e alence o child mo he hood.
To in es iga e u he how child-mo he bi h epo ing changes wi h he mo he ’s age a he ime o
epo ing, we used Mexican adminis a i e da a om 1995 o 2020, whe e delayed in o ma ion is a ailable.
We ind ha , while bi hs o mo he s aged 15-24 a e ypically egis e ed wi hin a yea (80-90% o cases), bi hs
o mo he s aged 10-14 show signi ican ly longe delays, wi h only 50-80% egis e ing wi hin a yea o he
bi h. The a e age egis a ion delay dec eases sha ply wi h he mo he ’s age, om 40 mon hs o 12-yea -old
mo he s o 10 mon hs o 15-yea -old mo he s. Bo h pa e ns a e consis en wi h epo ing cos s—whe he
social o judicial—declining as mo he s age.
Finally, we in es iga e h ee key mechanisms unde lying he obse ed unde - epo ing pa e ns. Fi s , we
ind e idence ha many bi hs o child mo he s a e no en i ely concealed bu a he misa ibu ed o olde
women in he household, pa icula ly hose beyond ep oduc i e age. Compa ing bi h coun s ac oss census
ounds e eals ha women aged 45-55 epo 80-200% mo e bi hs in baseline censuses han in subsequen
ones—a gap a exceeding wha mo ali y a es alone could explain. This sugges s amilies o en epo hese
bi hs bu a ibu e hem o g andmo he s o aun s.
Second, o assess he impo ance o judicial cos s in he epo ing decision, we exploi a ia ion in
minimum age o consen laws ac oss coun ies and he iming o he p egnancy ela i e o he mo he
eaching he minimum age o consen . Using a eg ession discon inui y design, we ind no e idence o
discon inuous inc eases in bi h epo ing when mo he s each he age o legal consen , sugges ing ha ea
o legal consequences o he abuse s may no be a subs an ial d i e o unde - epo ing.
4
Thi d, we examine how unde - epo ing a ies wi h social no ms su ounding emale sexuali y and
ep oduc i e igh s ac oss coun ies. Using da a om UN Demog aphic and Heal h Su eys on women’s
au onomy in sexual decision-making, we ind ha , con olling o income le els, a one s anda d de ia ion
inc ease in he sha e o women epo ing au onomous sexual decisions is associa ed wi h a 20% educ ion in
unde - epo ing o child-mo he bi hs.
This pape con ibu es o h ee bodies o esea ch. Fi s , we make a con ibu ion o he li e a u e on
child and eenage mo he hood. P io s udies ha e examined he causes and consequences o e ili y among
eenage gi ls, showing ha i is associa ed wi h nega i e ou comes o bo h mo he s and hei child en
(Co bacho e al.,2012;Du lo,2012;Kea ney and Le ine,2015;Klepinge e al.,1999;Lang and Weins ein,
2015). This social p oblem has a ac ed conside able policy a en ion, and he Uni ed Na ions included
educing e ili y a es among gi ls aged 10-19 as pa o i s Sus ainable De elopmen Goals (SDGs). Howe e ,
bo h he academic li e a u e and da a used o policy p ima ily ocus on e ili y among gi ls aged 15-19,
o whom eliable s a is ics a e widely a ailable. In con as , e ili y da a o gi ls younge han 15 emain
sca ce and incomple e (Schoumake and Sánchez-Páez,2022;WHO,2024). We con ibu e by p oducing
lowe -bound es ima es o unde - epo ing o 59 coun ies, demons a ing ha his phenomenon is mo e
widesp ead and mo e p ominen han p e iously hough , also a ec ing high- and middle-income coun ies.
Impo an ly, we show ha unde - epo ing occu s no only in census su eys bu also in bi h egis ies,
which a e o en conside ed he mos eliable da a sou ce o bi h coun s (Kea ney and Le ine,2012).
Second, we ex end p io esea ch on he economics o in o ma ion and public goods. P e ious s udies
ha e shown ha p i a e cos s can p e en indi iduals om u h ully disclosing socially aluable in o ma ion
in a a ie y o con ex s, such as audi o s epo ing i ms’ pollu ion le els (Du lo e al.,2013), schools epo ing
s uden en ollmen o he go e nmen (Sande u and Glassman,2015), and sel -employed indi iduals
decla ing hei income o ax au ho i ies (Hu s and Pugsley,2010). In he con ex o i al s a is ics, Blank
e al. (2009) show ha when ea ly ma iage is ou lawed, indi iduals ha e incen i es o mis epo , leading o
adminis a i e eco ds being an in e io da a sou ce compa ed o e ospec i e census da a o s udying he
e ec s o age-o -ma iage laws. Ou s udy inds ha adminis a i e eco ds can unde - epo child-mo he
bi hs, e en in he absence o legal epe cussions, when indi iduals ace non-pecunia y cos s s emming om
in o mal no ms and social p essu e.
Finally, we con ibu e o he li e a u e on social no ms. P io esea ch has shown ha conce ns o
epu a ion and social image can be as in luen ial as ma e ial incen i es (Bénabou and Ti ole,2006;Bu e a e al.,
2022;Daughe y and Reinganum,2009). Social p essu e can encou age socially bene icial beha io in con ex s
such as o ing (Ali and Lin,2013), wo kplace sa e y (Johnson,2020), sa ing (B eza and Chand asekha ,
5
wi h hei mo he s, as ques ions abou non- esiden child en a e a ely included. To mi iga e his conce n,
we es ima e bi hs only o g oups whe e he child would be a mos 15 yea s old a he ime o he second
measu emen , inc easing he likelihood ha he child is s ill li ing wi h he mo he . Howe e , e en wi h his
adjus men , he second measu emen o bi hs, aken yea s la e , is likely o be lowe due o olde child en
being less likely o eside wi h hei mo he s. Fac o s such as dea h, mig a ion, and ma iage dissolu ion also
con ibu e o his decline. As a esul , ou es ima es o bi hs unde - epo ing ep esen a lowe bound o he
ac ual unde - epo ing.
4 Es ima ion
We u n now o he empi ical analysis o child-mo he s’ bi h epo ing. Ou goal is o gauge he magni ude
o unde - epo ing and in es iga e, in ligh o ou model, how he mo he ’s age a bi h and a epo ing
in luences epo ing beha io .
4.1
Unde - epo ing o Child-Mo he Bi hs in Adminis a i e Reco ds
We begin by es ima ing he ex en o unde - epo ing in adminis a i e bi h eco ds o child mo he s (ages
10 o 14) in B azil, Mexico, and he Uni ed S a es— h ee la ge coun ies whe e he co e age o ci il egis a ion
and i al s a is ics (CRVS) is conside ed comple e o nea ly comple e (Kisambi a and Schmid,2022). We
do his by compa ing adminis a i e bi h coun s wi h bi h coun s ob ained om census da a collec ed 11
yea s la e .
7
Table 1 epo s he coun s om adminis a i e eco ds (column 1) and he subsequen census
es ima es (column 2). I also includes ou es ima e o un eco ded bi hs as a pe cen age o he census coun s
(column 3) o wo g oups: child mo he s (ages 10 o 14) and een and young adul mo he s (ages 15 o 24).
8
A posi i e alue in column 3 indica es unde - epo ing in adminis a i e eco ds ela i e o he census. In
con as , a nega i e alue sugges s ha adminis a i e eco ds exceed he census es ima es, which would be
expec ed in he absence o social o judicial cos s associa ed wi h epo ing because no all child en cohabi
wi h hei bi h mo he s.
The able displays a pa e n o sizable unde - epo ing o child mo he s in he adminis a i e eco ds
7
Census coun s a e compu ed o he yea p io o he mos ecen a ailable census in each coun y: 2009
o B azil and 2019 o Mexico and he Uni ed S a es. This is because census da a do no cap u e bi hs
occu ing in he same yea a e he census collec ion da es. The base yea is chosen o be 11 yea s p io .
8
These g oupings ollow he s anda d age-b acke s classi ica ion used by he Uni ed Na ions and o he
p o ide s o demog aphic s a is ics (Uni ed Na ions DESA,2017)
12
Table 1: Unde - epo ing in Adminis a i e Reco ds Rela i e o Subsequen Census
(1) (2) (3)
Adminis a i e Subsequen Adminis a i e
Reco ds Census Unde epo ing
(2)−(1)
(2)
B azil (1999)
Child mo he s (ages 10 o 14) 27518 34372 0.20
Teen mo he s (ages 15 o 24) 1737520 1244020 -0.40
Mexico (2009)
Child mo he s (ages 10 o 14) 9739 12558 0.22
Teen mo he s (ages 15 o 24) 1136376 841056 -0.35
Uni ed S a es (2009)
Child mo he s (ages 10 o 14) 5038 7112 0.29
Teen mo he s (ages 15 o 24) 1417630 839953 -0.69
No e: The able p o ides es ima es o he numbe o bi hs by age g oup a he ime o deli e y. The admin-
is a i e coun s e lec all bi hs eco ded wi hin he i s yea a e deli e y, as cap u ed in each coun y’s
adminis a i e egis y. The "subsequen census" coun s a e de i ed om he ages o child en and hei
mo he s in he mic oda a om he census conduc ed 10 yea s la e . The pe cen age o un eco ded bi hs is
calcula ed as he di e ence be ween he adminis a i e and census coun s, exp essed as a p opo ion o he
census coun s.
ac oss he h ee coun ies. In B azil, he numbe o bi hs o child mo he s eco ded in adminis a i e da a is
20% lowe han hose cap u ed by he subsequen census. The es ima ed unde - epo ing is 22% in Mexico
and 29% in he Uni ed S a es. This con as s sha ply wi h he pa e n obse ed o een and young adul
mo he s (ages 15 o 24), whe e adminis a i e eco ds exceed census bi h coun s by subs an ial ma gins:
40% in B azil, 35% in Mexico, and 69% in he Uni ed S a es.
These pa e ns a e e iden in Figu e 2, which plo s he es ima ed bi h coun s by age g oup and yea
using census and adminis a i e eco ds da a o B azil, Mexico, and he Uni ed S a es. Fo eenage and
young adul mo he s, on he le side, he bi h coun s om he egis ies a e always conside ably highe
han in he censuses—as expec ed due o he census limi a ions discussed. Howe e , o child mo he s, on
he igh side o Figu e 2, he bi h coun s es ima ed using a census occu ing 10 o mo e yea s in he u u e
a e signi ican ly highe han ha o he bi h egis ies.
Fo example, based on he 2020 Mexican census, he e we e 14,000 bi hs o mo he s aged 10 o 14 in
2010; howe e , in he adminis a i e eco ds, his numbe is less han 10,000. As we ge close o he census
in e iew da e (mo ing o he igh on he x-axis), mo he s a e younge a he ime o epo ing, and he
13
(a) Bi hs in B azil by mo he age g oup (1985-2020)
(b) Bi hs in Mexico by mo he age g oup (1985-2020)
(c) Bi hs in he USA by mo he age g oup (1985-2019)
Figu e 2: To al Bi hs by Mo he ’s Age G oup Acco ding o Census and Regis y Da a
No e: The igu e p esen s annual bi h es ima es de i ed om na ional censuses and bi h egis ies o B azil, Mexico, and he Uni ed
S a es. The le panels ep esen mo he s aged 15 o 25, while he igh panels depic hose aged 10 o 14. Panel (a) illus a es B azilian
bi h es ima es, u ilizing impu ed da a om he 2000 and 2010 Censuses in conjunc ion wi h he B azil Li e Bi h In o ma ion Sys em
(SINASC). Panel (b) po ays Mexican bi h es ima es, employing da a om he 2010 and 2020 Censuses supplemen ed by he bi h
egis y. Panel (c) exhibi s U.S. bi h es ima es using da a om he 2000 Census, he Ame ican Communi y Su eys (IPUMS) o 2000,
2010, 2019, and he bi h egis y. The 2000 su ey andomly sampled 0.13% o he popula ion, while he 2010 and 2019 su eys sampled
1%. Da a om immig an child en in he Uni ed S a es and in o ma ion om 2020, po en ially a ec ed by COVID-19 pandemic- ela ed
epo ing delays, a e excluded.
14
unde - epo ing p oblem inc eases o a poin whe e he census eco e s a lowe es ima e han ha o he
egis ies. This gene a es a single-c ossing in he es ima es be ween he census and he bi h egis ies, which
is he p edic ion o p oposi ion 4. The main disad an age o his me hodology is ha he " ue" census coun
o a yea close o a census (e.g., 2019) will only be a ailable when he 2030 census is conduc ed.
These es ima es and pa e ns demons a e ha he unde - egis a ion o bi hs—a well-documen ed issue
a ec ing many coun ies, especially in he de eloping wo ld (Uni ed Na ions DESA,2020a)—is s ikingly
mo e p onounced among child mo he s. The ex en o his bias is g ea e han p e iously ecognized. Using
ques ions on egis a ion om he Demog aphic and Heal h Su ey (DHS) o i e La in Ame ican coun ies,
Du yea e al. (2006) ind ha he likelihood o child en unde i e lacking a bi h ce i ica e inc eases by 4.5
o 6.5 pe cen age poin s when he mo he is a eenage , ela i e o an a e age o 14.5%, in Boli ia, Colombia,
and Pe u (bu no in B azil o Nica agua). Simila ly, Ebbe s and Smi s (2022), using DHS da a om 40
Sub-Saha an A ican coun ies, ind ha child en bo n o mo he s younge han 18 ha e 6% lowe odds o
being egis e ed.
4.2 Mo he s’ Age and Repo ing
This sec ion explo es how he age o he mo he a childbi h and a he ime o epo ing a ec s he epo ing
beha io . We i s discuss e idence based on adminis a i e eco ds and hen e idence based on censuses.
Adminis a i e Reco ds E idence
I he social cos o epo ing dec eases wi h he mo he ’s age a childbi h and he ime be ween epo ing
and he bi h, i should in luence epo ing beha io in adminis a i e eco ds in wo ways. Fi s , i he
mo he can somewha con ol he iming o epo ing, we would expec delays. These delays would happen
because, a he ime o bi h, he social cos is so high ha he op imal choice is no o epo . Howe e , as
sugges ed in P oposi ion 2, as ime p og esses and his cos dec eases, i may become bene icial o epo .
Second, he du a ion o he epo ing delay is a unc ion o he social cos , which, acco ding o P oposi ion
1, declines wi h he mo he ’s age a childbi h. The e o e, we would expec ha , condi ional on e en ual
epo ing, his delay would be longe he younge he mo he was a he ime o bi h.
Among he adminis a i e egis ies conside ed in his s udy, only Mexico p o ides de ailed da a on
delayed egis a ion, allowing us o es ima e epo ing delays o di e en ma e nal age g oups and bi h
coho s. We summa ize hese da a in Figu e 3. Panel (a) shows he sha e o child en o mo he s aged 15-24
15
(a) Sha e o Bi hs Regis e ed by Time Span since Yea o Bi h
(b) A e age delay in egis a ion (1995-2019)
Figu e 3: Bi h Regis a ion Delays in Mexico, by Mo he ’s Age G oup
No es: This igu e uses bi h egis y da a om Mexico (INEGI). Panel A plo s he sha e o bi hs ha ook
place in each yea ha we e egis e ed up o Decembe 2019, by maximum delay allowed. The 1 (5, 10, 20)
yea (s) delay line includes bi hs occu ed om Janua y o Decembe o each yea , egis e ed up o 12 (60,
120, 240) mon hs om bi h. Fo he 1 (5, 10, 20) yea (s) delay line, we only plo da a up o he yea 2018
(2014, 2009, 1999), which co esponds o he pe iod ha we ully obse e da a o he nex 12 (60, 120, 240)
mon hs. Panel B plo s he a e age delay in bi h egis a ion ac oss he age o he mo he a bi h in Mexico
(1995-2019), condi ional on he child en being egis e ed up o 15 yea s a e bi h.
(le panel) and 10-14 ( igh panel) egis e ed wi hin one, i e, en, o wen y yea s o hei bi h da e.
9
We expec some le el o delay in all egis a ions due o cos s such as ees o anspo a ion. These delays
ha e dec eased o e ime as he Mexican go e nmen has expanded access o ci il egis y se ices and
incen i ized egis a ion. Fo mo he s aged 15-24, be ween 1995 and 2005, jus o e 80% o bi hs o his age
g oup we e egis e ed wi hin one yea , ising o 90% be ween 2010 and 2018).
9
Fo consis ency wi h ou 2019 da a cu o , bi hs wi h up o 1 yea delay a e shown only h ough 2018.
This ensu es all 2018 bi hs a e cap u ed wi hin he 1-yea window wi hou equi ing 2020 da a. Simila ly,
bi hs wi h up o 5, 10, and 20 yea s o delay a e shown h ough 2014, 2009, and 1999, espec i ely.
16
In con as , o child mo he s aged 10-14, egis e ing wi hin one yea was subs an ially less common,
anging be ween 50 and 60% in he ea lie pe iod (1995-2005), imp o ing o jus abo e 80% by 2018. Panel (b)
o Figu e 3illus a es how he a e age delay in bi h egis a ion dec eases wi h he mo he ’s age, d opping
om app oxima ely 40 mon hs o 12-yea -old mo he s o a ound 10 mon hs o 15-yea -old mo he s, a e
which he delay s abilizes. These indings align wi h he wo k o Keskin and Ça lin (2020) in Tu key, who
ound ha , while 98% o bi hs o adolescen mo he s we e e en ually egis e ed be ween 2011 and 2015,
only 78% we e egis e ed wi hin he i s 30 days, compa ed o 95% o all bi hs.
Census E idence: B azil, Mexico, and he Uni ed S a es
Census su eys di e om adminis a i e eco ds in e ms o hei a ailabili y as a epo ing mechanism.
While he op ion o epo a bi h o he egis ies is con inuously a ailable, he op ion o epo i in he
census is gene ally a ailable only once e e y decade, which could occu sho ly a e he child’s bi h o
se e al yea s la e . C ucially, he iming o he census is exogenous o he child’s bi h da e, he mo he ’s age
a deli e y, and he mo he ’s age a he ime o epo ing. Thus, we can isola e he e ec o he mo he ’s age
a he ime o epo ing by compa ing es ima es om wo di e en census ounds o he same mo he ’s age
a childbi h and yea , he eby holding he epo ing me hod and age a childbi h cons an .
10
This p o ides
an addi ional es o P oposi ion 2, which posi s ha unde - epo ing should be less se e e he olde he
mo he is a he ime o epo ing.
In line wi h his p oposi ion, we ind sha ply di e en pa e ns be ween he een and young-adul
mo he s g oup and child-mo he s. When we conside mo he s aged 15 o 24 a he ime o deli e y—a g oup
o which we expec social and judicial cos s o epo ing o be smalle o non-exis en — we obse e ha
bi h coun s based on p io and la e censuses a e simila in B azil, Mexico, and he Uni ed S a es (Figu e
2, le side). In con as , bi h coun s o child mo he s— o whom we expec highe social and/o judicial
cos s—a e consis en ly lowe in he baseline census (when he mo he is s ill in he eens a he ime o
epo ing) han in he census conduc ed a decade la e (when he mo he is al eady in he wen ies). The
gap is la ge. Rela i e o he subsequen coun s, he baseline coun s a e smalle by 40 o 90%.
10
As p e iously discussed, his app oach allows us o iden i y epo ing di e ences ne o a i ion be ween
he baseline and subsequen census ounds, p o iding a lowe bound o unde - epo ing in he baseline
census.
17
Census E idence: Global Pa e ns o Unde - epo ing o Child-Mo he Bi hs
The unde - epo ing o child-mo he bi hs in B azil, Mexico, and he Uni ed S a es, as documen ed abo e,
aises he ques ion o whe he simila pa e ns o mis epo ing a e p e alen in o he coun ies. To add ess
his, we now examine he ex en o unde - epo ing o child-mo he bi hs in censuses ac oss a b oade se
o coun ies, using da a om 78 census pai s co e ing 59 coun ies.
11
While p e ious s udies on child and
eenage e ili y ha e p ima ily ocused on o mal bi h egis a ion in adminis a i e eco ds (e.g., Du yea
e al. 2006;Ebbe s and Smi s 2022;Wend e al. 2022), unde - egis a ion in demog aphic censuses and
su eys is impo an in i s own igh . These da ase s play a c i ical ole in shaping policies o add ess he
issue and in p oducing esea ch ha in o ms policy decisions (Uni ed Na ions DESA,2020a).
Figu e 4: Bi h Coun s in Each o he Fi e Yea s p io o he Baseline Census
No es: The sample comp ises 80 pai s o na ional censuses om 59 coun ies, wi h one obse a ion pe census pai , age o he mo he a
bi h and he yea o bi h. We es ima e he numbe o bi hs by iden i ying he age a bi h o he epo ed mo he . Samples exclude
child en o o eign ci izenship.
Figu e 4p esen s he c oss-coun y a e ages and 95% con idence in e als o bi h coun s ( om bo h
baseline and subsequen censuses) o each o he i e yea s p eceding he baseline census.
12
Consis en
wi h ou indings o B azil, Mexico, and he Uni ed S a es, we obse e sys ema ic di e ences in bi h coun s
be ween he baseline and subsequen censuses o child mo he s ( igh panel), bu no o olde mo he s
(le panel). Fu he mo e, he unde coun ing o bi hs in he baseline census is mos p onounced o bi hs
11
Es ima es a e de i ed om census mic oda a in eg a ed and s anda dized by IPUMS (In eg a ed Public
Use Mic oda a Se ies, IPUMS [da ase , accessed Ma ch,2024). To ou knowledge, no compa able sou ce
exis s o in e na ional bi h egis y da a.
12
Appendix Table A1 epo s he unde lying da a, along wi h child-mo he bi h unde coun ing es ima es
by he mo he ’s age a bi h o each baseline census.
18
occu ing in he yea immedia ely p eceding he census and diminishes as he ime be ween he bi h and he
census inc eases. This p o ides b oade empi ical suppo o he model’s p edic ion ha he p opensi y o
epo bi hs by young mo he s inc eases wi h he mo he ’s age a epo ing, all else being equal (P oposi ion
2).
Finally, we explo e how unde - epo ing in he baseline census a ies by he mo he ’s age a bi h
ac oss ou global sample. Figu e 5p esen s c oss-coun y a e ages and 95% con idence in e als o he
gap be ween baseline and subsequen census coun s, exp essed as a sha e o he subsequen census coun ,
o each ma e nal age. Unde - epo ing o child-mo he bi hs in con empo a y censuses is widesp ead.
Ac oss coun ies, bi h coun s a e consis en ly highe o mo he s aged 15 o younge in subsequen censuses
compa ed o baseline censuses conduc ed soon a e he bi h. Only o mo he s aged 17 o olde do baseline
censuses yield he highe coun s ha a e mechanically expec ed. In line wi h P oposi ion 1, he likelihood o
epo ing inc eases mono onically wi h he mo he ’s age a childbi h.
Figu e 5: Gap in Repo ed Childbi hs be ween Consecu i e Censuses by Mo he ’s Age
No es: The sample comp ises 80 pai s o na ional censuses om 59 coun ies, wi h one obse a ion pe census pai , age o he mo he a
bi h and he yea o bi h. We es ima e he numbe o bi hs by iden i ying he age a bi h o he epo ed mo he . The gap measu e is
calcula ed as one minus he a io o he numbe o bi hs es ima ed om he i s census o hose es ima ed om he second census
1−bi hs
bi hs +10
, o e he age o he mo he a bi h. The in e al be ween censuses a e ages 10 yea s, bu i includes spans anging
om 8 o 12 yea s. Samples exclude child en o o eign ci izenship.
We ind subs an ial unde - epo ing o child-mo he bi hs, wi h baseline censuses missing mo e han
60% bi hs o mo he s aged 13 o younge ha a e eco ded in subsequen censuses. Fo mo he s aged 14,
his igu e is a ound 50%, and o hose aged 15, i is close o 20%.
Based on census and su ey da a ha iden i y he sha e o young child en wi hou bi h ce i ica es,
p e ious esea ch has shown e y low le els o bi h egis a ion in he de eloping wo ld, wi h only abou
one- hi d o child en unde i e egis e ed in Sou h Asia (Kisambi a and Schmid,2022), a ound 45% in
Sub-Saha an A ica (Uni ed Na ions DESA,2020a), and be ween 8% and 25% in La in Ame ica (Du yea e al.,
2006). Howe e , ou indings indica e ha census and su ey samples hemsel es a e missing many child en,
pa icula ly hose bo n o e y young mo he s. The e o e, he le el o unde - egis a ion in bi h egis ies is
19
likely signi ican ly highe han p e iously es ima ed o his popula ion. This is because a signi ican ac ion
o he o sp ing o young mo he s is likely absen om hese censuses and su eys, and hose missing a e
also he mos likely no o ha e a bi h ce i ica e.
4.3 Ins umen al Payo s and he P opensi y o Repo
The e idence in Figu e 2also o e s insigh s in o he di e ences in ins umen al payo s ac oss epo ing
me hods and how hese payo s compa e o social and judicial cos s as mo he s age. Acco ding o P oposi ion
3, i he mo he ’s age a he ime o childbi h and epo ing a e held cons an , epo ing should be highe in
he me hod ha p o ides he g ea e ne ins umen al payo . Ou da a can app oxima e hese condi ions
when he census occu s close o he child’s bi h da e.
In Figu e 2, we obse e ha o bi hs occu ing in he yea s p io bu close o he census yea s, and o bo h
age g oups and ac oss all h ee coun ies, he numbe o bi hs eco ded in he egis ies exceeds he census
coun s. As pe P oposi ion 3, his sugges s ha he ne payo s o epo ing a e highe o adminis a i e
eco ds han o he census. This is unsu p ising, as bi h egis ies ypically o e mo e di ec bene i s, such
as access o se ices and public subsidies, while he bene i s o epo ing o he census a e less immedia e
and di ec . Thus, i a bi h o a young mo he occu s close o he ime o he census, households ha epo
o he census a e likely o also epo o he egis y, bu hose epo ing in he egis y may no epo in he
census.
I social and judicial cos s a e absen o minimal, he pa e n desc ibed abo e should pe sis : epo ing
will always be highe wi h he me hod ha p o ides a highe ne ins umen al payo . This is wha we
obse e in he h ee coun ies o mo he s aged 15 o 24 (Figu e 2, le column), whe e he numbe o bi hs
in he egis ies consis en ly exceeds hose es ima ed by he censuses.
13
Howe e , in he p esence o social
and/o judicial cos s ha decline wi h he mo he ’s age a epo ing, he model p edic s ha o e y young
mo he s, epo ing may become highe in a la e census when he mo he is olde , and hese cos s ha e
diminished. This is wha we obse e in all h ee coun ies o mo he s aged 10 o 14 (Figu e 2, igh column).
Fo bi hs occu ing close o and p io o he ime o a census, bi h coun s based on censuses a e lowe han
in he egis ies when using he con empo aneous census bu highe han he egis ies when using da a om
he subsequen census, collec ed 10 yea s la e . This phenomenon leads o he single-c ossing p edic ion
desc ibed in P oposi ion 4.
13
In he da a, his gap also e lec s he ac ha census es ima es only include bi hs whe e he mo he and
child eside in he same household a he ime o he census.
20
These esul s e eal ha e en adminis a i e egis ies–— he "gold s anda d" o e ili y measu emen –
—signi ican ly unde coun child-mo he bi hs. Cu en Uni ed Na ions guidelines ega d high-co e age
bi h egis ies as he mos eliable da a sou ce o es ima ing adolescen bi h a es, a o ing hem o e
su ey and census da a (Kisambi a and Schmid,2022). Howe e , we ind ha e ospec i e census bi h
coun s, despi e hei endency o unde es ima e he ac ual numbe o bi hs o olde age g oups, can p o ide
a mo e accu a e es ima e o child-mo he bi hs han egis ies.
5 Mechanisms
In his sec ion, we in es iga e he unde lying mechanisms con ibu ing o he unde - epo ing o child-mo he
bi hs.
5.1 A ibu ing Child en o Di e en Mo he s
Unde - epo ing a child-mo he ’s bi h does no always imply ha he child’s exis ence is en i ely un epo ed.
The household could acknowledge he child bu a ibu e hei pa en age o an olde woman in he household,
such as he child’s g andmo he o aun . This s a egy could help amilies a oid he social and legal cos s
associa ed wi h epo ing a bi h o a young mo he . Mo eo e , i o e s a plausible explana ion o he
child’s p esence when no iced by census enume a o s.
Figu e 6: Gap in Repo ed Childbi hs be ween Consecu i e Censuses by Mo he ’s Age,
Including Women beyond Fe ile Age
No es: The sample comp ises 80 pai s o na ional censuses om 59 coun ies, wi h one obse a ion pe census pai , age o he mo he a
bi h and he yea o bi h. We es ima e he numbe o bi hs by iden i ying he age a bi h o he epo ed mo he . The bias measu e is
calcula ed as one minus he a io o he numbe o bi hs es ima ed om he i s census o hose es ima ed om he second census
1−bi hs
bi hs +10
, o e he age o he mo he a bi h. The in e al be ween censuses a e ages 10 yea s, bu i includes spans anging
om 8 o 12 yea s. Samples exclude child en o o eign ci izenship.
21
Du yea, S., Olgia i, A., and S one, L. (2006). The Unde -Regis a ion o Bi hs in La in Ame ica. SSRN
Elec onic Jou nal, (Janua y).
Ebbe s, A. L. and Smi s, J. (2022). Household and con ex -le el de e minan s o bi h egis a ion in Sub-
Saha an A ica. PLoS ONE, 17(4 Ap il):1–17.
Hu s , E. and Pugsley, B. (2010). A e Household Su eys Like Tax Fo ms: E idence om he Sel Employed.
Re iew o Economics and S a is ics, 96(1):19–33.
In eg a ed Public Use Mic oda a Se ies, IPUMS [da ase , accessed Ma ch, . (2024). Minneso a Popula ion
Cen e .
Jayachand an, S. (2015). The Roo s o Gende Inequali y in De eloping Coun ies. Annual Re iew o Economics,
7(1):63–88.
Johnson, M. S. (2020). Regula ion by shaming: De e ence e ec s o publicizing iola ions o wo kplace
sa e y and heal h laws. Ame ican Economic Re iew, 110(6):1866–1904.
Ka ing, A. (2024). Social Signaling and Childhood Immuniza ion: A Field Expe imen in Sie a Leone. The
Qua e ly Jou nal o Economics, pages 2083–2133.
Kea ney, M. S. and Le ine, P. B. (2012). Why is he een bi h a e in he Uni ed S a es so high and why does
i ma e ? Jou nal o Economic Pe spec i es, 26(2):141–166.
Kea ney, M. S. and Le ine, P. B. (2015). Media in luences on social ou comes: The impac o m ’s 16 and
p egnan on een childbea ing. Ame ican Economic Re iew, 105(12):3597–3632.
Keskin, F. and Ça lin, A. (2020). Un egis e ed bi hs o adolescen mo he s in Tu key: In isible child en.
Compa a i e Popula ion S udies, 45:179–200.
Kisambi a, S. and Schmid, K. (2022). Selec ing adolescen bi h a es (10-14 and 15-19 yea s)
o moni o ing and epo ing on sus ainable de elopmen goals.
h ps://www.un.o g/
de elopmen /desa/pd/si es/www.un.o g.de elopmen .desa.pd/ iles/undesa_pd_2022_ p_
selec ing-ab -da a- o -sdg- epo ing.pd . [Accessed: 2023-03-17].
Klepinge , D., Lundbe g, S., and Plo nick, R. (1999). How does adolescen e ili y a ec he human capi al
and wages o young women? Jou nal o Human Resou ces, 34(3):421–448.
Lang, K. and Weins ein, R. (2015). The Consequences o Teenage Childbea ing be o e Roe . Wade. Ame ican
Economic Jou nal: Applied Economics, 7(4):169–197.
Meneses, E. and Ramí ez, M. (2018). Ni eles y endencias de la ecundidad en niñas y adolescen es de 10
a 14 años en méxico y ca ac e ís icas de las meno es y de los pad es de sus hijos e hijas, a pa i de las
es adís icas i ales de nacimien os de 1990 a 2016. (106):117–152.
Pullum, T. W. and Becke , S. (2014). E idence o omission and displacemen in DHS bi h his o ies . DHS
Me hodological Repo s No. 11, (Sep embe ).
Pullum, T. W., Becke , S., and Cha l on, C. (2018). Demog aphic and heal h su eys me hodology - household
ques ionnai e. Technical epo , The DHS P og am.
Sande u , J. and Glassman, A. (2015). The Poli ical Economy o Bad Da a: E idence om A ican Su ey and
Adminis a i e S a is ics. Jou nal o De elopmen S udies, 51(2):116–132.
Schoumake , B. and Sánchez-Páez, D. A. (2022). Unde -15 e ili y a ound he wo ld. Popula ion & Socie ies,
601(6):4.
28
Singh, S. and Da och, J. E. (2000). Adolescen p egnancy and childbea ing: Le els and ends in de eloped
coun ies. Family Planning Pe spec i es, 32(1):14–23.
Spoo enbe g, T. (2014). Re e se su i al me hod o e ili y es ima ion: An e alua ion. Demog aphic Resea ch,
31(1):217–246.
Uni ed Na ions (1983). Manual X. Indi ec echniques o Demog aphic es ima ions. Uni ed Na ions.
Uni ed Na ions DESA (2017). P inciples and ecommenda ions o popula ion and housing censuses, olume 3.
Uni ed Na ions DESA (2020a). Fe ili y among Young Adolescen s a Ages 10-14 Yea s - A global assessmen
Fe ili y among Young Adolescen s. Technical epo .
Uni ed Na ions DESA (2020b). Fe ili y among young adolescen s a ages 10-14 yea s – a global assess-
men .
h ps://www.un.o g/de elopmen /desa/pd/si es/www.un.o g.de elopmen .desa.pd/ iles/
desa_pd_2017_ e ili y_among_young_adolescen s.pd . [Accessed: 2023-03-17].
Wend , A., Hellwig, F., Saad, G. E., Faye, C., Boe ma, T., Ba os, A. J., and Vic o a, C. G. (2022). Bi h
egis a ion co e age acco ding o he sex o he head o household: an analysis o na ional su eys om
93 low- and middle-income coun ies. BMC Public Heal h, 22(1):1–11.
WHO, W. H. O. (2024). Adolescen bi h a e (pe 1000 women) – global heal h obse a o y da a. Accessed:
2024-09-24.
29
Appendix (Fo Online Publica ion)
A Addi ional Figu es and Tables
A1 Addi ional Figu es
(a) Bi hs in Mexico by Mo he Age G oup (1985-2018)
Figu e A1: To al Bi hs by Mo he ’s Age G oup Acco ding o Census and Regis y Da a
No e: The igu e p esen s annual bi h es ima es de i ed om na ional censuses and bi h egis ies o
Mexico. The le panels ep esen mo he s aged 15 o 25, while he igh panels depic hose aged 10 o 14. The
g aph po ays Mexican bi h es ima es, employing da a om he 2010 and 2020 Censuses supplemen ed by
he bi h egis y wi h only egis e ed bi hs wi hin one yea om bi h. In o ma ion om 2020, po en ially
a ec ed by COVID-19 pandemic- ela ed epo ing delays, is excluded.
1
Figu e A2: A e age o Bi hs Repo ed by Mo he ’s Age
2
Mo he s aged 10-14 Mo he s aged 15-19
Mo he s aged 20-24 Mo he s aged 25-29
Figu e A3: Exhibi 2: Mexico by Age G oups
No es: Sample does no include egis y in o ma ion om 2020 onwa ds due o he COVID-19 eme gency.
3
Figu e A4: Judicial Cos s and Minimum Age o Consen
No e: This g aph depic s a RD plo o ou bias measu e alongside he di e ence be ween a mo he ’s age a bi h and he minimum age o
consen . The sample excludes child en bo n in he yea o he i s census. MAoC anges be ween 12 and 18 yea s old.
4
A2 Addi ional Tables
Table A1: Es ima ed Un eco ded Bi hs, Rela i e o Subsequen Census
(1) (2) (3) (4) (5) (6) (7)
Mo he ’s age a child bi h To al (ages 10 - 14)
Baseline Census yea 10 11 12 13 14 Coun (in 1000s) Sha e o coun ed
High income coun ies (2023)
Aus ia 1991 0.050 . 0.030 0.020 0.100 0.200 95.24%
Chile 1992 0.290 0.440 0.320 0.190 0.240 1.480 42.53%
F ance 1990 0.360 0.636 0.800 0.924 1.484 4.204 95.11%
F ance 1999 0.101 0.181 0.072 0.123 0.107 0.585 70.90%
G eece 1991 . . 0.020 0.130 0.230 0.380 55.07%
Hunga y 1990 0.140 0.260 0.020 0.180 0.200 0.800 80.00%
Hunga y 2001 0.140 0.160 0.080 0.060 0.040 0.480 63.16%
Panama 1990 0.270 0.350 0.210 0.320 0.430 1.580 65.02%
Panama 2000 0.150 0.210 0.340 0.380 0.250 1.330 58.59%
Po ugal 1991 0.020 0.240 0.080 0.180 0.140 0.660 56.90%
Po ugal 2001 0.080 0.040 0.080 0.060 0.220 0.480 77.42%
Pue o Rico 1990 0.052 0.127 0.214 0.493 0.157 1.043 69.03%
Pue o Rico 2000 . . 0.414 . 0.216 0.630 71.92%
Romania 1992 0.190 0.140 0.070 0.120 0.770 1.290 52.02%
Romania 2002 0.040 0.030 0.240 0.210 0.130 0.650 46.76%
Russia 2002 1.380 2.200 1.420 2.220 2.980 10.200 88.24%
Spain 1991 0.200 0.140 0.120 0.240 0.037 0.737 67.03%
Spain 2001 0.717 0.396 0.345 0.347 0.317 2.122 94.65%
Swi ze land 1990 . 0.020 0.020 . 0.020 0.060 75.00%
T inidad and Tobago 2000 0.011 0.011 0.034 0.034 0.026 0.117 85.38%
Uni ed S a es 1990 4.935 7.949 12.000 19.397 30.122 74.403 91.15%
Uni ed S a es 2000 5.858 4.623 11.623 13.770 18.333 54.207 80.53%
U uguay 1996 0.082 0.018 0.076 . 0.269 0.445 56.54%
Uppe middle income coun ies (2023)
A gen ina 1991 1.435 1.811 2.045 2.131 3.886 11.308 68.66%
A menia 2001 0.050 0.050 0.080 0.100 0.160 0.440 91.67%
Bela us 1999 0.030 0.030 0.080 0.070 0.180 0.390 90.70%
Bo swana 1991 0.030 0.120 0.080 0.100 0.140 0.470 88.68%
Bo swana 2001 0.050 0.020 0.140 0.100 0.180 0.490 85.96%
B azil 1991 11.422 13.527 14.807 15.389 14.839 69.982 54.07%
B azil 2000 11.931 13.012 12.578 14.396 13.935 65.853 46.82%
China 1990 6.100 11.600 10.300 14.700 25.500 68.200 87.44%
Cos a Rica 2000 0.190 0.180 0.160 0.240 0.110 0.880 53.99%
Cuba 2002 0.310 0.440 0.350 0.300 0.430 1.830 57.55%
Dominican Republic 2002 0.460 0.920 1.110 1.450 2.180 6.120 56.25%
Ecuado 1990 1.250 0.970 . 1.070 . 3.290 34.41%
Ecuado 2001 0.580 0.680 0.800 1.100 1.370 4.530 47.73%
Fiji 1996 0.060 0.110 0.080 0.130 0.030 0.410 91.11%
Gua emala 1994 0.810 1.430 2.060 3.040 4.090 11.430 69.48%
Indonesia 1990 33.105 41.298 51.900 75.853 99.034 301.190 90.75%
Indonesia 2000 1.310 1.000 26.210 34.750 44.840 108.110 75.26%
Jamaica 1991 0.149 0.168 0.111 0.239 0.455 1.122 78.90%
Malaysia 1991 0.800 1.550 2.600 2.350 2.700 10.000 76.34%
Mau i ius 1990 0.040 . 0.030 0.020 0.040 0.130 81.25%
Mau i ius 2000 0.020 0.040 . 0.040 . 0.100 90.91%
Mexico 1990 5.553 7.694 8.866 12.756 20.231 55.100 74.18%
Mexico 2000 4.846 6.631 7.233 9.261 11.227 39.198 64.54%
Pa aguay 1992 0.332 0.332 0.474 0.729 0.917 2.785 81.08%
Sou h A ica 2001 1.650 1.863 3.400 3.544 5.304 15.760 86.67%
Thailand 1990 2.090 2.848 3.095 3.746 2.818 14.596 56.11%
Tu key 1990 4.280 5.420 7.420 9.420 12.920 39.460 76.71%
Lowe middle income coun ies (2023)
Bangladesh 1991 123.530 116.580 109.680 215.780 156.660 722.230 81.76%
Bangladesh 2001 33.360 50.050 86.830 167.520 195.530 533.290 80.63%
Benin 1992 2.600 2.560 4.020 5.060 3.910 18.150 71.12%
Benin 2002 4.890 4.440 7.450 11.000 5.760 33.540 84.27%
Boli ia 1992 0.590 0.650 0.930 1.280 1.970 5.420 76.23%
Boli ia 2001 1.423 1.815 1.454 1.595 1.948 8.234 84.42%
Cambodia 2004 0.315 0.019 0.185 0.019 0.199 0.737 100.00%
Egyp 1996 9.772 11.660 18.968 30.029 41.137 111.566 91.83%
Ghana 2000 5.390 5.790 6.150 9.530 11.000 37.860 94.37%
Kenya 1999 9.060 10.890 19.300 27.730 31.090 98.070 92.30%
Ky gyzs an 1999 0.530 0.490 0.320 0.490 0.620 2.450 97.61%
Nepal 2001 3.242 4.452 5.521 11.888 14.349 39.452 83.03%
Nica agua 1995 0.710 0.800 0.910 1.120 1.140 4.680 59.62%
Papua New Guinea 1990 0.500 0.480 1.870 3.250 3.520 9.620 83.72%
Philippines 1990 5.042 6.250 7.003 8.279 9.458 36.032 79.89%
Philippines 2000 0.475 0.996 1.056 1.217 0.174 3.917 22.01%
Senegal 2002 5.609 4.941 6.074 7.088 2.352 26.064 68.44%
Tanzania 2002 13.459 15.042 26.083 27.007 30.652 112.242 87.79%
Vie nam 1999 1.218 1.640 1.363 2.553 3.350 10.125 82.66%
Zambia 1990 3.870 5.010 4.340 6.470 6.960 26.650 90.80%
Zambia 2000 3.540 4.080 5.470 7.870 9.770 30.730 83.80%
Low income coun ies (2023)
Bu kina Faso 1996 4.450 5.290 4.180 5.960 3.760 23.640 63.91%
Malawi 1998 4.290 5.400 5.860 7.640 7.680 30.870 79.66%
Mali 1998 4.240 4.560 4.220 4.720 . 17.740 54.02%
Mozambique 1997 9.780 11.650 11.390 11.420 11.760 56.000 71.46%
Rwanda 1991 0.070 0.280 0.400 0.430 0.770 1.950 50.52%
Rwanda 2002 0.410 0.680 0.570 0.760 1.050 3.470 79.95%
Sie a Leone 2004 6.090 4.560 5.530 9.690 5.900 31.770 94.78%
Uganda 1991 9.286 11.325 14.480 19.492 22.164 76.747 92.68%
Uganda 2002 9.930 11.640 12.820 14.930 19.270 68.590 93.96%
No es: Columns 1-5 p esen he o al numbe o un eco ded bi hs by mo he ’s age a bi h, calcula ed as he di e ence be ween he es ima ed o al bi hs in
census +10 and census . Column 6 epo s he agg ega e numbe o un eco ded bi hs, while Column 7 epo s he sha e o un eco ded bi hs ela i e o he
o al eco ded bi hs in census +10.
5
C Da a Appendix
C1 Sample
Table C1: Coun y-Census Pai s in Sample
Coun y T0T1Coun y T0T1
A gen ina 1991 2001 Mau i ius 1990 2000
Aus ia 1991 2001 Mau i ius 2000 2011
Bangladesh 1991 2001 Mexico 1990 2000
Bangladesh 2001 2011 Mexico 2000 2010
A menia 2001 2011 Mo occo 2004 2014
Boli ia 1992 2001 Mozambique 1997 2007
Boli ia 2001 2012 Nepal 2001 2011
Bo swana 1991 2001 Nica agua 1995 2005
Bo swana 2001 2011 Panama 1990 2000
B azil 1991 2000 Panama 2000 2010
B azil 2000 2010 Papua New Guinea 1990 2000
Bela us 1999 2009 Pa aguay 1992 2002
Cambodia 2004 2013 Philippines 1990 2000
Chile 1992 2002 Philippines 2000 2010
China 1990 2000 Po ugal 1991 2001
Cos a Rica 2000 2011 Po ugal 2001 2011
Benin 1992 2002 Pue o Rico 1990 2000
Benin 2002 2013 Pue o Rico 2000 2010
Dominican Republic 2002 2010 Romania 1992 2002
Ecuado 1990 2001 Romania 2002 2011
Ecuado 2001 2010 Russia 2002 2010
Fiji 1996 2007 Rwanda 1991 2002
F ance 1990 1999 Rwanda 2002 2012
F ance 1999 2011 Senegal 2002 2013
Ghana 2000 2010 Sie a Leone 2004 2015
G eece 1991 2001 Vie nam 1999 2009
Gua emala 1994 2002 Spain 1991 2001
Hunga y 1990 2001 Spain 2001 2011
Hunga y 2001 2011 Swi ze land 1990 2000
Indonesia 1990 2000 Thailand 1990 2000
Indonesia 2000 2010 T inidad and Tobago 2000 2011
I eland 1991 2002 Tu key 1990 2000
I eland 2002 2011 Uganda 1991 2002
I aly 2001 2011 Uganda 2002 2014
Jamaica 1991 2001 Tanzania 2002 2012
Kenya 1999 2009 Uni ed S a es 1990 2000
Ky gyz Republic 1999 2009 Uni ed S a es 2000 2010
Malawi 1998 2008 Bu kina Faso 1996 2006
Malaysia 1991 2000 U uguay 1996 2006
Mali 1998 2009 Zambia 1990 2000
Zambia 2000 2010
No es: .
6
C2 De ini ions and Sou ces o Va iables
Table C2: De ini ions and Sou ces o Va iables Used in he Analysis
Measu e Desc ip ion Sou ce
Bias measu e
This a iable is calcula ed by one minus he a io be-
ween he numbe o bi hs epo ed in he i s census
o e he numbe o bi hs epo ed in he second cen-
sus. To es ima e he yea ly numbe o bi hs, we coun
he o al numbe o indi iduals bo n in a speci ic yea
om he census mic oda a ob ained om IPUMS in e -
na ional.
IPUMS In e na ional
% women making in-
o med sexual decisions
This indica o measu es he pe cen age o women aged
15-49 who a e ma ied o in a union and ha e au on-
omy in h ee key a eas: making decisions abou sexual
ela ions, con acep i e use, and hei own ep oduc-
i e heal h ca e. Speci ically, only women who ha e an
ac i e decision in he h ee componen s a e conside ed
o ha e au onomy in hese a eas.
UN Demog aphic and
Heal h Su eys
Log income pe capi a
This me ic is de i ed by compu ing he na u al loga-
i hm o he GDP a pu chase ’s p ices, which is mea-
su ed in cons an 2017 in e na ional dolla s.
Wo ld Bank
Di e ence be ween age and
MAoC
This a iable is calcula ed by sub ac ing he minimum
age o consen in a gi en coun y om he age o he
mo he a he ime o gi ing bi h.
ageo consen .ne
C3 Fu he De ails on Measu es Used
C3.1 Census Da a
The da a used in his p ojec we e ob ained di ec ly h ough IPUMS. IPUMS In e na ional ha monizes and
in eg a es census mic oda a om a ound he wo ld in o a consis en o ma o e ime and space, which
makes i use ul o c oss- empo al and c oss-na ional compa a i e esea ch.
7
C3.2 Adminis a i e Bi h Regis ies
Sis ema de In o mações sob e Nascidos Vi os, SINASC (B azil). This in o ma ion sys em
collec s da a on li e bi hs h ough he issuance o a Li e Bi h Ce i ica e (Decla ação de Nascido Vi o -
DNV) a he ime o bi h, p ima ily in hospi als and heal hca e acili ies. When a child is bo n, heal hca e
p o essionals ill ou he DNV, which includes comp ehensi e de ails abou he bi h. The hospi al is
esponsible o en e ing his in o ma ion in o he SINASC sys em. In cases o home bi hs, midwi es o
a ending heal hca e p o ide s a e esponsible o issuing he DNV. The DNV is manda o y and se es as
he p ima y documen o eco ding a li e bi h. Howe e , e iciency and e ec i eness o da a collec ion
and epo ing can a y ac oss s a es due o di e ences in local heal hca e in as uc u e and adminis a i e
p ac ices. Pa en s mus ake he DNV o a ci il egis y o ice (ca ó io) o o icially egis e he bi h. This
p ocess links he heal hca e da a collec ion wi h he ci il egis a ion sys em.
Despi e i s comp ehensi e co e age, a signi ican sha e o bi hs s ill go un egis e ed. This can happen
when bi hs do no occu in hospi als and ail o be egis e ed a e wa ds, because pa en s o hospi al-bo n
child en epo alse o e oneous in o ma ion, o because hospi al bi hs a e no egis e ed a a ci il egis y
o ice by he pa en s, and he c oss- alida ion sys em lags hose bi hs as equi ing u he checks.
Es adís ica de Nacimien os del Regis o Ci il, INEGI (Mexico). This da ase is pa o
Mexico’s o icial i a s a is ics (Es adís ica Vi ales), and is compiled by INEGI, he Mexican Na ional Ins i u e
o Geog aphy and S a is ics (Ins i u o Nacional de Es adís ica y Geog a ía), o each yea since 1985. Each
yea con ains de ailed mic oda a o all bi hs egis e ed in Mexico in ha yea , including da e
15
and place
o bi h, and socio-demog aphic in o ma ion on pa en s (including age a bi h). In Mexico, when a bi h
occu s, heal hca e p o ide s a e equi ed o issue a Ce i ica e o Bi h (Ce i icado de Nacimien o) wi hin
he i s 24 hou s. This ce i ica e is manda o y and se es bo h legal and s a is ical pu poses. The pa en s
a e hen equi ed o ake he Ce i ica e o Bi h o he local Ci il Regis y o ice ( egis o ci il) o ob ain
he o icial Ac a de Nacimien o (bi h ce i ica e). The Ci il Regis y o ices sp ead a ound he coun y hen
send hese Ac a de Nacimien o documen s di ec ly o INEGI, whe he in p in ed o m o elec onically. A e
ecei ing he da a, INEGI p ocesses, alida es and di uses he inal da ase .
As in he case o B azil, bi hs s ill can go un egis e ed in Mexico o simila easons: un egis e ed home
bi hs, non- egis a ion o bi hs in he ci il egis y, he pu pose ul o acciden al egis a ion o inco ec
in o ma ion, and lack o access o ci il egis y uni s. The e is s ill ano he po en ial sou ce o da a on li e
15
The da e o bi h is di e en om he egis a ion da e, such ha a gi en yea ’s da ase can include bi hs
ha ook place se e al yea s be o e he egis a ion da e.
8
