Muñoz S., E cio
Wo king Pape
In e gene a ional educa ional mobili y wi hin Chile
IDB Wo king Pape Se ies, No. IDB-WP-01707
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: Muñoz S., E cio (2025) : In e gene a ional educa ional mobili y wi hin Chile, IDB
Wo king Pape Se ies, No. IDB-WP-01707, In e -Ame ican De elopmen Bank (IDB), Washing on,
DC,
h ps://doi.o g/10.18235/0013513
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/324783
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/
In e gene a ional Educa ional Mobili y wi hin
Chile
E cio Munoz
WORKING PAPER No IDB-WP-01707
In e -Ame ican De elopmen Bank
Gende and Di e si y Di ision
Ap il 2025
In e gene a ional Educa ional Mobili y wi hin
Chile
E cio Munoz
In e -Ame ican De elopmen Bank
Gende and Di e si y Di ision
Ap il 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
Muñoz, E cio.
In e gene a ional educa ional mobili y wi hin Chile / E cio Muñoz.
p. cm. — (IDB Wo king Pape Se ies ; 1707)
Includes bibliog aphical e e ences.
1. Educa ional mobili y-Chile. 2. Educa ional equaliza ion-Chile. 3. Social
mobili y-Chile. I. In e -Ame ican De elopmen Bank. Gende and Di e si y
Di ision. II. Ti le. III. Se ies.
IDB-WP-1707
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 .
In e gene a ional Educa ional Mobili y wi hin Chile∗
E cio Mu˜noz†
In e -Ame ican De elopmen Bank
Ap il 2025
Abs ac
I p o ide es ima es o in e gene a ional mobili y (IGM) in educa ion a a disag-
g ega ed geog aphic le el o Chile, a coun y wi h high school-le el s a i ica ion by
socioeconomic s a us and a decen alized adminis a ion o public schools. I documen
wide a ia ion ac oss communes. Rela i e mobili y is co ela ed o he numbe o doc-
o s, he numbe o s uden s pe eache , and ea nings inequali y. Using a LASSO, I
ind ha he sha e o s uden s en olled in public schools, he numbe o s uden s pe
eache , popula ion densi y, and municipal budge a e he s onges p edic o s o IGM.
I also documen wi hin-coun y a iabili y in how pa en al educa ion is associa ed wi h
o he child en’s ou comes.
JEL-Codes: D63, I24, J62.
Keywo ds:Socioeconomic mobili y, Geog aphy, Chile, Educa ion.
∗I am g a e ul o Jona han Conning, Ch is os Giannikos, Roy an de Weide, N´u ia Rod ´ıguez-Planas,
Wim Vij e be g, and he pa icipan s o he mic o b own bag, he disse a ion semina a CUNY and he
i ual WEAI 2021 Con e ence o use ul commen s. Pablo Va gas p o ided ou s anding esea ch assis ance.
Any iews exp essed in his pape a e hose o he au ho s and should no be a ibu ed o he In e -Ame ican
De elopmen Bank.
†Email: [email p o ec ed]
1
I In oduc ion
How much o an indi idual’s educa ional achie emen is due o his o he pa en s’ educa ional
achie emen s? High pe sis ence in educa ional ou comes ac oss gene a ions can lead o
un ealized human capi al po en ial and ine icien alloca ion o esou ces and alen s ha
esul in lowe economic g ow h. Mo eo e , i can be a mechanism by which economic
ad an age is inhe i ed, as educa ion is linked o he capaci y o gene a e income and weal h.
Economis s ha e made impo an p og ess in documen ing he le el o in e gene a ional
mobili y (IGM) in educa ion (i.e., he ela ionship be ween educa ional ou comes o pa en s
and child en) o many coun ies (see Van de Weide e al. 2024).1Howe e , he e idence
a he coun y le el can hide impo an a ia ion wi hin coun ies, as i has been shown by
a g owing li e a u e ( o example, Alesina e al. 2021,2023,Ashe e al. 2024,Ca d e al.
2022,De enoncou 2022,Hilge 2016,Feigenbaum 2018,Munoz 2024) ha es ima es IGM
o small geog aphical uni s, ex ending he li e a u e on IGM in income ini ia ed by Che y
e al. (2014).
In his pape , I con ibu e o his li e a u e in h ee ways. Fi s , I es ima e in e gene a ional
mobili y in educa ion in Chile a he coun y, egion, and commune le el using census da a
o a coho bo n in he 1990s. I o e eigh indica o s ha desc ibe he associa ion be ween
child en’s and pa en s’ yea s o schooling,2cap u ing di e en policy- ele an conce ns and
aspec s o his associa ion (e.g., how much mo e educa ed child en wi h pa en s wi h an
ex a yea o schooling end o be, how likely a e child en om low-educa ed pa en s o
be low-educa ed, how likely a e child en o su pass he educa ion o hei pa en s, e c.),
he e o e p o iding a b oad iew o IGM.3I p o ide hese es ima es in an online da a ap-
pendix o u u e esea ch. Second, I show how o he child en’s ou comes, such as eenage
1See To che (2021) o a su ey ocused on de eloping coun ies.
2Th oughou he pape , I will e e o he coho o in e es as “child en,” and I will e e o hei pa en s
o olde ela i es li ing in he same household as “pa en s”. I will p ecisely de ine who will be conside ed as
a pa en in Sec ion II.
3Deu sche & Mazumde (2023) p o ide a amewo k ha highligh s he key concep s and p ope ies o
di e en indica o s o IGM.
2
p egnancy and e ia y educa ion a endance, a e also associa ed wi h pa en al educa ion a
he coun y le el and display wide a ia ion wi hin he coun y. Finally, I explo e how he
es ima es o educa ional IGM a e co ela ed wi h a ich se o a iables ela ed o income,
geog aphy, educa ion, municipal budge , and o he cha ac e is ics o he communes. Fu -
he mo e, I in es iga e by means o a lasso (leas absolu e sh inkage and selec ion ope a o ),
wha co ela es ha e he mos p edic i e powe o e IGM a he le el o commune.
IGM li e a u e o Chile. P e ious s udies ha e used di e en household and opinion
su eys (see o example, To che 2005,He z e al. 2007,Nunez & Mi anda 2010,Na ayan
e al. 2018,Celhay e al. 2010,Celhay & Gallegos 2015,Sapelli 2016,Neidh¨o e e al. 2018,
Van de Weide e al. 2024) o documen IGM in income, educa ion, and o he socioeconomic
measu es. Howe e , hey all ha e in common ha he samples a e no ep esen a i e a
he commune le el, so hey ocus on coun y-le el es ima es. Two excep ions a e Celhay &
Gallegos (2015) which also explo es mobili y a he egional le el ( he coa se adminis a i e
uni in which he coun y is di ided), and Co ´es O ihuela e al. (2023), which uses labo
ea nings in he o mal sec o om adminis a i e eco ds o es ima e income mobili y a he
egional le el and be ween communes ( he smalles adminis a i e uni ) in he Me opoli an
Region.
Ins i u ional backg ound. Chile is an in e es ing case s udy o analyze IGM a he sub-
na ional le el. On he one hand, he coun y is one o he iches economies in he La in
Ame ican egion and has shown signi ican p og ess in po e y educ ion and income pe
capi a g ow h in he las h ee decades. On he o he hand, income inequali y is ela i ely
high o OECD s anda ds, and p e ious esea ch has documen ed high school-le el s a i ica-
ion by socioeconomic s a us (Mizala & To che 2012,Gu i´e ez & Ca asco 2021), which has
ueled some educa ional e o ms in he las decade. In addi ion, he coun y is ma ked by
he ee-ma ke e o ms inhe i ed om he mili a y dic a o ship (1973-1990). This includes a
uni e sal ouche sys em and decen aliza ion o he adminis a ion o public schools, which
3
a e managed by municipali ies.4
In e ms o IGM a he coun y le el, he bes e idence a ailable a a global scale (Na ayan
e al. 2018,Van de Weide e al. 2024) shows some in e es ing indings o Chile. Among he
148 coun ies o which he e a e es ima es o educa ional mobili y o he coho bo n in he
1980s, he coun y anks ela i ely low when a summa y s a is ic o ela i e mobili y, such
as one minus he Pea son co ela ion coe icien be ween yea s o schooling o child en and
pa en s is used bu somewha mo e mobile acco ding o one minus he eg ession coe icien
be ween hese wo a iables.5In con as , he coun y seems much mo e mobile when we
look a a measu e o absolu e mobili y like he sha e o s uden s wi h highe educa ion
han pa en s. Howe e , when a measu e ha aims o cap u e di ec ional mobili y om he
bo om o he op is conside ed (i.e., “ ags o iches” o po e y o p i ilege a e as named in
Na ayan e al. 2018), hen he coun y appea s among he leas mobile ones (see Figu e A1
in he Appendix).6
The e olu ion ac oss di e en coho s o hese indica o s also show some in e es ing pa e ns
when compa ed o simple a e ages by egion as classi ied in Na ayan e al. (2018).7Chile
does no show much p og ess in mos o he indica o s ela i e o egional a e ages excep
o absolu e mobili y (sha e o s uden s wi h highe educa ion han pa en s) and ela i e
mobili y measu ed as 1 −β. In con as , ela i e mobili y measu ed as 1 −ρ(independen
o he ma ginal dis ibu ions o educa ion) has emained a lowe le els han all he egional
a e ages o all he coho s in he same way as he po e y o p i ilege a io (o ags o
iches).
4A ecen e o m s a ed a p ocess o cen aliza ion in 2018.
5The co ela ion coe icien can be ans o med in o he eg ession coe icien by mul iplying i by he
a io o he s anda d de ia ion o child schooling o pa en schooling. The e o e, di e ences be ween hem
a e explained by changes in inequali y ac oss gene a ions.
6Ranked 138 among he 148 a ailable es ima es.
7Figu e A2 in he Appendix plo s all hese indica o s ac oss coho s.
4
II Da a and Me hods
Da a. I use indi idual-le el da a om he 2017 census o housing and popula ion ob ained
om he Na ional Ins i u e o S a is ics.8This s a is ical ope a ion, which aimed o cap u e
he o al popula ion o Chile, includes demog aphic de ails such as age, sex, educa ion,
household composi ion, as well as de ailed geog aphical in o ma ion.
Sample de ini ion. The ull-coun census da abase con ains in o ma ion abou 17,574,003
indi iduals. I keep people bo n in Chile aged be ween 21 and 25 yea s and d op hose
conside ed domes ic se ice, li ing in collec i e housing, o in ansi , which educes he
sample o 1,155,207 indi iduals, 568,231 men and 586,976 women.9
Educa ion. The census da a con ains a a iable epo ing schooling, ega dless o he ack
o kind o s udy. When I s udy how he educa ional a ainmen o child en ela es o he
a ainmen o pa en s, I ake he highes a ainmen among he indi iduals in he olde
gene a ion.10 Gi en he ypical educa ional pa h in Chile, whe e s uden s s a i s g ade a
he age o 6, he a e age s uden would be able o a ain a mos 15 yea s o schooling a he
age o 21. To accommoda e o his, he indica o s a e compu ed using yea s o schooling
censo ed a 15 o bo h child en and pa en s.11
Geog aphy. Chile is di ided in o 16 egions, 56 p o inces, and 346 communes.12 The da a
se con ains in o ma ion on whe e he in e iew was conduc ed and he place o bi h in
e ms o hese h ee adminis a i e di isions. I use he la e o assign people o places and
es ima e IGM o he coun y, by egion, and by commune.
8The da a can be accessed a h ps://www.ine.gob.cl/es adis icas/sociales/censos-de-poblacion-y-
i ienda/censo-de-poblacion-y- i ienda.
9Van de Weide e al. (2024) use he same age ange o es ima e IGM in educa ion using su ey da a o
39 coun ies.
10The esul s a e quali a i ely simila i I use he a e age ounded o he nea es in ege ins ead o he
maximum.
11Simila censo ing o yea s o schooling is used in Neidh¨o e e al. (2018) wi h su ey da a o compu e
IGM a he coun y le el o 18 coun ies in La in Ame ica. Figu e A3 displays he dis ibu ion o educa ional
a ainmen o pa en s and child en.
12Chile does no ha e a commonly used designa ion o commu ing zones, such as he one used in Che y
e al. (2014) o he Uni ed S a es. Many communes a e wi hin commu ing dis ance in he coun y, pa ic-
ula ly wi hin he Me opoli an a ea. Howe e , he es ima es a he egional le el o his speci ic case a e
close o wha could be conside ed a commu ing zone.
5
case o women.22 These ou comes a e po en ially e y consequen ial in li e ajec o ies,
as he i s one is posi i ely associa ed wi h ea nings and o he indica o s o well-being
(O eopoulos & Pe onije ic 2013), while he second is nega i ely associa ed wi h income
and some indica o s o well-being (Fle che & Wol e 2009). Fu he mo e, hese ou comes
can be measu ed a ea lie ages han ou main ou come (i.e., yea s o schooling), educing
he magni ude o any po en ial co esidence bias, as co esidence a es dec ease wi h age.
Fi s , I es ima e he p obabili y o a ending a leas one yea o e ia y educa ion using a
sample o indi iduals be ween he ages o 19 and 21. Figu e A4c shows his likelihood o
each pa en al educa ional a ainmen , inding a posi i e slope app oxima ely equal o 0.046
wi h a somewha p ominen discon inui y a 12 yea s o schooling and a somewha nonlinea
ela ionship o low alues o pa en s’ yea s o schooling. This con as s wi h he i ually
linea ela ionship be ween pa en al income ank and college a endance documen ed o
he US in Che y e al. (2014). Despi e hese di e ences and o he di e ences in e ms
o measu emen and concep s, I ind simila gaps. The gap in he likelihood o a ending
e ia y educa ion o indi iduals wi h low-educa ed s. highly-educa ed pa en s is a ound
60 pe cen age poin s while Che y e al. (2014) documen ed a gap o 67.5 pe cen age poin s
in he US o indi iduals wi h lowes -income s. highes -income pa en s.
Second, I es ima e he p obabili y o becoming a mo he as a eenage , de ined as ha ing
a child o emales be ween he ages o 15 and 19. Figu e 2b shows his likelihood o
each pa en al educa ional a ainmen , inding a nega i e ela ionship close o linea wi h
a slope o -0.017. The gap be ween highly-educa ed and low-educa ed pa en s is a ound
20-25 pe cen age poin s (Che y e al. 2014, documen s a gap o 29.8 pe cen age poin s o
highes -lowes pa en s’ incomes).
22I use he same econome ic speci ica ion as in Equa ion 1wi h a di e en dependen a iable.
12
Figu e 2: O he child’s ou comes
0 .2 .4 .6 .8 1
O sp ing's likelihood o a ending e ia y educa ion
0 5 10 15 20
Pa en s' yea s o schooling
(a) Te ia y educa ion a endance
0 .1 .2 .3
Likelihood o being mo he as eenage
0 5 10 15 20
Pa en s' yea s o schooling
(b) Teenage bi h ( emales only)
No es: The i s plo displays he likelihood o comple ing a leas one yea o e ia y educa ion o
each le el o educa ion o he gene a ion abo e (highes yea s o schooling among pa en s and olde
ela i es li ing in he same household). The second plo displays he likelihood o ha ing a child as
eenage o each le el o educa ion o he gene a ion abo e. The samples include indi iduals wi h
age be ween 19 and 21 (le ) and 15 and 19 ( igh ). The size o he bubble a ies acco ding o he
numbe o indi iduals.
III.2 In e gene a ional mobili y wi hin Chile
Region-le el es ima es. Be o e p esen ing he mos disagg ega ed es ima es, Table 4sum-
ma izes he eigh measu es o in e es es ima ed o he 16 egions o Chile. Non-negligible
di e ences can be ound ac oss egions in mos o hese dimensions. Fo example, he chances
o eaching he op quin ile o he educa ional dis ibu ion o child en wi h pa en s a he
bo om quin ile (i.e., P1,5) is mo e han 200% highe in he no he n A ica y Pa inaco a
egion ela i e o Ays´en egion. Simila ly, in e ms o absolu e mobili y (i.e., α), he e a e
egions wi h mo e han one yea o di e ence, and ela i e mobili y (i.e., 1−β) is 17% highe
in A ica y Pa inaco a han in Me opoli ana de San iago o Los Rios. When I conside ela-
i e mobili y measu ed wi h he co ela ion coe icien (1−ρ), he le el in A ica y Pa inaco a
is app oxima ely 30% highe han in he egion wi h he lowes alue (A aucan´ıa).23
23Table A4 in he Appendix epo s he las h ee indica o s o IGM using he dis ibu ion o educa ional
a ainmen a he coun y le el e sus a he egion le el. The e is he e ogenei y in he di ec ion o change,
bu o he h ee indica o s, he ange o a ia ion dec eases using he local dis ibu ion.
13
Table 4: Region-le el es ima es o IGM S a is ics
Region α1−β¯
Y¯y≥1−ρ P1,5P1,1P5,5
Ta apac´a 9.66 0.74 11.53 0.60 0.70 0.10 0.37 0.32
An o agas a 9.25 0.71 11.61 0.57 0.68 0.08 0.40 0.31
A acama 9.54 0.73 10.99 0.62 0.68 0.07 0.40 0.29
Coquimbo 9.44 0.72 10.53 0.65 0.66 0.07 0.38 0.32
Valpa a´ıso 9.61 0.72 11.23 0.65 0.68 0.09 0.35 0.34
Libe ado Gene al Be na do O’Higgins 10.06 0.76 9.95 0.71 0.71 0.10 0.34 0.31
Maule 9.75 0.73 9.84 0.73 0.67 0.09 0.36 0.32
Biob´ıo 10.12 0.74 10.65 0.74 0.66 0.11 0.33 0.37
A aucan´ıa 9.58 0.71 9.88 0.75 0.61 0.07 0.38 0.37
Los Lagos 9.35 0.71 9.77 0.71 0.65 0.07 0.41 0.31
Ays´en del Gene al Ca los Ib´a˜nez del Campo 9.38 0.76 9.59 0.65 0.73 0.05 0.44 0.23
Magallanes y de la An ´a ica Chilena 10.36 0.77 11.33 0.66 0.72 0.10 0.30 0.30
Me opoli ana de San iago 9.38 0.70 11.33 0.64 0.65 0.09 0.37 0.36
Los R´ıos 9.46 0.70 10.18 0.72 0.62 0.06 0.38 0.34
A ica y Pa inaco a 10.76 0.82 11.49 0.61 0.79 0.14 0.28 0.31
˜
Nuble 10.02 0.74 9.86 0.76 0.68 0.11 0.33 0.36
No es: The able epo s egion-le el es ima es o absolu e mobili y, ela i e mobili y (1−β), a e age
pa en s’ educa ion, he sha e o child en wi h highe educa ion han pa en s, ela i e mobili y
(1−ρ), ags o iches, in e gene a ional low, and in e gene a ional high, espec i ely. A desc ip ion
o he measu es can be ound in Table 2. Rows a e so ed by he o icial designa ed numbe ha
each egion used o ha e un il 2018.
Commune-le el es ima es. I documen wide wi hin-coun y a ia ion. Rela i e mobili y
measu ed as 1 −β, excluding places wi h less han 50 indi iduals24, anges be ween 0.54
in Quemchi, a commune loca ed in he sou h o he coun y, and 0.97 in San Ped o de
A acama, a commune loca ed in he no h. Non-negligible a ia ion is ound in all he
indica o s s udied. Figu e A10 in he Appendix shows he dis ibu ions o he commune-
le el es ima es o he eigh measu es and Table A3 o he Appendix simila ly epo s some
desc ip i e s a is ics o hese es ima es. Fo all he indica o s I can ind communes wi h
le els a leas 100% g ea e han o he s, in some cases se e al imes g ea e .
The measu es o mobili y based on condi ional p obabili ies de i ed om quin iles o ed-
uca ional a ainmen a e cons uc ed using he dis ibu ion o a ainmen a he coun y
le el o child en and simila ly o pa en s. Simila measu es could be cons uc ed using he
dis ibu ion o a ainmen by commune. In his case, mo ing om he bo om o he op
24Figu e A9 in he Appendix displays he CDF o he sample size by commune, showing ha less han
5% o he communes ha e less han 50 obse a ions.
14
may equi e a highe numbe o yea s o schooling in some places compa ed o o he s and
cap u e a di e en aspec o mobili y. As an addi ional exe cise, I compu e hose measu es
and ind ha P1,5measu es cons uc ed in bo h ways a e highly co ela ed while P1,1is o a
lesse deg ee while in con as , P5,5is no co ela ed (see Figu e A11 in he Appendix.).
Figu e 3maps ela i e mobili y (1 −β) ac oss he coun y. The e a e some egions wi h
clus e s o communes showing ela i ely simila le els o IGM, such as he no he n egions
and mo e he e ogenei y in he cen e o he coun y. Figu e A15 in he Appendix plo s
ela i e mobili y di iding he map o he coun y in o h ee pa s, a no he n egion less he
me opoli an egion, he me opoli an egion, and a sou he n egion. These h ee egions
ha e communes wi h ela i ely low and high le els o in e gene a ional educa ional mobili y.
Howe e , in his map he a ie y in IGM le els o he me opoli an egion (whe e he highes
sha e o he popula ion li es) can be app ecia ed in mo e de ail.
Co ela ions among di e en measu es o IGM. Table 5p esen s he Pea son co ela-
ion coe icien s be ween he eigh mobili y s a is ics compu ed a he commune le el. I ind
he s onges posi i e co ela ion o be be ween absolu e and ela i e mobili y, bo h mea-
su ed wi h 1 −βand 1 −ρ. These h ee measu es a e a he same ime posi i ely co ela ed
o abo e pa en s and ags o iches, especially absolu e mobili y. In e gene a ional low is
nega i ely co ela ed wi h he o he six indica o s.
Table 5: Co ela ion among IGM s a is ics
α1−β¯
Y¯y≥1−ρ P1,5P1,1P5,5
Absolu e mobili y (α) 1
Rela i e mobili y (1 −β) 0.912∗∗∗ 1
A e age educa ion (¯
Y) -0.0175 -0.146∗∗ 1
Abo e pa en s (¯y≥) 0.268∗∗∗ 0.139∗-0.716∗∗∗ 1
Rela i e mobili y (1 −ρ) 0.713∗∗∗ 0.874∗∗∗ -0.0917 -0.0128 1
Rags o iches (P1,5) 0.478∗∗∗ 0.259∗∗∗ 0.296∗∗∗ 0.0604 0.228∗∗∗ 1
In e gene a ional low (P1,1) -0.730∗∗∗ -0.517∗∗∗ -0.207∗∗∗ -0.166∗∗ -0.380∗∗∗ -0.537∗∗∗ 1
In e gene a ional high (P5,5) -0.00472 -0.141∗0.0369 0.233∗∗∗ -0.140∗0.236∗∗∗ -0.0715 1
∗p < 0.05, ∗∗ p < 0.01, ∗∗∗ p < 0.001
O he ou comes wi hin Chile. I es ima e he ela ionship be ween pa en al educa ion and
15
Figu e 3: In e gene a ional educa ional (1 −β) mobili y wi hin Chile
No es: The map plo s ela i e IGM measu ed as one minus he eg ession coe icien (by commune)
be ween child en’s yea s o schooling (using indi iduals aged be ween 21 and 25) agains pa en s’
yea s o schooling. Educa ional a ainmen is censo ed a 15. Communes wi h less han 50 indi-
iduals a e le as missing (Figu e A9 in he Appendix displays he CDF o he sample size by
commune). Figu e A15 in he Appendix displays a e sion o he map di iding Chile in o h ee
a eas.
he wo al e na i e ou comes desc ibed in he p e ious sec ion: he likelihood o a ending
a leas one yea o e ia y educa ion, and he likelihood o being a mo he as a eenage
16
o emales.
Table 6 epo s hese es ima es a he egional-le el. The e is signi ican a ia ion ac oss
egions in he e ec o an addi ional yea o pa en al schooling on he chances o a ending
e ia y educa ion. A aucania shows he s onges e ec (0.044), which sugges ha he
gap be ween indi iduals wi h uneduca ed pa en s and hose wi h highly educa ed ones (21
yea s) in he chances o a ending e ia y educa ion is app oxima ely 92 pe cen age poin s
(21 ×0.044). A ca ea o no e is ha his calcula ion may o e es ima e he e ec in ligh
o he non-linea i y obse ed a he na ional le el in Figu e A4c o lowe le els o pa en al
educa ion. I I assume ha he e ec is null in he i s 5 yea s o educa ion, hen he gap is
app oxima ely 70 pe cen age poin s. On he o he ex eme, Ays´en egion shows he smalles
a e age e ec (0.019).
Simila ly, he e ec o an ex a yea o pa en s’ schooling on eenage bi h a es a ies sig-
ni ican ly ac oss egions. The e ec o one yea goes om a all in he likelihood o a eenage
bi h equal o 0.8 pe cen age poin s in ˜
Nuble o 1.6 pe cen age in An o agas a o Coquimbo.
This las e ec implies a gap be ween uneduca ed and highly educa ed pa en s o app oxi-
ma ely 33.6 pe cen age poin s, which again is meaning ul bu may be an o e es ima ion due
o non-linea i ies.
IV Co ela es o IGM wi hin Chile
In his sec ion, I examine whe he in e gene a ional educa ional mobili y a he commune
le el is co ela ed wi h a b oad se o a iables, including income dis ibu ion, educa ional
cha ac e is ics, municipal budge s, geog aphic ac o s, and o he local a ibu es. Unde -
s anding hese ela ionships is c ucial o iden i ying he main ac o s ha con ibu e o he
pe sis ence o educa ional ou comes ac oss gene a ions. Gi en he la ge numbe o po en ial
p edic o s, he analysis ollows a wo-s ep app oach. Fi s , I in es iga e he co ela ions be-
ween IGM and selec ed key a iables ha he li e a u e and heo e ical amewo ks sugges
17
Table 6: Pa en al educa ion e ec on o he ou comes
Region Te ia y educa ion Teenage bi h
Ta apac´a 0.038 -0.013
An o agas a 0.038 -0.016
A acama 0.042 -0.012
Coquimbo 0.040 -0.016
Valpa a´ıso 0.042 -0.014
Libe ado Gene al Be na do O’Higgins 0.028 -0.010
Maule 0.034 -0.012
Biob´ıo 0.039 -0.013
A aucan´ıa 0.044 -0.013
Los Lagos 0.035 -0.014
Ays´en del Gene al Ca los Ib´a˜nez del Campo 0.019 -0.015
Magallanes y de la An ´a ica Chilena 0.033 -0.009
Me opoli ana de San iago 0.043 -0.015
Los R´ıos 0.039 -0.013
A ica y Pa inaco a 0.026 -0.012
˜
Nuble 0.037 -0.008
No es: The able epo s he associa ion be ween an ex a yea o pa en s’ schooling and he
likelihood o comple ing a leas one yea o e ia y educa ion, as well as he likelihood o ha ing a
child as eenage o emales (compu ed using an OLS eg ession). The samples include indi iduals
aged be ween 19 and 21 (le ) and 15 and 19 ( igh ). Rows a e so ed by he o icial designa ed
numbe ha each egion used o ha e un il 2018.
as pa icula ly ele an . Second, I apply a LASSO (Leas Absolu e Sh inkage and Selec ion
Ope a o ) eg ession o iden i y he s onges p edic o s o mobili y. This me hod pe o ms
a selec ion o co ela es based on hei p edic i e s eng h.
The selec ion o co ela es ollows conside a ions o da a a ailabili y and is based on p e ious
esea ch, pa icula ly Alesina e al. (2021), Che y e al. (2014) and Van de Weide e al.
(2024). An impo an ca ea is ha his analysis should no be in e p e ed as causal. The
sole pu pose is o documen s ylized ac s ha can la e be used o heo e ically model o
es ima e empi ically he mechanisms behind local di e ences in in e gene a ional mobili y.
IV.1 Bi a ia e associa ions
This sec ion explo es he bi a ia e ela ionship be ween indica o s o IGM and a b oad
se o a iables, including income inequali y, go e nance, public in es men , educa ion, and
18
heal h se ices, among o he s. The ocus is on ela i e mobili y measu ed as one minus
he eg ession coe icien o simplici y and because i is a guably he mos widely used
indica o . Howe e , a summa y able wi h esul s o he emaining indica o s can be ound
in he Appendix (Table A5). The de ini ion o he co ela es and hei da a sou ces a e lis ed
in he Appendix (see Table A1). To allow o a lag be ween he con ex ual en i onmen and
he obse ed le el o mobili y, all a iables a e measu ed in he yea 2010.25
Figu e 4 epo s he coe icien s om eg essions wi h ela i e mobili y (s anda dized o ha e
mean 0 and a iance 1) as he dependen a iable and a gi en co ela e (s anda dized o ha e
mean 0 and a iance 1) as he independen a iable, which a e labeled as uncondi ional,
and hen he same eg ession con olling o a e age educa ion o he olde coho s.26 This
analysis p o ides an ini ial unde s anding o how hese ac o s ela e o IGM be o e applying
a selec ion me hod based on hei p edic i e abili y. The uncondi ional es ima es p o ide
a desc ip i e pe spec i e on he aw co ela ions be ween mobili y and each a iable, while
he condi ional es ima es help isola e he e ec abo e and beyond he a e age educa ional
a ainmen o pa en s.
Income dis ibu ion. Highe le els o inequali y a e o en associa ed wi h lowe mobil-
i y, as economic ad an ages and disad an ages pe sis ac oss gene a ions (Co ak 2013). I
include indica o s such as he Gini index, speci ic income quan iles (10 h, 50 h, 90 h, 95 h),
and income a ios (90/10, 90/50, 50/10) o cap u e di e en dimensions o inequali y. The
hypo hesis is ha g ea e income dispa i y, pa icula ly be ween he iches and poo es
households, limi s access o high-quali y educa ion and economic oppo uni ies, he eby e-
ducing mobili y.
I ind ha he Gini index, 90 h quan ile, 95 h quan ile, 90/10 a io, and 90/50 a io a e all
nega i ely and signi ican ly co ela ed wi h ela i e mobili y a he 5% le el. These esul s
sugges ha in Chile, highe in e gene a ional mobili y in educa ion is mo e s ongly associ-
25The excep ion in e ms o yea o measu emen is popula ion, which is compu ed using Census 2017.
26Following he app oach used in Alesina e al. (2021).
19
Figu e 4: Co ela es o ela i e mobili y (1 −β) a he commune-le el
Gini index
A g. labo ea nings
10 h quan ile
50 h quan ile
90 h quan ile
95 h quan ile
90 h o e 10 h
90 h o e 50 h
50 h o e 10 h
A ea
Dis . o eg. capi al
Densi y
Popula ion
Municipal quali y
Pe sonnel emale sha e
Felonies
Budge
To al expendi u e
Social expendi u e
Educa ional expendi u e
S uden s in public schools
S uden s pe eache
Seconda y es sco es
P ima y es sco es
Nu ses
Doc o s
In an mo ali y
Wa e ne wo k
-0.6 -0.4 -0.2 0 0.2 0.4 0.6
Uncondi ional
Condi ional on a g. educa ion o olde coho s
No es: The igu e epo s he coe icien s om eg essions wi h ela i e mobili y a he commune
le el measu ed as one minus he eg ession coe icien (s anda dized o ha e mean 0 and a iance
1) as he dependen a iable and a gi en co ela e (s anda dized o ha e mean 0 and a iance
1) as he independen a iable (in g een, labeled as uncondi ional), and he coe icien s om he
same eg ession con olling o a e age educa ion o he olde coho s (in o ange). 95% con idence
in e als a e included.
a ed wi h lowe le els o income inequali y in he uppe hal o he income dis ibu ion. This
con as s wi h indings om Co ak (2020), which show ha in Canada, mobili y in income is
mo e associa ed wi h inequali y in he lowe hal o he income dis ibu ion. Howe e , hese
esul s align wi h coun y-le el e idence epo ed by Na ayan e al. (2018), showing ha
income inequali y is posi i ely associa ed wi h in e gene a ional pe sis ence in educa ion,
meaning i is nega i ely associa ed wi h ela i e mobili y. These indings sugges ha he
well-documen ed ela ionship be ween inequali y and mobili y a he in e na ional le el may
also hold wi hin coun ies (see Figu e A13 in he Appendix).27
27I is wo h no ing ha he adminis a i e da ase used o cons uc measu es o income inequali y,
which comes om he unemploymen insu ance sys em, only conside s he o mal sec o . This could lead o
20
To complemen his analysis, I also examine he ela ionship be ween ela i e mobili y and
inequali y in educa ion among pa en s (indi iduals aged 40–60 a he ime o he Census).
This exe cise is equi alen o cons uc ing a “G ea Ga sby cu e” o educa ion wi hin
Chile, and I ind a nega i e ela ionship be ween educa ional inequali y and mobili y. This
sugges s ha he ela ionship documen ed ac oss coun ies by he Na ayan e al. (2018) also
holds wi hin coun ies (see Figu e A14 in he Appendix).
Demog aphic and geog aphic cha ac e is ics. Spa ial and demog aphic ac o s can
in luence mobili y by a ec ing access o educa ion, labo ma ke s, and public se ices. I
conside a iables such as popula ion size, densi y, commune a ea, and dis ance o he e-
gional capi al. The hypo hesis is ha la ge and mo e u banized communes, o hose close
o economic cen e s, may o e be e educa ional esou ces and job oppo uni ies, leading o
highe mobili y. Howe e , none o hese demog aphic a iables exhibi a s a is ically signi -
ican ela ionship wi h mobili y. This sugges s ha wi hin-coun y a ia ions in popula ion
dis ibu ion and geog aphy do no play as s ong a ole in educa ional mobili y.
Ins i u ional and go e nance quali y. The quali y o local go e nance and ins i u ional
s eng h may a ec mobili y by in luencing he e iciency o public se ice deli e y. I include
indica o s such as municipal go e nance quali y, emale ep esen a ion in he public sec o ,
and elonies. The hypo hesis is ha be e go e nance and lowe c ime le els may c ea e
an en i onmen mo e conduci e o highe in e gene a ional mobili y. I do no ind a signi i-
can ela ionship be ween hese go e nance indica o s and mobili y. This may indica e ha
a ia ions in local go e nmen pe o mance ac oss Chile a e no subs an ial enough o d i e
di e ences in mobili y ou comes.
Public spending and social in as uc u e. Public in es men in educa ion and social
se ices can mi iga e economic disad an ages and enhance mobili y. I analyze municipal
budge s, o al public expendi u e, social expendi u e, and educa ional expendi u e. The
hypo hesis is ha highe public spending, pa icula ly on educa ion, should be posi i ely
an unde es ima ion o income inequali y, as in o mal wo ke s a e excluded.
21
He z, T., Jayasunde a, T., Pi aino, P., Selcuk, S., Smi h, N. & Ve ashchagina, A. (2007),
‘The Inhe i ance o Educa ional Inequali y: In e na ional Compa isons and Fi y-Yea
T ends’, The B.E. Jou nal o Economic Analysis Policy 7(2).
Hilge , N. (2016), ‘The G ea Escape: In e gene a ional Mobili y in he Uni ed S a es Since
1940’, NBER Wo king Pape (21217).
Mazumde , B. (2016), Es ima ing he in e gene a ional elas ici y and ank associa ion in he
uni ed s a es: O e coming he cu en limi a ions o ax da a, in ‘Inequali y: Causes and
consequences’, Vol. 43, Eme ald g oup publishing limi ed, pp. 83–129.
Mizala, A. & To che, F. (2012), ‘B inging he schools back in: he s a i ica ion o educa-
ional achie emen in he chilean ouche sys em’, In e na ional Jou nal o Educa ional
De elopmen 32(1), 132–144.
Munoz, E. (2024), ‘The Geog aphy o In e gene a ional Mobili y in La in Ame ica and he
Ca ibbean’, Econom´ıa LACEA Jou nal 23, 333–354.
Munoz, E. & Si a egna, M. (2023), ‘When Measu e Ma e s: Co esidence Bias and In e -
gene a ional Mobili y Re isi ed’, IDB Wo king Pape 01469.
Na ayan, A., Van de Weide, R., Cojoca u, A., Lakne , C., Redaelli, S., Ge szon Mahle ,
D., Ramasubbaiah, R. & Thewissen, S. (2018), Fai P og ess?: Economic Mobili y Ac oss
Gene a ions A ound he Wo ld, The Wo ld Bank.
Neidh¨o e , G., Se ano, J. & Gaspa ini, L. (2018), ‘Educa ional Inequali y and In e gene a-
ional Mobili y in La in Ame ica: A New Da abase’, Jou nal o De elopmen Economics
134, 329–349.
Nunez, J. I. & Mi anda, L. (2010), ‘In e gene a ional Income Mobili y in a Less-De eloped,
High-Inequali y Con ex : The Case o Chile’, The B.E. Jou nal o Economic Analysis
Policy 10(1).
28
O eopoulos, P. & Pe onije ic, U. (2013), ‘Making college wo h i : A e iew o esea ch on
he e u ns o highe educa ion’.
Sapelli, C. (2016), Cap 3: La Mo ilidad In e gene acional de la Educacion en Chile, in ‘Chile
: ¿M´as Equi a i o? Una mi ada a la dinamica social del Chile de aye , hoy y ma˜nana’.
To che, F. (2005), ‘Unequal Bu Fluid: Social Mobili y in Chile in Compa a i e Pe spec i e’,
Ame ican Sociological Re iew 70, 422–450.
To che, F. (2021), Educa ional Mobili y in he De eloping Wo ld, in V. I e sen, A. K -
ishna & K. Sen, eds, ‘Social Mobili y in De eloping Coun ies: Concep s, Me hods, and
De e minan s’, Ox o d Uni e si y P ess.
Van de Weide, R., Lakne , C., Ge szon Mahle , D., Na ayan, A. & Ramasubbaiah, R.
(2024), ‘In e gene a ional Mobili y a ound he Wo ld: A New Da abase’, Jou nal o De-
elopmen Economics 166, 103167.
29
Appendices
The appendix p o ides addi ional ables and igu es, and o he ele an in o ma ion.
Table A1 lis s he se o co ela es ha I use oge he wi h a sho desc ip ion and da a
sou ces.
Table A2 epo s all he indica o s compu ed by sub-popula ions (male s. emale, indigenous
s. non-indigenous, and u ban s. u al).
Table A3 epo s some desc ip i e s a is ics o he es ima es o IGM a he le el o commune.
Figu e A1 plo s di e en measu es o in e gene a ional mobili y in educa ion a coun y-le el
highligh ing whe e Chile alls ela i e o La in Ame ica and he Ca ibbean and he wo ld.
Figu e A2 plo s di e en measu es o in e gene a ional mobili y in educa ion o Chile com-
pa ed o simple a e ages o egions o i e di e en coho s.
Figu e A3 displays an his og am wi h he dis ibu ions o educa ional a ainmen o pa en s
and child en.
Figu e A4 displays he e olu ion o mobili y ac oss bi h coho s in ecen li e a u e e sus
my es ima e.
Figu e A5 plo s he a e age co esidence a e agains IGM indica o s a he commune le el.
Figu e A6 plo s he a e age co esidence a e by le el o educa ion.
Figu e A7 plo s he a e age co esidence a e by age.
Figu e A8 displays he ansi ion p obabili ies be ween educa ional a ainmen o pa en s
and child en (classi ied in o h ee ca ego ies).
Figu e A9 shows he cumula i e dis ibu ion o he sample size by commune.
Figu e A10 displays he dis ibu ion o all he measu es a commune-le el.
Figu e A11 displays sca e plo s compa ing indica o s o mobili y ( ha use quin iles) using
coun y le el dis ibu ion o educa ional a ainmen s. local dis ibu ion.
Figu e A15 maps he le el o educa ional in e gene a ional mobili y a he commune le el
sepa a ing he coun y in o no h, me opoli an egion, and sou h.
30
Figu e A12 epo s he esul s o he co ela ions wi h a se o a iables using all he measu es
o IGM.
Figu e A13 shows a binsca e plo be ween ela i e mobili y and income inequali y measu ed
wi h he Gini coe icien a he commune le el.
Figu e A14 shows a binsca e plo be ween ela i e mobili y in educa ion and educa ional
inequali y measu ed wi h he s anda d de ia ion o yea s o schooling a he commune le el.
31
Table A1: Co a ia es
Label Sou ce Desc ip ion
Gini Index UID Gini Index
A e age ea nings UID A e age ea nings in he o mal sec o
10 h quan ile UID 10 h pe cen ile o ea nings in he o mal sec o
50 h quan ile UID 50 h pe cen ile o ea nings in he o mal sec o
90 h quan ile UID 90 h pe cen ile o ea nings in he o mal sec o
95 h quan ile UID 95 h pe cen ile o ea nings in he o mal sec o
Ra io 90-10 UID Ra io 90 h o 10 h pe cen ile o ea nings in he o mal sec o
Ra io 90-50 UID Ra io 90 h o 50 h pe cen ile o ea nings in he o mal sec o
Ra io 50-10 UID Ra io 50 h o 10 h pe cen ile o ea nings in he o mal sec o
A ea SINIM Log o he o al su ace o commune
Dis ance o egional capi al SINIM Log o he dis ance be ween he commune and he egional capi al
Popula ion densi y pe km2 SINIM Log o popula ion densi y pe km2 by commune
Popula ion SINIM Log o commune’s es ima ed popula ion in June 2012
Municipal p o essionaliza ion SINIM Sha e o college educa ed wo ke s in he municipali y
Female Sha e in Municipali y SINIM Sha e o emale wo ke s o e he o al wo ke s in pe sonnel o he municipali y
C imes CEAD Log o he numbe o c imes wi h g ea e social conno a ion
Budge a ailabili y SINIM Log o commune’s budge a ailabili y pe capi a
To al expendi u e SINIM Log o commune’s o al expendi u e pe capi a
Social expendi u e SINIM Log o he commune’s o al expendi u e in he social p og ams a ea pe capi a
Educa ion expendi u e SINIM Log o he commune’e o al expendi u e educa ion p og ams
S uden s in public schools ACE Numbe o s uden s en olled in public schools o e o al en ollmen
S uden s pe eache SINIM Log o s uden s pe eache a io in he municipal educa ion sys em
S anda ized es - seconda y ACE A e age sco e be ween ma h and language in SIMCE aken in high school
S anda ized es - p ima y ACE A e age sco e be ween ma h and language in SIMCE aken in 4 h g ade
Nu ses by 100K inhabi an s SINIM Log o numbe o nu ses by 100.000 inhabi an s wi hin he commune
Doc o s by 100K inhabi an s SINIM Log o numbe o doc o s by 100.000 inhabi an s wi hin he commune
In an mo ali y a e SINIM Numbe o child en unde 1 yea o age who die o e e y 1.000 li e bi hs
Wa e ne wo k SINIM Pe cen age o homes connec ed o d inking wa e ne wo k in he commune
Pa en al educa ion Census A e age educa ion o indi idual olde han 24 bu younge han 66
Unemploymen insu ance da abase (UID) can be accessed a :
h ps://www.spensiones.cl/apps/bdp/index.php.
Na ional sys em o municipal in o ma ion (SINIM) can be accessed a :
h p://da os.sinim.go .cl/da os municipales.php.
Cen e o c ime s udies and analysis (CEAD) can be accessed a :
h p://cead.spd.go .cl/es adis icas-delic uales/.
Resea ch uni , educa ion quali y agency da a (ACE) can be accessed a :
h ps://in o maciones adis ica.agenciaeducacion.cl/#/bases.
Census 2017 da a can be eques ed om he Na ional Ins i u e o S a is ics a :
h ps://www.ine.cl.
32
Table A2: IGM a coun y-le el o subg oups
Male Female Non-indigenous Indigenous U ban Ru al
α9.049 10.129 9.535 9.881 9.622 9.476
β0.688 0.742 0.710 0.748 0.717 0.718
¯
Y11.126 11.123 11.260 10.247 11.336 9.203
¯y≥0.628 0.707 0.663 0.690 0.657 0.754
ρ0.624 0.658 0.640 0.676 0.652 0.649
P15 0.068 0.108 0.089 0.082 0.092 0.071
P11 0.419 0.310 0.365 0.367 0.359 0.390
P55 0.331 0.378 0.357 0.306 0.353 0.358
The able epo s coun y-le el es ima es o absolu e mobili y, ela i e mobili y (1 −β), a e age
pa en s’ educa ion, sha e o child en wi h highe educa ion han pa en s, ela i e mobili y (1 −ρ),
ags o iches, in e gene a ional low, and in e gene a ional high, espec i ely, all compu ed by
subg oup. A desc ip ion o he measu es can be ound in Table 2.
Table A3: Desc ip i e s a is ics o IGM a commune-le el
Mean SD Min Max N
α9.79 0.66 7.16 11.73 330
1−β0.74 0.05 0.54 0.97 330
¯
Y10.00 1.19 6.13 14.50 330
¯y≥0.71 0.07 0.48 0.90 330
1−ρ0.68 0.06 0.50 0.96 330
P15 0.09 0.03 0.02 0.22 312
P11 0.36 0.06 0.09 0.57 313
P55 0.33 0.05 0.10 0.46 190
The able epo s desc ip i e s a is ics o es ima es o absolu e mobili y, ela i e mobili y (1 −β),
a e age pa en s’ educa ion, sha e o child en wi h highe educa ion han pa en s, ela i e mobili y
(1−ρ), ags o iches, in e gene a ional low, and in e gene a ional high, espec i ely, all o hem a
he commune-le el. I omi es ima es wi h less han 50 obse a ions. A desc ip ion o he measu es
can be ound in Table 2.
33
Figu e A1: Chile ela i e o he wo ld in e ms o educa ional IGM
.2 .4 .6 .8 1
1-be a
Chile La in Ame ica & Ca ibbean O he coun ies
(a) Rela i e mobili y (1 −β)
.2 .4 .6 .8 1
1- ho
Chile La in Ame ica & Ca ibbean O he coun ies
(b) Rela i e mobili y (1 −ρ)
0 .2 .4 .6 .8 1
Sha e (weakly)
Chile La in Ame ica & Ca ibbean O he coun ies
(c) Sha e o s uden s wi h highe educa ion han pa en s (sha e)
.05 .1 .15 .2 .25
BHQ4
Chile La in Ame ica & Ca ibbean O he coun ies
(d) Di ec ional IGM - P obabili y child om bo om hal ends up in Q4 (BHQ4)
.2 .3 .4 .5 .6
Q4Q4
Chile La in Ame ica & Ca ibbean O he coun ies
(e) Di ec ional IGM - P obabili y child om Q4 ends up in Q4 (Q4Q4)
.25 .3 .35 .4 .45
BHQ1
Chile La in Ame ica & Ca ibbean O he coun ies
( ) Di ec ional IGM - P obabili y child om bo om hal ends up in Q1 (BHQ1)
Sou ce: Elabo a ion by he au ho wi h da a om Na ayan e al. (2018).
34
Figu e A2: Mobili y in Chile e sus a e age by egion o i e coho s
(a) Rela i e mobili y (1 −β)(b) Rela i e mobili y (1 −ρ)
(c) Sha e wi h highe educa ion han pa en s (d) P ob. child om bo om hal ends up in Q4
(e) P obabili y child om Q4 ends up in Q4 ( ) P ob. child om bo om hal ends up in Q1
Sou ce: Elabo a ion by he au ho wi h da a om Na ayan e al. (2018). Regional a e ages a e
unweigh ed. Regions a e EAP: Eas Asia & Paci ic; ECA: Eu ope and Cen al Asia; HI: High
income; LAC: La in Ame ica and he Ca ibbean; MENA: Middle Eas and No h A ica; SA:
Sou h Asia; SSA: Sub-saha ian A ica.
35
Figu e A3: His og am o educa ion
0 .1 .2 .3 .4
Densi y
0 5 10 15
Yea o schooling
Child en Pa en s
36
Figu e A4: Own es ima es e sus ecen li e a u e a he coun y le el
(a) Rela i e mobili y (1 −β) om Na ayan
e al. (2018) and own es ima e.
(b) Rela i e mobili y (1 −β) om Neidh¨o e
e al. (2018) and own es ima e.
(c) Rela i e mobili y (1 −ρ) om Na ayan
e al. (2018) and own es ima e.
(d) Rela i e mobili y (1 −ρ) om Neidh¨o e
e al. (2018) and own es ima e.
The igu e shows es ima es o in e gene a ional educa ional mobili y ob ained om eg essing chil-
d en yea s o schooling agains pa en s’ yea s o schooling, and he Pea son co ela ion coe icien
be ween he same wo a iables. Na ayan e al. (2018) uses CASEN su ey while Neidh¨o e e al.
(2018) also uses La inoba ome o su ey (LBM). The o me uses 10-yea coho s, he la e uses
4-yea coho s ( he mos ecen one is 1992-1995), and my es ima e uses indi iduals app oxima ely
bo n be ween yea s 1991-1995. The las ou coho s using LBM su ey con ain smalle samples
(831, 413, 179, and 24 obse a ions), and hence a e somewha un eliable.
37
Figu e A11: Compa ison o indica o s using coun y le el dis ibu ion o educa ional
a ainmen s. local dis ibu ion
(a) Rags o iches (P1,5)(b) In e gene a ional low (P1,1)
(c) In e gene a ional high (P5,5)
The igu e compa es es ima es o ags o iches, in e gene a ional low, and in e gene a ional high
measu es compu ed using quin iles based on coun y-le el educa ional a ainmen e sus commune-
le el a ainmen (deno ed local). Each uses a sample o indi iduals o age 21-25 omi ing communes
wi h less han 50 indi iduals. Fo de ails abou he indica o s see Table 2.
44
Figu e A12: Co ela es o he IGM a he commune-le el (all he indica o s)
(a) Uncondi ional (b) Condi ional on educa ion o old coho s
45
Figu e A13: In e gene a ional mobili y in educa ion s. income inequali y
.7 .72 .74 .76 .78
Rela i e IGM in educa ion
.25 .3 .35 .4 .45
Gini coe . o indi iduals ages 18-60 (labo ea nings)
No es: The igu e shows a binsca e plo be ween ela i e IGM (measu ed as one minus he
eg ession coe icien be ween child’s yea s o schooling agains pa en s’ yea s o schooling) and
he Gini coe icien compu ed using labo ea nings in 2010 o indi iduals ages 18-60. Educa ional
a ainmen is censo ed a 15 and he sample includes indi iduals wi h age be ween 21 and 25.
Communes wi h less han 50 obse a ions a e no included.
46
Figu e A14: In e gene a ional mobili y in educa ion s. inequali y in educa ion
Pea son's co ela ion coe icien : -.123
.64 .66 .68 .7 .72
Rela i e IGM in educa ion
3.2 3.4 3.6 3.8 4 4.2
S anda d de ia ion o yea s o schooling (age 40-60)
No es: The igu e shows a binsca e plo be ween ela i e IGM (measu ed as one minus he Pea son
co ela ion coe icien be ween child’s yea s o schooling agains pa en s’ yea s o schooling) and
he s anda d de ia ion o yea s o schooling compu ed using indi iduals ages 40-60 ha a e used
as pa en s. Educa ional a ainmen is censo ed a 15. Communes wi h less han 50 obse a ions
a e no included.
47
Table A4: Region-le el es ima es o IGM S a is ics
Region P1,5Plocal
1,5P1,1Plocal
1,1P5,5Plocal
5,5
Ta apac´a 0.10 0.11 0.37 0.33 0.32 0.31
An o agas a 0.08 0.10 0.40 0.36 0.31 0.33
A acama 0.07 0.10 0.40 0.34 0.29 0.36
Coquimbo 0.07 0.08 0.38 0.34 0.32 0.31
Valpa a´ıso 0.09 0.10 0.35 0.36 0.34 0.33
Libe ado Gene al Be na do O’Higgins 0.10 0.10 0.34 0.34 0.31 0.31
Maule 0.09 0.09 0.36 0.37 0.32 0.31
Biob´ıo 0.11 0.09 0.33 0.36 0.37 0.32
A aucan´ıa 0.07 0.07 0.38 0.39 0.37 0.36
Los Lagos 0.07 0.08 0.41 0.38 0.31 0.32
Ays´en del Gene al Ca los Ib´a˜nez del Campo 0.05 0.09 0.44 0.35 0.23 0.32
Magallanes y de la An ´a ica Chilena 0.10 0.10 0.30 0.31 0.30 0.33
Me opoli ana de San iago 0.09 0.09 0.37 0.36 0.36 0.36
Los R´ıos 0.06 0.06 0.38 0.37 0.34 0.35
A ica y Pa inaco a 0.14 0.14 0.28 0.30 0.31 0.32
˜
Nuble 0.11 0.09 0.33 0.35 0.36 0.27
The able epo s egion-le el es ima es o ags o iches, in e gene a ional low, and in e gene a-
ional high (a desc ip ion o he measu es can be ound in Table 2). I compa es measu es ha
assign indi iduals in o quin iles using he dis ibu ion o educa ional a ainmen a he coun y
le el wi h measu es ha use he dis ibu ion o each egion ( hose wi h he supe sc ip “local”).
48
Table A5: Co ela es o in e gene a ional mobili y. All he indica o s.
Rela i e
mobili y
(1 −β)
Absolu e
mobili y
(α)
A e age
educa ion
(¯
Y)
Rela i e
mobili y
(1 −ρ)
Abo e
pa en s
(¯y≥)
Rags o
iches
P(1,5)
In e gen.
low
P(1,1)
In e gen.
high
P(5,5)
Gini index ⊗ ⊗ × ⊗ ⊗ ◦
A g. labo ea nings × × × × ×
10 h quan ile ⊗ ⊗ × ◦
50 h quan ile × × × × ×
90 h quan ile ⊗ ⊗ × ⊗ ⊗ ⊗ ×
95 h quan ile ⊗ ⊗ × ⊗ ⊗ ⊗ ×
90 h o e 10 h ⊗ ⊗ × ⊗ × ◦ ×
90 h o e 50 h ⊗ ⊗ × ⊗ ⊗
50 h o e 10 h × ⊗ ⊗ × ◦
A ea × ⊗ ⊗
Dis . o eg. capi al × ◦ ⊗ ⊗
Densi y × ⊗ ×
Popula ion × ⊗ × × ◦
Municipal quali y (p o )
Pe sonnel emale sha e ⊗
Felonies × ⊗ ◦
Budge × ⊗ ⊗ × ×
To al expendi u e × ⊗ ×
Social expendi u e ⊗ ⊗
Educa ional expendi u e × ⊗
S uden s in public schools × ⊗ × × ⊗ ◦
S uden s pe eache ⊗ ⊗ ⊗ ⊗ × ×
Seconda y es sco es × × × × ⊗ ⊗
P ima y es sco es ⊗ ◦ ×
Nu ses × × ×
Doc o s ⊗ ⊗ ⊗ ×
In an mo ali y
Wa e ne wo k ⊗ ⊗ ×
No es: The able epo s he s a is ical signi icance (a he 5% le el) o he coe icien s om eg es-
sions wi h an indica o o in e gene a ional mobili y a he commune le el (s anda dized o ha e
mean 0 and a iance 1) as he dependen a iable and a gi en co ela e (s anda dized o ha e mean
0 and a iance 1) as he independen a iable (labeled as uncondi ional model), and he coe icien s
om he same eg ession con olling o a e age educa ion o he olde coho s (labeled as condi-
ional model). ×i s a is ically signi ican in uncondi ional model. ◦i s a is ically signi ican in
condi ional model. ⊗i s a is ically signi ican in bo h models.
49
Table A6: LASSO a iable selec ion ac oss IGM indica o s
Rela i e
mobili y
(1 −β)
Absolu e
mobili y
(α)
A e age
educa ion
(¯
Y)
Rela i e
mobili y
(1 −ρ)
Abo e
pa en s
(¯y≥)
Rags o
iches
P(1,5)
In e gen.
low
P(1,1)
In e gen.
high
P(5,5)
Gini index ×
A g. labo ea nings
10 h quan ile
50 h quan ile × × ×
90 h quan ile
95 h quan ile
90 h o e 10 h × × × ×
90 h o e 50 h ×
50 h o e 10 h ×
A ea × × × × × ×
Dis . o eg. capi al × × × ×
Densi y × × × ×
Popula ion ×
Municipal quali y (p o ) × × × × ×
Pe sonnel emale sha e × × × × ×
Felonies × × ×
Budge × × × ×
To al expendi u e
Social expendi u e × × ×
Educa ional expendi u e ×
S uden s in public schools × × × × ×
S uden s pe eache × × ×
Seconda y es sco es × × ×
P ima y es sco es × × ×
Nu ses × × × × ×
Doc o s × × × ×
In an mo ali y × ×
Wa e ne wo k × × × × ×
No es: The able p esen s he a iables selec ed wi h a LASSO es ima ion using he op imal alue
o λ, highligh ing he se o co ela es ha emain nonze o a e egula iza ion o each indica o
o in e gene a ional mobili y.
50
Figu e A15: In e gene a ional educa ional mobili y wi hin Chile
No h
Me opoli an Region
Sou h
Rela i e mobili y
0.54 o 0.69 0.69 o 0.70 0.70 o 0.71 0.71 o 0.72 0.72 o 0.73 0.73 o 0.75 0.75 o 0.76 0.76 o 0.78 0.78 o 0.80 0.80 o 0.97 Missing
(a) Rela i e mobili y by commune - Chile, 2017
No es: The map plo s ela i e IGM measu ed as one minus he eg ession coe icien (by commune) be ween child’s yea s o schooling
(using age be ween 21 and 25) agains pa en s’ yea s o schooling. Educa ional a ainmen is censo ed a 15. Communes wi h less han
50 obse a ions a e le as missing.
51
