F os , Ma ga e ; Kim, Sangeun; Sca ascini, Ca los G.; Zamo a, Paula; Zechmeis e ,
Elizabe h J.
Wo king Pape
Disas e and poli ical us : A na u al expe imen om he
2017 Mexico Ci y ea hquake
IDB Wo king Pape Se ies, No. IDB-WP-1192
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: F os , Ma ga e ; Kim, Sangeun; Sca ascini, Ca los G.; Zamo a, Paula;
Zechmeis e , Elizabe h J. (2024) : Disas e and poli ical us : A na u al expe imen om he 2017
Mexico Ci y ea hquake, IDB Wo king Pape Se ies, No. IDB-WP-1192, In e -Ame ican De elopmen
Bank (IDB), Washing on, DC,
h ps://doi.o g/10.18235/0005511
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/289909
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-nc-nd/3.0/igo/legalcode
Disas e and Poli ical T us :
A
Na u al Expe imen om he 2017 Mexico Ci y
Ea hquake
Ma ga e F os
Sangeun Kim
Ca los Sca ascini
Paula Zamo a
Elizabe h J. Zechmeis e
IDB WORKING PAPER SERIES Nº IDB-WP-1192
Janua y 2024
Depa men o Resea ch and Chie Economis
In e -Ame ican De elopmen Bank
Janua y 2024
Disas e and Poli ical T us :
A
Na u al Expe imen om he 2017 Mexico Ci y Ea hquake
Ma ga e F os *
Sangeun Kim**
Ca los Sca ascini***
Paula Zamo a****
Elizabe h J. Zechmeis e *****
* Uni e si y o Rhode Island
** Aspen Ins i u e
*** In e -Ame ican De elopmen Bank
‡ Texas A&M Uni e si y
‡‡ Vande bil Uni e si y
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
Disas e and poli ical us : a na u al expe imen om he 2017 Mexico Ci y
ea hquake / Ma ga e F os , Sangeun Kim, Ca los Sca ascini, Paula Zamo a,
Elizabe h J. Zechmeis e .
p. cm. — (IDB Wo king Pape Se ies ; 1192)
Includes bibliog aphic e e ences.
1. Ea hquake elie -Poli ical aspec s-Mexico-Econome ic models. 2. Na u al
disas e s-Poli ical aspec s-Mexico-Econome ic models. 3. T us -Poli ical aspec s-
Mexico-Econome ic models. I. F os , Ma ga e . II. Kim, Sangeun. III. Sca ascini,
Ca los G., 1971- IV. Zamo a, Paula. V. Zechmeis e , Elizabe h J., 1972- VI. In e -
A
me ican De elopmen Bank. Depa men o Resea ch and Chie Economis . VII.
Se ies.
IDB-WP-1192
Copy igh © In e -Ame ican De elopmen Bank. This wo k is licensed unde a C ea i e Commons IGO 3.0 A ibu ion-
NonComme cial-NoDe i a i es (CC-IGO BY-NC-ND 3.0 IGO) license (h p://c ea i ecommons.o g/licenses/by-nc-nd/3.0/igo/
legalcode) and may be ep oduced wi h a ibu ion o he IDB and o any non-comme cial pu pose, as p o ided below. No
de i a i e wo k is allowed.
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 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 CC-IGO license.
Following a pee e iew p ocess, and wi h p e ious w i en consen by he In e -Ame ican De elopmen Bank (IDB), a e ised
e sion o his wo k may also be ep oduced in any academic jou nal, including hose indexed by he Ame ican Economic
Associa ion's EconLi , p o ided ha he IDB is c edi ed and ha he au ho (s) ecei e no income om he publica ion. The e o e,
he es ic ion o ecei e income om such publica ion shall only ex end o he publica ion's au ho (s). Wi h ega d o such
es ic ion, in case o any inconsis ency be ween he C ea i e Commons IGO 3.0 A ibu ion-NonComme cial-NoDe i a i es license
and hese s a emen s, he la e shall p e ail.
No e ha link p o ided abo e includes addi ional e ms and condi ions o he license.
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 .
h p://www.iadb.o g
2024
Abs ac
Poli ical us is ounda ional o democ a ic legi imacy, ep esen a i e go e nance,
and he p o ision o e ec i e public policy. Va ious shocks can in luence his us ,
s ee ing coun ies on o posi i e o nega i e ajec o ies. This s udy examines whe he
na u al disas e s can impac gene al poli ical us and i disas e elie e o s can
mi iga e hese e ec s. We in es iga e he ela ionships be ween disas e , us , and aid
using no el su ey da a collec ed be o e and a e a 7.1-magni ude ea hquake s uck
Mexico Ci y in Sep embe 2017. Ou indings e eal ha he disas e esul ed in an
11% dec ease in gene al poli ical us . Addi ionally, we demons a e ha geog aphical
p oximi y o disas e elie e o s may coun e balance his decline in us . This s udy
con ibu es o he schola ship on he poli ics o disas e s and o e s policy implica ions,
highligh ing he ole o disas e assis ance in po en ially es o ing gene al poli ical us
a e a disas e .
JEL classi ica ions: H84, Q54, D72, Z13
Keywo ds: Poli ical us , Na u al disas e , Na u al expe imen , Aid elie , De elopmen
Da a collec ion was IRB app o ed by he Beha io al Sciences Commi ee o Vande bil Uni e si y’s
Ins i u ional Re iew Boa d. The au ho s decla e ha hey ha e no known compe ing inancial in e es s o
pe sonal ela ionships ha could ha e appea ed o in luence he wo k epo ed in his pape . Nei he he
IDB no any o he pa y had any ole in he s udy design, da a collec ion and analysis, decision o publish,
o p epa a ion o he manusc ip . The in o ma ion and opinions p esen ed he ein a e en i ely hose o he
au ho s, and no endo semen by he In e -Ame ican De elopmen Bank, i s Boa d o Execu i e Di ec o s,
o he coun ies hey ep esen is exp essed o implied.
1 In oduc ion
Poli ical us is essen ial o he well- unc ioning o democ acy. Rep esen a i e democ acy
es s on delega ion, which equi es us (U bina i and Wa en, 2008). T us inc eases public
demand o wel a e-enhancing policies and acili a es social cohesion, which in u n enhances
us (Kee e and Sca ascini, 2022), and us is c ucial in imes o disas e when p i a e-
public collabo a ion is necessa y o e ec i e eco e y (Ald ich, 2017). I ma e s, hen, i
disas e s unde mine us in public o icials p ecisely when i is mos needed. Schola ship
on pos -disas e us le els has mos o en ocused on incumben s and/o speci ic o ices
(Reinha d , 2015,2019), go e nmen s (Han e al., 2011; Nicholls and Picou, 2013), and
social us (Bai and Li, 2021; Becche i e al., 2017; Ca lin e al., 2014b; S ephane, 2021).
We add o his body o esea ch by add essing he e ec o disas e on gene al poli ical us ,
and he condi ioning ole o disas e elie , in a compa a i ely less de eloped con ex .
Ou co e con en ion is ha la ge-scale disas e s lowe gene al poli ical us —meaning,
indi iduals’ con idence in he eliabili y and in eg i y o public o icials. We a gue us alls
pos -disas e ia wo pa hs: s essing s a e capaci y and inc easing a enues o co up ion.
Fi s , disas e s signal compe ence and may o e whelm s a e capaci y. Subsequen nega i e
expe iences ough o a ec assessmen s o go e nmen bene olence, compe ence, and he
abili y o deli e o ci izens (Olson and Gaw onski, 2010). Addi ionally, he chao ic a e -
ma h o a disas e and he need o as disbu semen s c ea e oppo uni ies o co up ion
(Nikolo a and Ma ino , 2017; Yamamu a, 2014) a a ime when he public is highly sensi i e
o mal easance (Gaw onski and Olson, 2000). The e o e, especially whe e weak ins i u ions
p e ail, gene al poli ical us would decline in he a e ma h o a majo na u al disas e .
Bu can his all in us be p e en ed? A p omising a enue o heo izing on he ela-
ionship be ween disas e and gene al poli ical us comes om esea ch sugges ing ha he
p o ision o wel a e mi iga es agains nega i e u ns in poli ical e alua ions (Laza e e al.,
2014) and social o in e pe sonal us (Ca lin e al., 2014b; De Juan e al., 2020; Pe o a
and Ros old, 2024). In heo y, aid ca ies he po en ial o he s a e o demons a e capac-
i y and bene icence. Some schola s ha e shown ha aid can gene a e incumben suppo
(Healy and Malho a, 2009; Bech el and Hainmuelle , 2011; Gallego, 2018; Laza e e al.,
2014). O he s, howe e , ind ha aid is insu icien o coun e inc eases in nega i e poli ical
e alua ions, e en when dis ibu ed ai ly (Cole e al., 2012; Hee sink e al., 2017). Mo eo e ,
aid in lows can be mishandled (Leeson and Sobel, 2008; Yamamu a, 2014) o ine ec i ely
dis ibu ed (Eichenaue e al., 2020; F ancken e al., 2012; Sobel and Leeson, 2006), leading
o declines in poli ical e alua ions. In sho , whe he aid can mi iga e agains cu en s ha
2
diminish us is an open ques ion in need o empi ical es ing.
We es ela ionships among disas e , us , and aid using da a ga he ed immedia ely
be o e and a e a 7.1-magni ude ea hquake s uck Mexico Ci y on Sep embe 19, 2017. The
ea hquake killed 369 people and inju ed app oxima ely 2,000 (Andone e al.,2017;Uni ed
S a es Geological Su ey,2017;Villegas and Ahmed,2017). To assess he consequences
o disas e o gene al poli ical us , we employ a no el design based on wo ounds o
su eys: one conduc ed immedia ely be o e (569 obse a ions) and ano he wo mon hs
a e he ea hquake (1,164 obse a ions) in he g ea e Mexico Ci y me opoli an egion.
These 1,164 indi iduals we e su eyed o ma ch he p e-ea hquake sample, wi h each p e-
ea hquake indi idual ma ched o a leas wo compa able (on loca ion, gende , and age)
pos -ea hquake indi iduals.
Ou i s objec i e is o assess he e ec o he ea hquake on poli ical us , measu ed
ia a composi e o ou ques ions on con idence in poli icians and ci il se an s. Al hough
unning a andomized con olled ial o exposu e o an ea hquake is no possible, he
ma ched (by design) sample suppo s a causal in e p e a ion o he analyses unde a se o
assump ions: ha he app oach c ea es wo o he wise homogeneous g oups and ea hquake
damage is independen o p e-disas e le els o us .1We u he es he obus ness o
ou ini ial esul s ia s a is ical ma ching. We conclude ha he disas e e en nega i ely
a ec ed gene al poli ical us : he ea hquake caused a subs an ial d op o 11% in gene al
poli ical us .
We hen conside he ques ion o disas e elie : a e age esul s may mask subs an ial
he e ogenei y i hose who ecei e disas e elie ha e hei gene al poli ical us es o ed
by his demons a ion o s a e e ec i eness and bene olence. We un analyses wi h wo
measu es o access o aid: subjec i e awa eness and objec i e loca ion da a. We ind a
posi i e, signi ican ela ionship be ween us and pe cei ed p oximi y o dis ibu ion cen e s
o ood, wa e , and o he essen ial i ems. P ecisely, hose who epo ed ha ing access o
such assis ance epo ed ela i ely highe le els o us . Al hough no as signi ican , we also
ind a posi i e ela ionship o analyses o ac ual p oximi y o a dis ibu ion cen e . Bo h
measu es ha e limi a ions. Al hough we con ol o poli ical a iables a he municipali y
le el be o e he ea hquake, subjec i e epo s may be endogenous o ac o s we a e no
conside ing. While we scou ed he web o ind aid cen e da a, he lis o ac ual loca ions
is likely incomple e. We conclude ha he esul s p o ide sugges i e e idence ha aid can
coun e he ad e se e ec s o a disas e on gene al poli ical us .
1We also assume no ele an composi ional changes in he popula ion in su eyed a eas.
3
The pape makes ou con ibu ions. Fi s , unlike p e ious s udies ocusing on pa icula
poli ical igu es o o ices, we examine he connec ion be ween disas e and gene al poli ical
us . This scope is ele an because, once low, gene al poli ical us is mo e di icul o
es o e han us in speci ic indi iduals’/o ices’ handling o a disas e (Le i and S oke ,
2000). Second, we add s ong e idence ha disas e dec eases us using a unique da ase
wi h ad an ageous iming and su ey design in which we ma ched pos -disas e esponden s
o hose in a p e-disas e su ey. Thi d, we ex end schola ship on he consequences o pos -
disas e wel a e p o ision beyond hei implica ions o incumben suppo and social us
o he domain o gene al poli ical us . We show ha , e en in a con ex in which e icien
and g a - ee aid dis ibu ion is a challenge, he es ablishmen o dis ibu ion cen e s—o en
h ough public-p i a e o NGO pa ne ships—con ains he po en ial o coun e declines in
gene al poli ical us . Fou h, we ex end he scope o s udies on disas e s and us om
he Uni ed S a es and Wes e n Eu ope o a compa a i ely less de eloped con ex . In his
con ex , baseline gene al us is low, and co up ion, slow deli e y, and he po en ial co-
op ing o aid deli e y by hi d pa ies o en complica e aid dis ibu ion. Concen a ing on a
de eloping coun y is impo an because mos p io s udies ha e ocused on con ex s whe e
gene al us is mode a e o high, as is he s a e’s capaci y—albei pe haps no i s willingness
o o ganiza ion— o deli e aid quickly and e icien ly.
2 Analy ical F amewo k
Mos s udies o pos -disas e poli ical us ha use sys ema ically-collec ed empi ical da a
concen a e on public con idence in incumben s in speci ic go e nmen o ices (Reinha d ,
2015,2019), go e nmen a di e en le els (Han e al., 2011; Nicholls and Picou, 2013), and
socie y (Ca lin e al., 2014b). Table A.1 in he Online Appendix Asumma izes mos o
hese s udies despi e no being an exhaus i e lis . O hese, Alb ech (2017) is closes o
ou s udy, as i examines us in gene al poli icians wi h mul iple ounds o c oss-coun y
su eys and, in e es ingly, inds no e idence o a connec ion be ween disas e and poli ical
us . Ne e heless, his c oss-na ional s udy includes compa a i ely mino disas e s (e.g.,
ex eme empe a u es) ha may ha e mo e ma ginal implica ions o us .
Acco ding o Ha din (1993, p.506), us is de ined as a h ee-pa ela ionship whe e
“A” us s “B” o pe o m “X.” Ou esea ch expands upon exis ing li e a u e by conside -
ing a b oade scope o “B” and “X.” Speci ically, we in es iga e he public’s gene al us
in poli icians and ci il se an s (B) conce ning hei commi men o keeping p omises—
4
eliabili y—and compliance wi h he law—in eg i y (X). In essence, ou ocus cen e s on he
public’s gene al poli ical us , whe e hey place hei us in he gene al poli ical commu-
ni y (public o icials) in ma e s o co e esponsibili ies ha include bu also ex end beyond
disas e elie .
We use se e al lines o schola ship in o a uni ied amewo k on he connec ions be ween
disas e and gene al poli ical us . We de ail a mic o-logic ha suppo s he expec a ion
ha gene al poli ical us will end o all a e a majo disas e . Then, we conside whe he
p o iding disas e aid om any sou ce may mi iga e ha d op. In his case, we s a e an
open expec a ion. On he one hand, aid a ailabili y should es o e pe cep ions o s a e
compe ence. The li e a u e has shown ha incumben s a e ewa ded o aid p o ision.
Howe e , because aid can be co up ed and ine ec i e, pa icula ly in de eloping coun ies,
he connec ion be ween disas e aid and gene al us is a om gua an eed.
Ou co e hesis is ha disas e s place s ess on poli ical us due o wo non- i al dy-
namics. The i s ela es o he dimension o us conce ning compe ence (Le i and S oke ,
2000). Na u al disas e s o e load sys ems, placing a signi ican bu den on s a e capaci y.
Disas e s dis up economies, dis u b in as uc u e, and diso ganize bu eauc acies, while
hey mul iply socie al demands (D u y and Olson, 1998; Olson, 2000; Schneide , 1992).
Wi h limi ed ime ho izons and a mo i a ion o ca e o public opinion, go e nmen s end
o unde -in es in p epa a ion (Healy and Malho a, 2009). The combina ion o p e-disas e
unde -p epa edness and pos -disas e dis up ions and demands isks lea ing s a es ulne a-
ble o appea ing incompe en ollowing na u al disas e s (Olson and Gaw onski, 2010).
The second dynamic ela es o he s a e’s ulne abili y o co up ion and accusa ions o
co up ion ollowing disas e . Pe cep ions o pe asi e co up ion ma e because hey signal
a lack o bene icence, which may shape poli ical us (Chang and Chu, 2006). Widesp ead
co up ion also unde mines poli ical us by hampe ing he s a e’s capaci y o deli e e-
sou ces o he people in need (La allée e al., 2008). T us wo hy agen s do no ac oppo -
unis ically e en when i is bene icial o hem (Kee e e al., 2020). Ne e heless, na u al
disas e s can po en ially inc ease oppo uni y and demand o co up ion, especially in de-
eloping coun y con ex s (Nikolo a and Ma ino , 2017; Yamamu a, 2014). Resea ch e eals
ha indi iduals a e highly sensi i e o co up ion in bad imes, including ollowing disas-
e s (Gaw onski and Olson, 2000; Olson and Gaw onski, 2010; Zechmeis e and Zizumbo-
Colunga, 2013). In b ie , o he ex en ha co up ion seeps in o pos -disas e dynamics
and pe cep ions, he e is eason o suspec ha poli ical us will decline a e a disas e .
Declines in us ’s wo co e dimensions—inc eased bu den and ulne abili y o co up ion—
5
who had no con idence in poli icians and ci il se an s be o e and a e he ea hquake. Less
han 25% o esponden s hough ha poli icians and ci il se an s ul illed hei p omises
o complied wi h he law be o e he ea hquake. This ac no only o e s a glimpse in o he
low le els o us in he coun y bu may also gene a e a ceiling e ec in ou es ima ions
o educed us .
Fo ou co e analysis, we conduc ed P incipal Componen Analysis (PCA) using he ou
su ey i ems: i) ul illmen o poli icians’ p omises, ii) ul illmen o ci il se an s’ p omises,
iii) law compliance o poli icians; and 4) law compliance o ci il se an s. All ou a iables
exhibi a high deg ee o posi i e co ela ion on he i s componen .11 Subsequen ly, we use
i as a la en a iable o “gene al poli ical us .”
Figu e 2: PCA: T us be o e and a e he Ea hquake
0 .5 1 1.5 2 2.5
Densi y
0 .2 .4 .6 .8 1
“P e” “Pos ”
Figu e 2shows he dis ibu ion o he dependen a iable be o e and a e he ea h-
quake once he i s componen o he PCA is aken as he gene al poli ical us a iable
and is no malized. The la en a iable anges om low (0) o high (1) us le els. The high
concen a ion o pos -ea hquake esponden s’ answe s on he le -hand side o he dis ibu-
ion indica es ha mo e people epo ed ha ing “ e y low" gene al poli ical us a e he
ea hquake han be o e he e en .
and a e he ea hquake (Wilcoxon ank-sum es : p- alue = 0.020).
11See Sec ion D.2 in he Online Appendix o de ails ega ding he P incipal Componen Analysis (PCA).
12
Table 1: E ec o he Ea hquake on Gene al Poli ical T us
Gene al poli ical us
(1) (2) (3) (4) (5) (6)
Ea hquake -0.029*** -0.029*** -0.029*** -0.029*** -0.029*** -0.029***
(0.010) (0.010) (0.010) (0.010) (0.010) (0.010)
Male -0.000 0.001 -0.000 -0.000 -0.000 0.001
(0.009) (0.009) (0.009) (0.009) (0.009) (0.009)
Age -0.002*** -0.002*** -0.002*** -0.002*** -0.002*** -0.002***
(0.000) (0.000) (0.000) (0.000) (0.000) (0.000)
Las yea o educa ion app o ed -0.002 -0.002 -0.002 -0.002 -0.002 -0.002*
(0.001) (0.001) (0.001) (0.001) (0.001) (0.001)
Numbe o adul s in he household -0.001 -0.001 -0.001 -0.001 -0.001 -0.002
(0.003) (0.003) (0.003) (0.003) (0.003) (0.003)
Numbe o child en in he household 0.002 0.002 0.002 0.002 0.002 0.002
(0.004) (0.004) (0.004) (0.004) (0.004) (0.004)
PRI: p e 0.031 0.069
(0.056) (0.073)
PRD: p e 0.146*** 0.167***
(0.053) (0.057)
Weal h: p e 0.013 0.087
(0.081) (0.118)
Risk A e sion: p e 0.003 -0.002
(0.004) (0.003)
Pa ience: p e 0.001 0.000
(0.003) (0.003)
Social T us : p e -0.023 -0.017
(0.036) (0.042)
Cons an 0.340*** 0.314*** 0.333*** 0.238** 0.405*** 0.358*
(0.028) (0.031) (0.049) (0.120) (0.103) (0.196)
Obse a ions 1,651 1,651 1,651 1,651 1,651 1,651
R-squa ed 0.039 0.043 0.039 0.039 0.039 0.044
No es: Clus e ed s anda d e o s a e in pa en heses. ∗p < 0.10,∗∗p < 0.05,∗∗∗p < 0.01. The a e age le els
o poli ical a ilia ion, isk a e sion, pa ience, and social us a e calcula ed a he municipali y le el be o e he
ea hquake.
We es ima e he e ec o he ea hquake on he gene al poli ical us a iable using
o dina y leas squa es. Table 1p esen s he esul s and suppo s he conclusion ha us
subs an ially dec eased a e he ea hquake, independen ly o he se o con ols included
in he eg essions. The i s model con ols o a iables unlikely o be a ec ed by he
ea hquake (i.e., gende , age, educa ion, and numbe o adul s and child en pe household)
o imp o e e iciency. Models (2) - (5) include poli ical a ilia ion, weal h, isk a e sion,
pa ience, and social us . Gi en ha he ea hquake may a ec hese a iables, we use p e-
ea men a e ages a he municipali y le el.12 Fo poli ical a ilia ion, we ocus on he wo
12We conside ed a e aging by clus e s, bu each clus e included only abou 10 esponden s, while mu-
13
main pa ies in his sub egion: PRI (Ins i u ional Re olu iona y Pa y) and PRD (Pa y
o he Democ a ic Re olu ion). Fo example, suppose 60% o he esponden s in a gi en
municipali y suppo ed PRI be o e he ea hquake. In ha case, we assign his pe cen age
o esponden s who esided in he municipali y du ing he pos -ea hquake su ey. All
models con ol o egional indica o a iables.
The e o e, he esul s suppo H1: he ea hquake educed he le el o gene al poli ical
us .13 The es ima ed e ec o he ea hquake on gene al us emains he same: -0.029,
which is equi alen o a d op o 11% (0.029/0.26). In e es ingly, he able also shows a
posi i e and signi ican co ela ion be ween p e-ea hquake suppo o he pa y o he
head o he Fede al Dis ic /Mexico Ci y a he ime o he ea hquake (PRD: p e) and he
e ec on gene al poli ical us .
Wha could be d i ing he d op in us ? We men ioned wo pa hs, one in ol ing s a e
capaci y and ano he one in ol ing co up ion. We do no ha e mic o-le el da a o es
hese mechanisms in de ail. Howe e , we do ha e da a on blame a ibu ion. Pe cep ions
o go e nmen c edibili y and in eg i y ough o be a ec ed i ci izens blame public ac o s
o he disas e ’s occu ence. The pos -ea hquake su ey shows ha 62.2% o esponden s
belie ed he ea hquake damage migh ha e been a e ed i building egula ions had been
en o ced.
6 Disas e Relie ’s Role in Rebuilding Poli ical T us
Thus a , he e idence suppo s he hypo hesis ha se e e disas e s unde mine poli ical us .
Howe e , he esul s p esen ed o he 2017 Mexico ea hquake may mask whe he disas e
elie in he a e ma h o he ea hquake, pa icula ly aid dis ibu ion cen e s, posi i ely
a ec ed poli ical us . I ha we e he case, he e a e wo impo an implica ions. Fi s ,
disas e elie would coun e ail he impac o he ea hquake. Second, he esul s om he
es o H1 would be a lowe bound o he ac ual e ec ; wi hou aid, he d op in us would
ha e been highe . Thus, his sec ion ocuses on one o he mos common o ms o pos -
ea hquake aid: dis ibu ion cen e s ( ables and en s) se up in a ec ed neighbo hoods o
indi iduals o pick up essen ial goods (e.g., wa e and ood). We in es iga e he ele ance
o dis ibu ion cen e s ia how close hey we e o esponden s’ households in wo ways: i)
nicipali ies a e aged abou 60 su ey esponden s.
13Addi ional analyses show ha he esul holds when con olling o he ac ual alues o poli ical suppo ,
weal h, pa ience, isk a e sion, and social us wi hou a e aging a he municipali y le el.
14
esponden s’ sel - epo s abou he p esence o dis ibu ion cen e s in hei neighbo hood,
and ii) ga he ing he loca ion da a o dis ibu ion cen e s.
We no e ha bo h measu es ha e s eng hs and weaknesses. Subjec i e assessmen s
a e help ul because hey measu e awa eness o he esou ce. I a dis ibu ion cen e exis s,
bu an indi idual is no awa e o i , in heo y, i should no a ec a i udes. Howe e ,
subjec i e assessmen s may be endogenous o ac o s ha shape poli ical us o subjec o
expec a ions o ecei e addi ional aid. We con ol o some po en ial con ounde s bu canno
elimina e his h ea . Objec i e indica o s a oid subjec i e biases bu ha e wo challenges:
indi iduals may no be awa e o hem, and i is impossible o de elop a comple e lis o all
aid dis ibu ion loca ions. The la e is because, in he chao ic a e ma h o a se e e disas e
like ha in Mexico, aid s a ions a y in size and du a ion, hey a e se up by many di e en
ac o s, and he e is no single au ho i a i e lis o he placemen and iming o all cen e s.
We assess he e idence o e ed by hese measu es wi h hese ca ea s in mind.
15
6.1 Sel - epo ed P oximi y o Dis ibu ion Cen e s
Table 2: Associa ion be ween Sel - epo ed Dis ibu ion Cen e s and Gene al Poli ical
T us
Gene al poli ical us
(1) (2) (3) (4) (5) (6)
Sel - epo ed DC 0.032** 0.030** 0.032** 0.032** 0.032** 0.030**
(0.013) (0.013) (0.013) (0.013) (0.013) (0.013)
Male 0.003 0.004 0.003 0.003 0.003 0.004
(0.011) (0.011) (0.011) (0.011) (0.011) (0.011)
Age -0.002*** -0.002*** -0.002*** -0.002*** -0.002*** -0.002***
(0.000) (0.000) (0.000) (0.000) (0.000) (0.000)
Las yea o educa ion app o ed -0.001 -0.001 -0.001 -0.001 -0.001 -0.002
(0.002) (0.002) (0.002) (0.002) (0.002) (0.002)
Numbe o adul s in he household 0.001 0.001 0.001 0.001 0.001 0.001
(0.003) (0.003) (0.003) (0.003) (0.003) (0.003)
Numbe o child en in he household 0.003 0.003 0.002 0.003 0.003 0.003
(0.004) (0.004) (0.004) (0.004) (0.004) (0.004)
PRI:p e 0.034 0.054
(0.065) (0.087)
PRD:p e 0.148** 0.192***
(0.065) (0.071)
Weal h: p e -0.057 0.014
(0.087) (0.143)
Risk A e sion: p e -0.002 -0.006
(0.005) (0.005)
Pa ience: p e 0.001 0.000
(0.003) (0.003)
Social T us : p e 0.012 -0.008
(0.037) (0.047)
Cons an 0.284*** 0.260*** 0.316*** 0.318* 0.252** 0.440*
(0.035) (0.037) (0.056) (0.163) (0.106) (0.254)
Obse a ions 1,069 1,069 1,069 1,069 1,069 1,069
R-squa ed 0.056 0.061 0.057 0.057 0.056 0.063
No es: Clus e ed s anda d e o s a e in pa en heses. ∗p < 0.10,∗∗p < 0.05,∗∗∗p < 0.01. The a e age le els
o poli ical a ilia ion, isk a e sion, pa ience, and social us a e calcula ed a he municipali y le el be o e he
ea hquake.
Table 2 epo s he esul s o sel - epo ed dis ibu ion cen e s on gene al poli ical us . This
able includes he same con ol a iables and egional ixed e ec s in he p e ious sec ion. We
conduc his analysis only o pos -ea hquake esponden s (p oximi y o dis ibu ion cen e
in o ma ion is i ele an be o e he ea hquake). We also no e ha he “ ea men ” (i.e.,
dis ibu ion cen e ) is no longe assumed o ha e been assigned andomly. The able shows
ha esponden s who epo ed ha ing (a leas ) one dis ibu ion cen e in hei neighbo hood
16
epo ed highe gene al poli ical us in poli icians and ci il se an s, and i s e ec size is
sizeable ( anging om 0.030 o 0.032, i ep esen s an inc ease o abou 12% wi h espec
o he p e-ea hquake mean). As he a e age ea men e ec o he ea hquake was -0.029,
hose who epo ed obse ing a dis ibu ion cen e in hei neighbo hood would appea o
ha e had hei gene al poli ical us es o ed o he p e-ea hquake le el.
6.2 P oximi y o he Objec i e Loca ion o Dis ibu ion Cen e s
While he sel - epo s on dis ibu ion cen e s cap u e he awa eness o he a ailabili y o
aid, hey may be con ounded wi h esponden s’ p io belie s abou he poli icians and ci il
se an s ( e e se causali y: hose who us mo e epo ha ing seen a dis ibu ion cen e .)
The e migh also be sel -selec ion— hose who needed hem mo e ook he ex a s eps o
loca e and use hem. To add ess hese conce ns, we es he obus ness o he esul s using
an objec i e measu e.
To loca e dis ibu ion cen e s, we ga he ed and c oss- e e enced in o ma ion om go -
e nmen , newspape , and social media sou ces epo ing on dis ibu ion cen e s es ablished
in Mexico Ci y in he wake o he ea hquake (see sec ion E.1 o de ailed sou ces). These
sou ces p o ided pa ial iden i ying in o ma ion o each dis ibu ion cen e , such as he
name, p o ide , and, in some cases, la i ude and longi ude o a dis ibu ion cen e . Some
in o ma ion was lacking, such as he s ee name, neighbo hood name, municipali y name,
and zone numbe . We illed in he missing in o ma ion o each dis ibu ion cen e using
publicly a ailable maps and go e nmen documen s on pos al in o ma ion and zoning in
Mexico Ci y. We checked each obse a ion wice, using wo di e en maps. We ook mos o
he p ima y sou ce ma e ials abou dis ibu ion cen e s om he Mexico Ci y go e nmen
websi e and he Collabo a i e Map “Rescue Mexico,” which hen we e co obo a ed wi h
news a icles om El Uni e sal,Animal Polí ico, and Mexican public ins i u ions’ Twi e
eeds.14 We could no loca e simila da a on dis ibu ion cen e s o he S a e o Mexico, so
ou analysis he e is limi ed o Mexico Ci y and, hus, o ewe su ey esponden s. Figu e 3
shows he loca ion o he dis ibu ion cen e s in Mexico Ci y iden i ied h ough ou scou ing
o a ailable in o ma ion.
14Acco ding o mos sou ces, his collabo a i e map was he mos used and accu a e sou ce o in o ma ion
a e he ea hquake.
17
Figu e 3: Loca ion o Dis ibu ion Cen e s in Mexico Ci y
Sou ce: Au ho s’ compila ion.
No es: The g ay lines indica e he bounda ies o he colonias (“neighbo hoods” in Mexico Ci y). The diamonds ep esen he dis ibu ion cen e s
ound.
We de e mined whe he a dis ibu ion cen e was in o nea a esponden ’s neighbo hood
i he di e ence in zip codes was smalle han 30 and in he same municipali y ( o example,
77810 was he zip code o he place whe e he dis ibu ion was loca ed and 77830 was he
code o he esponden ’s esidence).15 Responden s we e assigned a alue o 1 o ha ing
a dis ibu ion cen e in hei neighbo hood and 0 o he wise. We a i ed a 30 o he
cu -o ia an induc i e app oach, whe e we es ima ed he dis ance be ween dis ibu ion
cen e s and su ey esponden s’ houses in kilome e s and walking ime in minu es using hei
geog aphical coo dina es.16 The summa y s a is ics in Table 3demons a e ha esponden s
15We use he neighbo hood and no he municipali y because we wan o ma ch he ques ion asked in he
su ey. See Online Appendix C.
16We calcula ed he dis ance be ween each household and i s closes dis ibu ion cen e s. A e ob-
aining he in o ma ion om Google Maps, we checked he co espondence o he add esses and he ZIP
codes using he o icial ables o ZIP codes and human se lemen s p o ided by he Mexican go e n-
men , which can be ound a h ps://da os.gob.mx/busca/da ase /codigos-pos ales-coo denadas-y-colonias/
esou ce/7675 085-6a8 -4b20-8091- 5117 e964e.
18
coded o be esiding nea a dis ibu ion cen e li e nea he cen e s by he dis ance measu es.
Table 3: Summa y Time and Dis ance be ween Dis ibu ion Cen e and Households
Mean S.D. Min p25 Median p75 Max
Time (minu es) 21.21 15.10 1.10 10.70 15.13 32.30 50.30
Dis ance (Km) 1.66 1.21 0.10 0.83 1.12 2.56 4.01
Table 4 epo s he esul s o objec i e p oximi y o dis ibu ion cen e s, as opposed o
pe cei ed closeness o he cen e s, on gene al poli ical us . The eg essions emain he
same as be o e, excep ha he a iable o in e es now cap u es p oximi y o a e i ied
loca ion o dis ibu ion cen e s. The able shows ha ha ing a dis ibu ion cen e nea by
posi i ely co ela es wi h esponden s’ gene al poli ical us . No ably, addi ional analyses
demons a e ha whe he he go e nmen p o ides he dis ibu ion cen e o no does no
a ec he conclusion.
The coe icien is la ge bu less p ecisely es ima ed han in Table 2. Since he sample
size has educed subs an ially, i lowe s he s a is ical powe o de ec a signi ican and
consis en e ec . In he mos pa simonious model shown in column 1, he coe icien o he
dis ibu ion cen e is signi ican a he 10% le el. Howe e , he signi icance a ies depending
on he con ols we include.17 Conside ing hese esul s and he p io ones ha use he
sel - epo ed exis ence o dis ibu ion cen e s, we conclude ha he e is sugges i e e idence
suppo ing H2a.
17We also an he eg essions using di e en coding schemes o he Objec i e DC a iable by elaxing
he zip code di e ences o be 50 and 100. The s a is ical signi icance anishes wi h he al e na i e a iables,
al hough he posi i e co ela ions emain. Howe e , he wo al e na i e coding schemes inc eased he dis ance
be ween su ey esponden s’ esidences and dis ibu ion cen e s o abou 30 and 40 minu es, meaning ha
he dis ibu ion cen e s we e po en ially no p ecisely in he neighbo hood.
19
Table 4: Associa ion be ween Objec i e Dis ibu ion Cen e s and Gene al Poli ical T us
Gene al poli ical us
(1) (2) (3) (4) (5) (6)
Objec i e DC 0.045* 0.042* 0.045* 0.039 0.045* 0.026
(0.025) (0.024) (0.024) (0.024) (0.025) (0.025)
Male -0.002 -0.002 -0.002 -0.002 -0.002 -0.002
(0.016) (0.016) (0.016) (0.016) (0.016) (0.016)
Age -0.002*** -0.002*** -0.002*** -0.002*** -0.002*** -0.002***
(0.001) (0.001) (0.001) (0.001) (0.001) (0.001)
Las yea o educa ion app o ed -0.001 -0.001 -0.001 -0.001 -0.001 -0.001
(0.002) (0.002) (0.002) (0.002) (0.002) (0.002)
Numbe o adul s in he household -0.001 -0.001 -0.001 -0.002 -0.001 -0.002
(0.005) (0.005) (0.005) (0.005) (0.005) (0.005)
Numbe o child en in he household -0.003 -0.004 -0.003 -0.003 -0.003 -0.003
(0.006) (0.006) (0.006) (0.006) (0.006) (0.006)
PRI:p e -0.163 -0.408
(0.196) (0.282)
PRD:p e 0.040 0.094
(0.124) (0.179)
Weal h: p e 0.039 0.505
(0.263) (0.501)
Risk A e sion: p e 0.008 0.007
(0.009) (0.009)
Pa ience: p e 0.001 0.005
(0.005) (0.006)
Social T us : p e 0.005 0.065
(0.062) (0.106)
Cons an 0.275*** 0.288*** 0.254 0.059 0.260 -0.398
(0.043) (0.055) (0.156) (0.287) (0.170) (0.582)
Obse a ions 515 515 515 515 515 515
R-squa ed 0.038 0.039 0.038 0.039 0.038 0.043
No es: Clus e ed s anda d e o s a e in pa en heses. ∗p < 0.10,∗∗p < 0.05,∗∗∗p < 0.01. The a e age le els
o poli ical a ilia ion, isk a e sion, pa ience, and social us a e calcula ed a he municipali y le el be o e he
ea hquake.
6.3 Co espondence be ween Subjec i e and Objec i e Measu es
The somewha a ying esul s om sel - epo ed s. objec i e loca ions o dis ibu ion cen-
e s aise he ques ion: How much o e lap exis s be ween he wo a iables in Mexico Ci y?
We ound ha he a iables based on sel - epo s and ac ual loca ions ma ch o 62.64% o
esponden s.
How we e he dis ibu ion cen e s a ge ed? Analyses o p edic o s o he sel - epo and
20
objec i e loca ion measu es a e in o ma i e, hough hey also di e ge in some ways. Wi h
he sel - epo a iable, we ind ha hose who expe ienced wa e se ice in e up ions, had
di icul ies inding ood and eme gency p oduc s o hei household, o suppo ed PRD (i.e.,
he pa y o he head o he Fede al Dis ic /Mexico Ci y a he ime o he ea hquake) a e
mo e likely o epo ha he e was a dis ibu ion cen e in hei neighbo hood compa ed
o hose who did no expe ience hose di icul ies (see Table E.1). Analyses o he ac ual
loca ion-based a iable yield no s a is ically signi ican co ela ions be ween dis ibu ion
cen e s in esponden s’ neighbo hood and expe iencing wa e se ice dis up ions o ha ing
di icul ies loca ing ood and o he necessi ies. Ins ead, only he poli ical a iables show
s a is ically signi ican co ela ions wi h he ac ual dis ibu ion cen e loca ion a iable (See
Table E.2). The p obabili y o ha ing a dis ibu ion cen e nea by is posi i ely co ela ed
wi h suppo o he incumben pa y. Al oge he , he e idence indica es some misalignmen
be ween ac ual and pe cei ed eali ies, al hough we cau ion ha he po en ial incomple eness
o he objec i e cen e lis may be pa ially esponsible o hese di e ences.
6.4 Robus ness Checks and Placebo Analysis
While he wo c oss-sec ional da a se s a e use ul o assessing he ea hquake’s impac on
us , we wan o ensu e he e was no selec ion on he sample su eyed in he second wa e.18
We use weigh ing and di e en ma ching echniques o accoun o he ac ha people’s
baseline condi ions, like educa ion, numbe o child en, numbe o adul s in he household,
and egion, migh a ec he p obabili y o acing a ha de si ua ion a e he ea hquake,
ha ing a dis ibu ion cen e in he neighbo hood, and expe iencing lowe le els o us .19
To gua an ee ha he ma ching es ima o s consis en ly es ima e he e ec s o in e es ,
we assume ha su eying a pe son in ound one o wo was independen o he ou comes,
condi ional on he co a ia es and ha he p obabili y o being su eyed in he second wa e
is bounded away om ze o and one. Using weigh ing and ma ching echniques, we con i m
ha he esul s a e obus and ha he wo samples a e compa able. The esul s in F.1
in he Online Appendix show ha gene al poli ical us dec eased signi ican ly a e he
ea hquake ac oss models con olling o di e en con ol a iables and ma ching me hods.
Addi ionally, as we ha e shown in Sec ion 6, he ac ual dis ibu ion cen e a iable may
include some noise. To add ess his conce n and ensu e ha he epo ed s a is ical sig-
188%(93 indi iduals) o he pos -quake sample epo ha hey we e su eyed in bo h wa es.
19Reweigh ing is compe i i e wi h he mos e ec i e ma ching es ima o s when he o e lap is good (Busso
e al., 2014), as in ou case.
21
Nikolo a, Elena, and Nikolay Ma ino . 2017. “Do Public Fund Wind alls Inc ease Co -
up ion? E idence om a Na u al Disas e .” Compa a i e Poli ical S udies 50, no. 11
(Sep embe ): 1455–88.
O’Donnell, Guille mo. 2004. “The Quali y o Democ acy: Why he Rule o Law Ma e s.”
Jou nal o Democ acy 15 (4): 32–46. Accessed Sep embe 29, 2021.
Olson, Richa d S ua . 2000. “Towa d a Poli ics o Disas e : Losses, Values, Agendas, and
Blame.” C isis Managemen 18 (2): 154.
Olson, Richa d S ua , and Vincen T. Gaw onski. 2010. “F om Disas e E en o Poli ical
C isis: A “5C +A” F amewo k o Analysis.” In e na ional S udies Pe spec i es 11 (3):
205–21.
Pe o a, K is ina, and Elisabe h L. Ros old. 2024. “Mi iga ing he legacy o iolence: Can
lood elie imp o e people’s us in go e nmen in con lic -a ec ed a eas? E idence
om Pakis an.” Wo ld De elopmen 173:106372. h ps://www.sciencedi ec .com/scien
ce/a icle/pii/S0305750X23001900.
Reinha d , Gina Yanni ell. 2015. “Fi s -Hand Expe ience and Second-Hand In o ma ion:
Changing T us ac oss Th ee Le els o Go e nmen .” Re iew o Policy Resea ch 32 (3):
345–64.
. 2019. “The In e sec ionali y o Disas e s’ E ec s on T us in Public O icials.” Social
Science Qua e ly 100 (7): 2567–80. Accessed Augus 3, 2021.
Rudolph, Thomas J. 2003. “Ins i u ional Con ex and he Assignmen o Poli ical Respon-
sibili y.” The Jou nal o Poli ics 65, no. 1 (Feb ua y): 190–215.
Sca ascini, Ca los, and Joanna Valle L. 2020. “The Elusi e Ques o G ow h in La in
Ame ican and he Ca ibbean: The Role o T us .” Washing on, DC.
Schneide , Saund a K. 1992. “Go e nmen al Response o Disas e s: The Con lic be ween
Bu eauc a ic P ocedu es and Eme gen No ms.” Public Adminis a ion Re iew 52, no.
2 (Ma ch): 135.
Sobel, Russell S., and Pe e T. Leeson. 2006. “Go e nmen ’s Response o Hu icane Ka ina:
A Public Choice Analysis.” Public Choice 127 (1-2): 55–73.
S ephane, Vic o . 2021. “Hiding behind he eil o ashes: Social capi al in he wake o na u al
disas e s.” Wo ld De elopmen 145:105518. h ps://www.sciencedi ec .com/science/
a icle/pii/S0305750X21001303.
28
Townshend, I an, Olu Awosoga, Judi h Kulig, e al. 2015. “Social Cohesion and Resilience
Ac oss Communi ies Tha Ha e Expe ienced a Disas e .” Na u al Haza ds 76:913–38.
h ps://doi.o g/10.1007/s11069-014-1526-4.
Uni ed S a es Geological Su ey. 2017. Magni ude 7.1 Ea hquake in Mexico. USGS News,
Sep embe . h ps://www.usgs.go /news/ ea u ed-s o y/magni ude-71-ea hquake-
mexico.
U bina i, Nadia, and Ma k E Wa en. 2008. “The Concep o Rep esen a ion in Con empo-
a y Democ a ic Theo y.” Annual Re iew o Poli ical Science 11:387–412.
Villegas, Paulina, and Azam Ahmed. 2017. Powe ul Ea hquake S ikes Mexico, Killing
Dozens. The New Yo k Times, Sep embe . h ps://www.ny imes.com/2017/09/19/
wo ld/ame icas/mexico-ea hquake.h ml.
Wong, Ma hew Y. H., and Ying-ho Kwong. 2021. “Wa ning signal: Poli ical T us , Typhoons
and he My h o he ‘Li’s Field’ in Hong Kong.” Asia Paci ic Viewpoin 62 (2): 206–22.
Wo man, Camille B. 1976. “Causal A ibu ions and Pe sonal con ol.” In New Di ec ions
in A ibu ion Resea ch, edi ed by John H. Ha ey, William J. Ickes, and Robe F.
Kidd, 23–52. Hillsdale, NJ: E lbaum.
Yamamu a, Eiji. 2014. “Impac o Na u al Disas e on Public Sec o Co up ion.” Public
Choice 161 (3-4): 385–405.
You, Yu, Yi an Huang, and Yuyi Zhuang. 2020. “Na u al Disas e and Poli ical T us : A
Na u al Expe imen S udy o he Impac o he Wenchuan Ea hquake.” Chinese Jou nal
o Sociology 6, no. 1 (Janua y): 140–65.
Zechmeis e , Elizabe h J., and Daniel Zizumbo-Colunga. 2013. “The Va ying Poli ical Toll o
Conce ns abou Co up ion in Good e sus Bad Economic Times.” Compa a i e Poli ical
S udies 46 (10): 1190–218.
29
Online Appendix
A Li e a u e e iew
Table A.1: Selec ed Recen Empi ical In es iga ions o
Poli ical T us A e A Na u al Disas e
Pape Case Da a Explana o y T us ee (DV)
a iable
Alb ech 10 disas e s Su ey Disas e Poli icians
2017 in Eu ope (c oss-na ional,
mul iple ounds)
Ald ich Ea hquake Obse a ions, Disas e Cen al &
2017 & Tsunami in e iews, local go
in Japan & news a icles
Ca lin e al. Ea hquakes in Su ey Di ec s. indi ec People in he
2014b El Sal ado , (c oss-na ional, disas e expe ience communi y
Hai i & Chile pos -disas e )
Han e al. & Ea hquake Obse a ions, Disas e Local go
2011 in China in e iews, &
ocus g oup
Ma hew & Typhoon News a icles T us in poli icians Bu eauc a
Kwong 2021 in Hong Kong a e disas e
Nicholls & Hu icane Su ey Di ec s. indi ec Local, s a e,
Picou 2013 in he U.S. disas e expe ience ede al go
Reinha d Hu icane Su ey Di ec s. indi ec p esiden , go e no ,
2015 in he U.S. disas e expe ience Mayo , FEMA,
s a e/local EMA,
Reinha d Hu icane Su ey Gende Police &
2019 in he U.S. ambulance se ice
Sco e al. Coal was e Su ey Time Expe , gene al
2016 up u e ppl, public
in he U.S. o icials, local go
Sco e al. Coal was e In e iews Disas e Co po a ions,
2005 up u e go , & egula o y
in he U.S. au ho i ies
S ömbäck Tsunami Su ey, Disas e Poli icians,
& No d 2006 in Indonesia ocus g oup go , media
You e al. Ea hquake Su ey Disas e Local, p o incial,
2020 in China (2 ounds) cen al go o icial
No es: DV lis ed include only ou comes ela ed o us . S udies abou incumben app o al a ings
and elec o al ou comes a e ela ed bu no men ioned he e.
B Su ey Design and Implemen a ion
The p e-ea hquake su ey (Augus 31 - Sep embe 19, 2017) and he pos -ea hquake su -
ey (No embe 18, 2017 - Janua y 18, 2018) had IRB app o al and con o med o APSA’s
P inciples and Guidance o Human Subjec s Resea ch. Rec ui ed indi iduals we e ead a
s udy in o ma ion shee ha in o med hem ha hey we e being in i ed o pa icipa e in a
1
s udy abou public opinion in Mexico, conduc ed by LAPOP and adminis e ed by DATA-
OPM. They we e hen allowed o consen o decline o pa icipa e. The e was no decep ion
in he s udy. Those who consen ed o he s udy could decline any ques ion and e mina e
he s udy a any ime. As is common o s anda d public opinions in Mexico, and elsewhe e,
pa icipan s we e no compensa ed. The s udy did no in e ene wi h eal-wo ld e en s and
only unc ioned as an opinion su ey.1
We designed he o iginal su ey o be ep esen a i e o non-ins i u ionalized adul s (18
o abo e) li ing in he g ea e Mexico Ci y me opoli an a ea. The o iginal a ge was 900
in e iews in he p e-ea hquake su ey. Each egion o he me opoli an a ea was designa ed
a p opo ion o he o al numbe o in e iews based on popula ion da a. A sample o people
was hen chosen in ou s ages, as shown in Table B.1.
We used geog aphical s a i ica ion since his inc eases p ecision by yielding smalle an-
dom sample e o s han hose ob ained wi h a simple andom sampling o he same sample
size (Ge be and G een, 2012). A s a i ied sample also ends o be mo e ep esen a i e and
dispe sed since i gua an ees he inclusion o municipali ies in he en i e me opoli an a ea.
Table B.1: S a i ied Mul i-s age Clus e ed Sampling Design
Sampling Uni Uni selec ed Sampling Selec ion P ocess
S a i ica ion
S a a Regions S a i ied Sampling
Mul is age sample om each s a um
Fi s s age Seconda y Sampling Uni (SSU)* Elec o al sec ions‡PPS sampling§
Second s age Te ia y Sampling Uni (TSU) Blocks o dwellings PPS sampling
Thi d s age Qua e na y Sampling Uni (QSU) Households Sys ema ic sampling
Fou h s age Final Sampling Uni s (FSU) Pe son in household Quo a by sex and age
No es: *‡An elec o al sec ion is he basic geog aphical uni in o which he na ional e i o y is di ided o elec o al pu poses. §PPS deno es P obabili y
P opo ional o Size.
We used a p obabili y p opo ional o he adul popula ion sampling (PPS) me hod o
selec he elec o al sec ions (Seconda y Sampling Uni s, SSUs) wi hin each egion.2Since six
in e iews we e o be conduc ed pe elec o al sec ion, he numbe o in e iews pe egion
di ec ly de e mined he numbe o elec o al sec ions o be andomly d awn om each egion.
1In addi ion o su ey esponses, he de ices we e p og ammed o cap u e loca ion, audio segmen s,
a pic u e o he enume a o om he on - acing came a, and iming da a; he local eam audi ed 100%
o su eys o quali y con ol on each o hese dimensions, and ou eam conduc ed a second audi o jus
o e 20%. A small p opo ion o ini ially- eco ded in e iews we e canceled o quali y con ol issues (abou
6-7% o comple ed in e iews); he compu e -assis ed quali y con ol p ocedu es mean ha hose poo
quali y in e iews we e de ec ed and eplaced wi h alid high-quali y in e iews om he co ec blocks
while ieldwo k was s ill in p og ess.
2The e a e ede al elec o al dis ic s and elec o al sec ions. We use elec o al sec ions, he basic geo-
g aphical uni in o which he na ional e i o y is di ided o elec o al pu poses. (INE, FEPADE, UNAM,
T ibunal Elec o al del Pode Judicial de la Fede ación,2016).
2
Table B.2 shows each egion’s pa icipa ion in he su ey, he numbe o municipali ies and
elec o al sec ions, he numbe o in e iews ha had o be ca ied ou in each egion, and
he numbe o in e iews ha we e ac ually ca ied ou be o e and a e he ea hquake.
Table B.2: Numbe o In e iews pe Region ( a ge and ac ual sample)
Sample Ac ual sample
(Ta ge ) P e - Ea hquake Pos - Ea hquake
Region Numbe o Numbe o Numbe o P opo ion Numbe o Numbe o Numbe o P opo ion Numbe o Numbe o P opo ion
Municipali ies SSUs In e iews Municipali ies SSUs In e iews SSUs In e iews
Cen e 6 28 168 18.67%4 19 109 19.16%32 229 19.67%
No h 14 27 162 18.00%9 19 111 19.51%24 229 19.67%
Sou h 10 25 150 16.67%6 13 78 13.71%26 156 13.40%
Eas 9 47 282 31.33%6 36 213 37.43%62 431 37.03%
Wes 6 23 138 15.33%4 10 58 10.19%38 119 10.22%
To al 45 150 900 100%29 97 569 100%162 1,164 100%
Once we had andomly selec ed he elec o al sec ions, we used a new PPS sampling
me hod o selec he blocks o dwellings in each SSU. The numbe o households in i de-
ined he size o each block. (see Na ional Housing In en o y 2016). Nex , enume a o s
sys ema ically chose households, skipping wo housing uni s a e each comple ed in e iew
wi hin a block.3Finally, a pe son in each household was chosen acco ding o gende and
age quo as, which we e es ima ed based on he dis ibu ion o he popula ion egis e ed in
he elec o al sec ions.4Th ee di e en quo a o ms (A, B, and C), shown in Table B.3, we e
used o app oxima e he e e ence pa ame e s o each SSU.
Table B.3: Household Fo ms
Fo m A Fo m B Fo m C
Gende /Age 18- 29 30- 50 <50 To al 18- 29 30- 50 <50 To al 18- 29 30- 50 <50 To al
Men 1 2 0 3 1 1 1 3 1 1 1 3
Women 1 1 1 3 1 2 0 3 1 1 1 3
To al 2 3 1 6 2 3 1 6 2 2 2 6
The p e-ea hquake su ey was abou wo- hi ds comple e (569 in e iews) when he
ea hquake s uck. The pos -quake su ey was conduc ed wo mon hs a e he ea hquake.
The pos -ea hquake su ey was designed o d aw 900 indi iduals om he same SSUs as
in he o iginal design. In addi ion, we laye ed in an o e sample o indi iduals who i he
p o ile o he 569 indi iduals who we e in e iewed be o e he ea hquake. We only keep he
da a om esponden s in he same SSUs as esponden s in he i s wa e o he su ey.
3The enume a o s we e ins uc ed o loca e he block’s no h-eas e n poin o s a he in e iews and
con inue walking clockwise. In case o ejec ion, acan housing, o people’s absence, he enume a o selec ed
he adjacen housing. When su ey pe sonnel eached he end o a block wi hou comple ing he quo a o 3
in e iews pe block, hey con inued o he nex block, ollowing he same ou ine as in he p e ious block.
4In e iewees had o eside pe manen ly in he household, no be a domes ic wo ke o isi o . I mo e
han wo people in a house we e in he same age g oup and gende , we ec ui ed a pe son wi h he closes
bi hday.
3
C Su ey Ques ions, Va iables, and Balance Check
Table C.1: Ques ions
Ques ions o gene al poli ical us a iable
Va iable name Ques ion Response op ions No e
Gene al poli -
ical us
1./2. [Poli icians in gene al/ public o icials],
do you hink i is e y common, somewha
common, no e y common, o no common
a all ha when hey p omise some hing hey
ul ill i ?
3./4. [Poli icians in gene al/ public o icials],
do you hink i is e y common, somewha
common, no e y common, o no common
a all ha hey comply wi h he laws and
egula ions o he coun y?
1. No common a all
2. No e y common
3. Somewha common
4. Ve y common
One composi e a iable is cons uc ed
om he ou su ey ques ions using
PCA
Ques ion o sel - epo ed dis ibu ion cen e a iable
Code Ques ion Response op ions No e
Sel - epo ed
DC
A e he ea hquake, we e he e dis ibu ion
cen e s o ood, wa e and o he essen ial
i ems in you neighbo hood?
1. Yes
2. No
Recoded as:
1. Yes 0. No
4
Table C.2: Ques ions
Ques ions o con ol a iables
Code Ques ion Response op ions No e
Male Gende 1. Male
2. Female
Recoded as:
1. Male
0. Female
Age How old a e you? Nume ic alue
Educa ion How many yea s o schooling ha e you com-
ple ed?
Ranges om 0 (none) o 24 (doc o a e
deg ee)
N. Adul s Including you, how many adul s li e in you
home?
Nume ic alue
N. Child en How many child en unde he age o 18 li e
in you home?
Nume ic alue
PRI/PRD O hese pa ies, which a e you mos willing
o suppo ?
1. PRI
2. PAN
3. PRD
4. MORENA
5. G een Pa y
6. New Alliance
7. Ci izen Mo emen
8. Labo Pa y
Fi s , coded one i a esponden sup-
po ed PRI/PRD, 0 o he wise. Then,
es ima ed he p opo ion o pa y sup-
po e s be o e he ea hquake and as-
signed he a e age o esiden s o a co -
esponding municipali y
Weal h Could you ell me i you ha e he ollow-
ing in you house: (1) Ba h oom inside
he house (2) Sala ied employee(s)/Domes ic
wo ke /Domes ic se ice (3) Au omobile(
s)/ca (s) (4) Mic ocompu e (s), lap ops,
able s, ipads and ne books (5) Dishwashe
(6) Re ige a o (7) F eeze(s) (8) Washing
machine (9) Mic owa e (10) Mo o cycle (11)
Clo hes d ye (12) Tele ision (13) Cable ele-
ision, sa elli e ele ision, Ne lix (14) Land-
line elephone (15) Cellula elephone (16)
D inking wa e inside he house (17) In e -
ne se ice inside he house
1. Yes 2. No Recoded each su ey ques ion as an in-
dica o a iable assigning a alue o 1 i
owned, hen summed up he numbe o
i ems ha a esponden possessed, and
es ima ed he municipali y a e age o i
be o e he ea hquake and assigned he
a e age o esiden s o a co esponding
municipali y
Pa ience Imagine you ha e won 1,800 Mexican pesos
in a lo e y and ha e he op ion o ecei e he
ewa d in a paymen oday o 1,800 pesos o
a highe paymen ha you would ecei e in
12 mon hs. I am going o ask you a se ies o
ques ions abou you p e e ences abou e-
cei ing 1800 oday o ecei ing a highe pay-
men in a yea .
The sequence o ques ions has 32
possible o de ed ou comes, such ha
we can de i e a measu e o pa ience
anging om 1 o 32. Example: Be-
ween ecei ing a p ize o 1,800 pesos
oday o 2,770 pesos in 12 mon hs,
wha would you p e e ?
1. To ecei e 1,800 pesos oday.
2. To ecei e 2,770 pesos in a yea .
We use he a iable con aining alues
om 1 o 32, whe e 1 indica es he low-
es le el o pa ience and 32 he highes
le el. We calcula ed he a e age be o e
he ea hquake and assigned i o esi-
den s o he co esponding municipali y
a e he ea hquake.
Risk a e sion Imagine you can choose be ween ecei ing
a su e p ize in pesos, o ecei ing a lo e y
icke wi h which you could win 5,400 pesos,
bu you could also win no hing. Below I am
going o name di e en p ize al e na i es so
you can ell me which ones you would choose.
The sequence o ques ions has 32
possible o de ed ou comes, such ha
we can de i e a measu e o isk a e -
sion anging om 1 o 32. Example:
Be ween a lo e y icke , in which you
ha e an equal chance o winning 5,400
pesos o no winning any hing, and
ecei ing a su e p ize o 360 pesos,
which op ion would you choose?
1. The lo e y icke .
2. The su e p ize o 360 pesos.
We use he a iable con aining alues
om 1 o 32, whe e 1 indica es he
lowes le el o isk a e sion (also called
isk-lo ing pe son) and 32 he highes
le el o isk a e sion ( isk-a e se pe -
son). We calcula ed he a e age be o e
he ea hquake and assigned i o esi-
den s o he co esponding municipali y
a e he ea hquake.
Social us In gene al, would you say ha he majo -
i y o people a e e y us wo hy, some-
wha us wo hy, no e y us wo hy, o
un us wo hy?
1. Ve y us wo hy
2. Somewha us wo hy
3. No e y us wo hy
4. Un us wo hy
Es ima ed he municipali y a e age be-
o e he ea hquake and assigned he
a e age o esiden s o a co esponding
municipali y
5
Table C.3: Desc ip i e S a is ics
P e-ea hquake Pos -ea hquake
Mean s.d. Mean s.d. Di e ence p- alue Np e Npos
Male 0.511 (0.500) 0.508 (0.500) 0.004 0.885 569 1,164
Age 39.306 (16.023) 38.942 (15.563) 0.364 0.651 568 1,164
Educa ion 10.743 (4.292) 10.683 (3.979) 0.060 0.774 569 1,149
N. Adul s 3.497 (1.830) 3.331 (2.004) 0.166 0.097 563 1,148
N. Child en 1.528 (1.612) 1.431 (1.565) 0.098 0.229 562 1,145
PRI 0.148 (0.355) 0.141 (0.348) 0.007 0.707 569 1,164
PRD 0.130 (0.337) 0.083 (0.277) 0.047 0.002 569 1,164
Weal h 9.396 (3.026) 9.259 (3.187) 0.137 0.397 560 1,116
Social us 2.692 (0.782) 2.704 (0.805) -0.012 0.768 569 1,164
Cen e 0.192 (0.394) 0.197 (0.398) -0.005 0.799 569 1,164
No h 0.195 (0.397) 0.197 (0.398) -0.002 0.935 569 1,164
Sou h 0.137 (0.344) 0.134 (0.341) 0.003 0.861 569 1,164
Eas 0.374 (0.484) 0.370 (0.483) 0.004 0.869 569 1,164
Wes 0.102 (0.303) 0.102 (0.303) -0.000 0.985 569 1,164
Pa ience 11.380 (11.715) 11.989 (11.868) -0.609 0.343 560 850
Risk A e sion 28.073 (7.212) 26.011 (8.853) 2.062 0.000 560 816
Gene al T us 0.260 0.200 0.193 0.232 0.027 0.007 560 1,134
No es: S anda d de ia ion in pa en heses. The e is 1 missing alue in age, 15 in educa ion, 22 in he
numbe o adul s, 26 in he numbe o child en, 357 in isk a e sion, 323 in ime-p e e ences (pa ience),
and 39 in gene al us
.
D Gene al Poli ical T us
D.1 Dis ibu ion o Dependen Va iables
Table D.1: Dis ibu ion o Dependen Va iables
Panel 1 - P omises
Poli icians Ci il se an s
P e Pos To al P e Pos To al
N. Col % N. Col % N. Col % N. Col % N. Col % N. Col %
(1) No common a all 240 42.25% 575 49.83% 815 47.33% 254 45.12% 600 52.04% 854 49.77%
(2) No e y common 228 40.14% 453 39.25% 681 39.55% 209 37.12% 391 33.91% 600 34.97%
(3) Somewha common 77 13.56% 94 8.15% 171 9.93% 89 15.81% 128 11.10% 217 12.65%
(4) Ve y common 23 4.05% 32 2.77% 55 3.19% 11 1.95% 34 2.95% 45 2.62%
To al 568 100% 1,154 100% 1722 100% 563 100% 1,153 100% 1,716 100%
Pea son chi2(3) = 17.852 p- alue < 0.001 Pea son chi2(3) = 12.813 p- alue = 0.005
6
Panel 2 - Compliance
Poli icians Ci il se an s
P e Pos To al P e Pos To al
N. Col % N. Col % N. Col % N. Col % N. Col % N. Col %
(1) No common a all 225 39.68% 477 41.41% 702 40.84% 227 39.96% 513 44.49% 740 43.00%
(2) No e y common 255 44.97% 491 42.62% 746 43.40% 223 39.26% 456 39.55% 679 39.45%
(3) Somewha common 78 13.76% 157 13.63% 235 13.67% 108 19.01% 162 14.05% 270 15.69%
(4) Ve y common 9 1.59% 27 2.34% 36 2.09% 10 1.76% 22 1.91% 32 1.86%
To al 567 100% 1,152 100% 1,719 100% 568 100% 1,153 100% 1,721 100%
Pea son chi2(3) = 1.804 p- alue = 0.614 Pea son chi2(3) = 7.843 p- alue = 0.049
D.2 P incipal Componen Analysis
The p incipal componen analysis on ou su ey i ems shows ha he a ia ions in he
da a can be explained by ou componen s, indica ed as in Componen s 1-4 in Table D.2.
Howe e , he i s la en a iable alone explains abou 60% o he a ia ion and only i s
eigen alue exceeds 1, meaning ha i is he only unde lying la en a iable wo h explo ing
as o he s do no ha e signi ican co a iance.
Table D.2: Componen s om PCA
Componen Eigen alue Di e ence P opo ion Cumula i e
Componen 1 2.295 1.480 0.573 0.573
Componen 2 0.814 0.327 0.204 0.777
Componen 3 0.487 0.084 0.122 0.899
Componen 4 0.403 0.000 0.1007 1.000
No es: Numbe o obse a ions = 1,694.
Table D.3 shows ha he loadings o all ou su ey i ems a e high on he i s componen .
The e o e, he e a e su icien a ia ions ha co ela e o one ano he and can be in e p e ed
as apping in o one la en a iable.
Table D.3: Loadings
Va iable Componen 1
P omises - Poli icians 0.483
P omises - Ci il se an s 0.500
Compliance - Poli icians 0.506
Compliance - Ci il se an s 0.510
7
E Dis ibu ion Cen e s
E.1 Obse ed Dis ibu ion Cen e s Da a Collec ion
We analyzed pos -ea hquake social media pos s in h ee phases. Fi s , we sea ched Google
Schola o li e a u e on he Mexico Ci y go e nmen ea hquake esponse and he ole o
social media. Second, we c ea ed a lis o hash ags o Twi e sea ches. Thi d, we collec ed
pho os o dis ibu ion cen e s ha p o ided aid in he a e ma h o he ea hquake.
As a i s s ep, we ound a icles ha examined he go e nmen esponse ia—and in
compa ison o—ci il socie y esponses o he ea hquake. These a icles ecoun ha mem-
be s o ci il socie y o ganiza ions in he hou s a e he Sep embe 19 ea hquake assembled
a la ge social ne wo k o exchange eal- ime in o ma ion abou building collapses, missing
pe sons, and la ge-scale damage on Wha sApp (Campos Ri e a, 2018; Mo a e al., 2018).
The in o ma ion being ansmi ed ia Wha sApp abou building collapses and damage
became a collabo a i e map called “Mapeo colabo a i o Resca eMX,” which ci il socie y
and go e nmen al ac o s used o ind su i o s o he ea hquake, assess whe e damage had
occu ed, and loca e shel e s and dis ibu ion cen e s.
A e we conduc ed he Google Schola sea ches, we c ea ed a lis o po en ially popu-
la pos -ea hquake Twi e hash ags. Ou goal in hese sea ches was o ind in o ma ion
abou who an dis ibu ion cen e s, whe e hey we e loca ed, and how his in o ma ion
was dissemina ed. We ound ha indi iduals p ima ily communica ed publicly ia Twi -
e ( ollowing Mo a e al. (2018)). We hen used he i s se o wee s o c ea e a longe
lis o he mos popula pos -ea hquake hash ags: #19S, #Ve i icado19S, #Cen odeA-
copio, #sismoCDMX, #19sAcopio, #19s olun a ios, #AyudaSismo, #AyudaCDMX, # e-
cons ucciónCDMX, #6meses19S, #Fue zaMéxico, #Ampli icaMexico, and #Edoméx19S.
In addi ion, we sea ched o hash ags ha men ioned he Mexican A my (SEDENA), Na y
(SEMAR), and Plan DN-III-E—a pos -disas e assis ance plan, in which he mili a y is
deployed o assis in disas e eco e y.
Thi d, we conduc ed Google sea ches o dis ibu ion cen e images. We downloaded 40
pho os in which he image me ada a associa es he cen e wi h he Sep embe 19 ea hquake.
Mos indi iduals in he pho os a e no wea ing uni o ms; howe e , in i e pho os, indi iduals
a e wea ing a b igh ly colo ed es (neon g een o pu ple) o a s icke ha says “CDMX.”
In wo o he pho os, olun ee s a e wea ing es s ha clea ly say “C uz Roja” (Red C oss).
Dis ibu ion cen e s we e s a ed by indi iduals om di e se o ganiza ions, including ci il
8
Table G.2: Ma ginal E ec s o Obse ed Dis ibu ion Cen e on Gene al T us
Gene al poli ical us
(1) (2) (3) (4) (5) (6)
Objec i e DC -0.024 -0.017 -0.027 -0.024 -0.023 -0.030
(0.022) (0.022) (0.021) (0.020) (0.022) (0.024)
Male -0.018 -0.018 -0.017 -0.018 -0.018 -0.017
(0.018) (0.018) (0.018) (0.018) (0.018) (0.019)
Age -0.001 -0.001 -0.001 -0.001 -0.001 -0.001
(0.001) (0.001) (0.001) (0.001) (0.001) (0.001)
Las yea o educa ion app o ed -0.006* -0.006** -0.006** -0.006** -0.006** -0.006**
(0.003) (0.003) (0.003) (0.003) (0.003) (0.003)
Numbe o adul s in he household -0.013* -0.013* -0.013* -0.014* -0.013* -0.013*
(0.007) (0.007) (0.007) (0.007) (0.007) (0.007)
Numbe o child en in he household -0.004 -0.004 -0.004 -0.004 -0.004 -0.003
(0.009) (0.009) (0.009) (0.009) (0.009) (0.009)
PRI:p e 0.256 -0.036
(0.233) (0.288)
PRD:p e -0.110 -0.098
(0.151) (0.127)
Weal h: p e 0.387 0.568
(0.244) (0.479)
Risk A e sion: p e -0.012** -0.015**
(0.006) (0.006)
Pa ience: p e 0.001 0.002
(0.005) (0.006)
Social T us : p e -0.052 0.108
(0.095) (0.132)
Cons an 0.407*** 0.396*** 0.195 0.739*** 0.553** 0.220
(0.051) (0.059) (0.151) (0.196) (0.259) (0.540)
Obse a ions 271 271 271 271 271
Clus e ed s anda d e o s a e in pa en heses. ∗p < 0.10,∗∗ p < 0.05,∗∗∗ p < 0.01
15
Re e ences
Abadie, Albe o, Da id D ukke , Jane Lebe He , and Guido W. Imbens. 2004. “Imple-
men ing Ma ching Es ima o s o A e age T ea men E ec s in S a a.” S a a Jou nal 4
(3): 290–311.
Abadie, Albe o, and Guido W. Imbens. 2006. “La ge Sample P ope ies o Ma ching Es i-
ma o s o A e age T ea men E ec s.” Econome ica 74 (1): 235–67.
. 2011. “Bias-co ec ed Ma ching Es ima o s o A e age T ea men E ec s.” Jou nal
o Business & Economic S a is ics 29:1–11.
. 2016. “Ma ching on he Es ima ed P opensi y Sco e.” Econome ica 84 (2): 781–807.
Busso, Ma ias, John DiNa do, and Jus in McC a y. 2014. “New E idence on he Fini e
Sample P ope ies o P opensi y Sco e Reweigh ing and Ma ching Es ima o s.” The
Re iew o Economics and S a is ics 96 (5): 885–97.
Campos Ri e a, Héc o . 2018. “# Ve i icado19s: la Fo aleza de las Redes Sociales an e un
Te emo o.”
Ga ido, Melissa, Amy Kelley, Julia Pa is, Ka he ine Roza, Diane Meie , R Sean Mo i-
son, and Melissa Ald idge. 2011. “Me hods o Cons uc ing and Assessing P opensi y
Sco es.” Heal h Se ices Resea ch 49 (5).
Ge be , A.S., and D.P. G een. 2012. Field Expe imen s: Design, Analysis, and In e p e a ion.
W. W. No on. h ps://books.google.com.co/books?id=yxEGywAACAAJ.
INE, FEPADE, UNAM, T ibunal Elec o al del Pode Judicial de la Fede ación. 2016. Com-
pendio Legislación Nacional Elec o al. h ps://po alan e io .ine.mx/a chi os3/po al/
his o ico/ ecu sos/IFE- 2/DS/DS- Va ios/docs/2016/CompendioLegislacionNal/
Compendio-TomoII.pd .
Leu en, Edwin, and Ba ba a Sianesi. 2003. “PSMATCH2: S a a Module o Pe o m Full
Mahalanobis and P opensi y Sco e Ma ching, Common Suppo G aphing, and Co a i-
a e Imbalance Tes ing.” Accessed on June 11, 2020. A ailable a h ps://econpape s.
epec.o g/so wa e/bocbocode/s432001.h m.
Mo a, Ma iana, Ma ía Paula Sa on, and Pablo Gómez. 2018. “In es igación-acción du an e
Desas es: Uso de Redes y De echos.” Re is a Mexicana de Sociología 80 (SPE): 95–119.
16
