Labor scarcity, technology adoption and innovation: evidence from the cholera pandemics in 19th century France

F anck, Raphaël
A icle — Published Ve sion
Labo sca ci y, echnology adop ion and inno a ion:
e idence om he chole a pandemics in 19 h cen u y
F ance
Jou nal o Economic G ow h
P o ided in Coope a ion wi h:
Sp inge Na u e
Sugges ed Ci a ion: F anck, Raphaël (2024) : Labo sca ci y, echnology adop ion and inno a ion:
e idence om he chole a pandemics in 19 h cen u y F ance, Jou nal o Economic G ow h, ISSN
1573-7020, Sp inge US, New Yo k, NY, Vol. 29, Iss. 4, pp. 543-583,
h ps://doi.o g/10.1007/s10887-024-09241-3
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/315325
S anda d-Nu zungsbedingungen:
Die Dokumen e au EconS o dü en zu eigenen wissenscha lichen
Zwecken und zum P i a geb auch gespeiche und kopie we den.
Sie dü en die Dokumen e nich ü ö en liche ode komme zielle
Zwecke e iel äl igen, ö en lich auss ellen, ö en lich zugänglich
machen, e eiben ode ande wei ig nu zen.
So e n die Ve asse die Dokumen e un e Open-Con en -Lizenzen
(insbesonde e CC-Lizenzen) zu Ve ügung ges ell haben soll en,
gel en abweichend on diesen Nu zungsbedingungen die in de do
genann en Lizenz gewäh en Nu zungs ech e.
Te ms o use:
Documen s in EconS o may be sa ed and copied o you pe sonal
and schola ly pu poses.
You a e no o copy documen s o public o comme cial pu poses, o
exhibi he documen s publicly, o make hem publicly a ailable on he
in e ne , o o dis ibu e o o he wise use he documen s in public.
I he documen s ha e been made a ailable unde an Open Con en
Licence (especially C ea i e Commons Licences), you may exe cise
u he usage igh s as speci ied in he indica ed licence.
h p://c ea i ecommons.o g/licenses/by/4.0/
Vol.:(0123456789)
Jou nal o Economic G ow h (2024) 29:543–583
h ps://doi.o g/10.1007/s10887-024-09241-3
1 3
Labo sca ci y, echnology adop ion andinno a ion:
e idence om hechole a pandemics in19 h cen u y F ance
RaphaëlF anck1,2,3,4
Accep ed: 14 Janua y 2024 / Published online: 29 Feb ua y 2024
© The Au ho (s) 2024
Abs ac
To analyze he impac o labo sca ci y on echnology adop ion and inno a ion, his s udy
uses he di e en ial sp ead o chole a ac oss F ance in 1832, 1849 and 1854, be o e he
ansmission mode o his disease was unde s ood. The esul s sugges ha a la ge sha e o
chole a dea hs in he popula ion, which can be causally linked o summe empe a u e le -
els, had a posi i e and signi ican sho - un e ec on echnology adop ion and inno a ion
in ag icul u e bu a nega i e and signi ican sho - un impac on echnology adop ion in
indus y. These esul s can be explained by he posi i e impac o labo sca ci y on human
capi al o ma ion.
Keywo ds Epidemics· Labo sca ci y· Technology adop ion· Technology-skill
complemen a i y
JEL Classi ica ion I15· N13· O33
I hank Yoshiaki Azuma, G aziella Be occhi, Guillaume Blanc, B uno Cap e ini, F ancesco
Cinni ella, Céd ic Chamb u, E e Colson-Sih a, Nicola Fuchs-Schündeln, Oded Galo , Vé onique Gille,
Ta ek Ha chaoui, Ma iko Klasing, Pe os Milionis, Masao Ogaki, Nuno Palma, Josep Pijoan-Mas,
Niklas Po a ke, James Rockey, Ca la Sal o, Shmuel San and Joseph Zei a, con e ence pa icipan s
a he CEPR Mac oeconomics and Economic G ow h Mee ing, Economic His o y Socie y, Eu opean
Public Choice and Royal Economic Socie y, as well as semina pa icipan s a Cem i, CesI o, Doshisha
Uni e si y, G adua e Ins i u e o Policy S udies, Heb ew Uni e si y o Je usalem, HSE Moscow, IAST
Toulouse, Keio Uni e si y, Kyo o Uni e si y, Pa is-Dauphine Uni e si y, Uni e si y o Bi mingham,
Uni e si y o G oningen, Uni e si y o Manches e , Uni e si y o Sou he n Denma k and Uni e si y o
Tokyo o help ul commen s. Pa o his a icle was w i en a he CesI o Resea ch Ins i u e in Munich
whom I hank o i s hospi ali y. Idan Been p o ided excellen esea ch assis ance. I emain solely
esponsible o any mis akes.
This a icle is dedica ed o he memo y o Ya i Welzman.
* Raphaël F anck
Raphael.F anc[email p o ec ed]
1 Depa men o Economics, The Heb ew Uni e si y o Je usalem, Moun Scopus, 91905Je usalem,
Is ael
2 CEPR, London, UK
3 CesI o, Munich, Ge many
4 GLO, Essen, Ge many
544
Jou nal o Economic G ow h (2024) 29:543–583
1 3
1 In oduc ion
To explain echnology adop ion, heo e ical s udies ha e de eloped he mac oeconomic
implica ions o p oduc ion ac o s which can be ei he complemen a y o subs i u e (see,
e.g., Aghion & Howi , 1992; Zei a, 1998; Howi , 1999; Acemoglu, 2007; 2010; Alesina
e al., 2018). I labo and echnology a e complemen a y ac o s o p oduc ion, hen labo
sca ci y, whe eby skilled and/o unskilled wo ke s a e needed o ope a e machine y, is de -
imen al o echnology adop ion.1 I hey a e subs i u e, hen labo sca ci y leads o high
wages and is conduci e o echnology adop ion. Howe e , he e a e only a ew empi i-
cal analyses o he e ec s o labo sca ci y on echnology adop ion because ob aining a
quasi-expe imen al amewo k ha could p o ide causal e idence has u ned ou o be
challenging.
This s udy makes use o da a abou he chole a pandemics in 1832, 1849 and 1854
ac oss F ance o p o ide educed o m es ima es o he e ec o labo sca ci y on ech-
nology adop ion and inno a ion.2 In so doing, i asks he ollowing ques ions: (i) is labo
sca ci y conduci e o echnology adop ion in ag icul u e and in indus y o no , i.e., a e
p oduc ion ac o s in ag icul u e and in indus y complemen a y o subs i u e? (ii) is labo
sca ci y conduci e o echnological inno a ion? and (iii) is labo sca ci y conduci e o
echnology adop ion and inno a ion in bo h he sho - un and he mid- o long- un?
19 h cen u y F ance appea s well sui ed o such an empi ical analysis. Fi s , he coun-
y was hi ha shly by he chole a epidemics: i los 102,739 indi iduals in 1832, 102,500
in 1849 and 142,749 in 1854, i.e., abou 1% o he popula ion died o e 22 yea s.3 Howe e
some a eas we e hi mo e in ensely han o he s. Fo ins ance, he depa men o A iège in
he Sou h-Wes o F ance los 4.2% o i s popula ion du ing he 1854 pandemic. Second,
i was one o he i s coun ies o expe ience he indus ial e olu ion. Thi d, he F ench
e i o y had been di ided in small adminis a i e di isions o nea ly equal size in 1790
and hus, be o e he sp ead o chole a. Du ing he pe iod unde s udy, he e we e 85 depa -
men s which we e subdi ided in o 357 a ondissemen s: he a e age size o depa men s
was 6,228km
2
while ha o a ondissemen s was 1966km
2
.
In he cou se o he 19 h cen u y, scien is s o e ed compe ing heo ies on he sp ead
o chole a and i s cu e. Al hough English physician John Snow had al eady published his
i s indings in 1849, i was only in 1855 wi h he second edi ion o his book ha he con-
clusi ely demons a ed he ole o con amina ed wa e in he sp ead o he disease (Snow,
1855). And while I alian scien is Filippo Pacini had isola ed he Vib io Chole ae Bac e-
ium in 1854, i was only in 1884 ha Ge man scien is Robe Koch would iden i y he
Vib io Chole ae Bac e ium as he sou ce o he disease and subsequen ly p o ide a ea -
men (Koch, 1884). Scien is s ha e, by now, iden i ied he di e en modes o ansmission
o chole a (Glass & Black, 1992). In pa icula , o a coun y like F ance whose wea he
1 Se e al s udies (e.g., K eme , 1993; Ash a & Galo , 2011) no ed ha his o ically, echnological inno a-
ion occu ed in densely-popula ed a eas.
2 This pape hus di e s om s udies which use CES and/o Cobb–Douglas p oduc ion unc ions o assess
he a e o subs i u ion be ween labo and echnology. In his li e a u e (e.g., Knoblach & S öckl, 2019, o
a ecen su ey), speci ic assump ions on es ima ion equa ions and echnology dynamics ha e a subs an ial
impac on he es ima ed pa ame e s. We do no a emp o ep oduce ou main educed o m eg ession
esul s wi h a CES p oduc ion unc ion gi en he speci ici ies o ou da a as we discuss below.
3 To pu hese igu es in pe spec i e, es ima es sugges ha he Spanish lu in F ance killed abou 0.61% o
he popula ion a e WWI (238,000 ou o 39,108,000 inhabi an s) while he Co id-19 pandemic had killed
0.19% by 31 Decembe 2021 (123,805 ou o 66,314,842 inhabi an s) (Ansa e al., 2009)
545
Jou nal o Economic G ow h (2024) 29:543–583
1 3
is no wa m h oughou he yea , chole a is pa icula ly p one o ansmission in he sum-
me and speci ically, in egions which a e humid. In such an en i onmen , ansmission is
o en possible because he Vib io chole ae bac e ium can su i e o six o se en weeks
on d y clo hes which we e p e iously damp and swea y. In ac , because he basic ules o
mic obe ansmission and social dis ancing we e unknown a he ime, chole a was o en
sp ead du ing une al wakes when mou ne s would ouch he body o he dead and his/he
d y clo hes, he eby leading o he mis aken belie ha he disease sp ead h ough ai bo ne
“miasmas”.
Bu e en i he sp ead o chole a be o e 1855 was no unde s ood and could no be p e-
en ed, i is possible o conjec u e in hindsigh ha he di usion o he pandemics was co -
ela ed wi h local cha ac e is ics. While ou empi ical s a egy con ols o ime-in a ian
cha ac e is ics wi h ixed e ec s, i migh be he case ha chole a sp ead mo e easily in
a eas nea i e s whe e popula ion densi y inc eased be ween 1832 and 1854. Mo eo e ,
he ela ionship be ween labo sca ci y and echnology adop ion may ul ima ely e lec he
po en ial e ec o ins i u ional, geog aphical, and cul u al cha ac e is ics on he join e o-
lu ion o he labo supply and echnological p og ess. Gi en he po en ial endogenei y in
he ela ionship be ween labo sca ci y and echnology, and in ligh o he his o ical e i-
dence linking summe empe a u e le els and humidi y o he sp ead o chole a in F ance
(Delapo e, 1986; Bou delais & Raulo , 1987), his pape uses he his o ical wea he da a
o Lu e bache e al. (2004), Lu e bache e al. (2006) and Pauling e al. (2006) o es ablish
he causal impac o he chole a on echnology adop ion. The empi ical analysis shows ha
summe empe a u es in 1832, 1849 and 1854 ha e a causal impac in he local in ensi y o
chole a dea hs in he popula ion o each depa men . This inding is obus o using Ace-
moglu e al. (2020)’s maximum likelihood s a egy ha accoun s o in e pola ion conce ns
in he measu emen o empe a u e ac oss geog aphic uni s. Mo e gene ally, ou esul s a e
obus o alsi ica ion es s showing ha he sha e o chole a dea hs canno be explained by
o he seasonal empe a u e and ain all le els in o he yea s as well as o p e- ends es s o
obse able demog aphic and economic cha ac e is ics.
The esul s es ablish ha in he sho - un, a la ge sha e o chole a dea hs in he popula-
ion had a posi i e and signi ican e ec on echnology adop ion and inno a ion in ag i-
cul u e bu a nega i e and signi ican impac on echnology adop ion in indus y. As such,
ou esul s sugges ha labo and capi al a e subs i u e ac o s o p oduc ion in ag icul u e
and complemen a y in he indus ial sec o , in line wi h ecen s udies on he impac o
labo sca ci y ha ely on policy a ia ions in mig a ion (e.g., Ab ami zky e al., 2023;
San, 2023). Howe e ou indings indica e ha he e ec s o he chole a pandemics on
echnology adop ion and inno a ion we e quan i a i ely limi ed.4 A depa men expe ienc-
ing a median loss in popula ion because o he chole a epidemics (0.057%) would ha e
adop ed 0.28 addi ional mechanized ploughs pe day labo e o e he ollowing yea s bu
would ha e had 3.68 ewe s eam-powe ed machines pe wo ke in he yea a e each epi-
demic. These esul s a e obus o accoun ing o spa ial au oco ela ion using Colella e al.
(2020)’s app oach as well as o he e ogeneous ea men e ec s using he wo-way ixed
e ec s es ima o s o de Chaisema in and D’Haul oeuille (2020).
4 I is possible ha pandemics only ha e a majo economic e ec on economies when he dea h oll eaches
a high h eshold, e.g., when one hi d o he popula ion died du ing he Black Dea h in he Middle Ages.
Howe e , since he 19 h cen u y, no pandemic in coun ies ou o he Mal husian ap has killed ha many
people. The public policy implica ions o ou esul s he e o e call o a ca e ul app oach as he economic
consequences o pandemics may no be as dis up i e as one would hink.
546
Jou nal o Economic G ow h (2024) 29:543–583
1 3
Mo eo e , ou s udy sugges s ha he posi i e impac o labo sca ci y on human capi-
al accumula ion can explain ou main esul s. As popula ion loss inc eased he expec ed
e u ns o li e acy and li e a e wo ke s we e sough ou in indus ial wo k (e.g., Ka z &
Ma go, 2014; A ack e al., 2019; F anck & Galo , 2022), he ise in he sha e o li e acy
wo ke s in he popula ion o se he immedia e nega i e e ec o he popula ion losses on
echnology adop ion in indus y. In pa allel, his inc ease in li e a e wo ke s, who would
mos likely a oid low-paying wo k in ag icul u e, os e ed ag icul u al mechaniza ion.
Addi ional eg essions show ha his human capi al channel o ou esul s is obus o
accoun ing o mig a ion, u baniza ion, a cul u al shi as p oxied by a change in eligios-
i y, e ili y and nup iali y pa e ns as well as local inancial in e media ion.
This s udy is ela ed o h ee s ands o he economics li e a u e bu seeks o p o ide
a di e en pe spec i e. Fi s , i is ela ed o esea ch on pandemics, income shocks and
economic g ow h (e.g., Chak abo y e al., 2010; Adda, 2016; Rasul, 2020; Albanesi &
Kim, 2021). Pandemics could spu g ow h by inc easing a ailable esou ces o su i ing
indi iduals, especially o economies a he Mal husian s age o de elopmen (Lage lö ,
2003; Young, 2005; Siuda & Sunde, 2021).5 Howe e , i is di icul o asce ain he impac
o pandemics o coun ies ou o he Mal husian ap: while Amb us e al. (2020) ind a
long- e m impac o he 1854 chole a pandemic on po e y wi hin London, s udies on he
1918–1920 Spanish lu (e.g., Ba o e al., 2020; Jo dà e al., 2020; Lin & Meissne , 2020)
concu ha i had sho - e m nega i e e ec s bu di e as o i s ac ual long- un pe sis ence.
Second, his pape is ela ed o esea ch seeking o explain echnology adop ion du -
ing he indus ial e olu ion in he 19 h cen u y (e.g., Moky , 2009; Akcigi e al., 2017;
Juhász, 2018; Cap e ini & Vo h, 2020; F anck & Galo , 2022). Resea ch s a ing wi h
Habakukk (1962) has a gued ha labo sca ci y, and he ensuing high wages, led o he
adop ion o machine y. I is howe e unclea whe he high wages in England and he USA
ac ually s emmed om he ela i e abundance o coal o land, o om he p esence o
skilled wo ke s wi h high le els o p oduc i i y(see, e.g., Kelly e al., 2014; S ephenson,
2018). Rela edly, he ecen s udy o Vo h e al. (2022) uses exogeneous local a ia ion in
gende imbalance igge ed by mass consc ip ion in England du ing he Re olu iona y and
Napoleonic wa s, and inds ha his ype o labo sca ci y os e ed echnology adop ion in
he ea ly phase o he indus ial e olu ion.
Thi d, his s udy is ela ed o esea ch assessing he impac o labo ma ke condi ions
on he adop ion o labo -sa ing echnology: hese include Acemoglu and Finkels ein (2008)
on heal hca e, Manuelli and Seshad i (2014) and Ho nbeck and Naidu (2014) on ag icul-
u e, Lewis (2011) on manu ac u ing, Acemoglu and Res epo (2022) on he link be ween
demog aphic ac o s and echnology adop ion as well as Dechezlep ê e e al. (2019) on he
e ec s o labo cos s on au oma ion.6 In his espec , mos o he ecen li e a u e on labo
sca ci y akes ad an age o changes in mig a o y policies in he sho - and mid- un (e.g.,
Mose e al., 2014; Clemens e al., 2018; Ab ami zky e al., 2023; San, 2023). This s udy
howe e seeks o gi e a di e en pe spec i e by p o iding causal e idence o e a 50-yea
pe iod o he e ec s o labo sca ci y caused by a disease whose ansmission mode was
hen no unde s ood and which had no cu e.
5 The Black Dea h in Wes e n Eu ope seems o ha e been conduci e o g ow h in he long- un bu i s
e ec s we e di e en in Eas e n Eu ope (e.g., Voig lände & Vo h, 2013; Jedwab e al., 2019).
6 O he s udies dealing wi h he ela i e sca ci y o p oduc ion ac o s on echnological adop ion include
Newell e al. (1999) and Hassle e al. (2021) on he ise o ene gy p ices and sca ce na u al esou ces as
well as Hanlon (2015) on co on.

547
Jou nal o Economic G ow h (2024) 29:543–583
1 3
The emainde o his a icle is as ollows. Sec ion2 p esen s he da a and Sec .3 he
empi ical s a egy. Sec ion4 discusses he main esul s. Sec ion5 shows ha he inc ease
in human capi al explains ou main esul s and es ablishes ha al e na i e mechanisms do
no p o ide con incing explana ions. Sec ion6 concludes.
2 Da a
The da ase comp ises in o ma ion on he 85 depa men s and 357 a ondissemen s in
mainland F ance, as well as on indi iduals li ing ac oss he coun y, du ing and a e he
1832, 1849 and 1854 chole a pandemics.7 As we no e below, in o ma ion is some imes
missing o some ou come a iables immedia ely a e 1832 and in hose ins ances, we a e
he e o e compelled o es ic he sample o he a e ma h o he 1849 and 1854 pandem-
ics. Table A.1 epo s he desc ip i e s a is ics o he a iables in he empi ical analysis
ac oss he depa men s and a ondissemen s as well as o he a iables used in he indi-
idual-le el analysis. Tables A.2 and A.3 p o ide desc ip i e s a is ics o he addi ional
a iables employed in alsi ica ion es s and obus ness analyses.
2.1 Chole a ou b eaks
2.1.1 Chole a ansmission channels
Chole a is a wa e bo ne disease which is mos ib an be ween 15 and 25 Celsius deg ees.
Bu i d inking con amina ed wa e emains he mos well-known mode o ca ching he dis-
ease because o Snow (1855)’s seminal s udy, mode n esea ch (Glass & Black, 1992) has
demons a ed ha he e a e se e al ansmission channels o he chole a. The bac e ium
can indeed su i e and adap o di e en en i onmen s so ha i is no only obse ed in
popula ed loca ed ma i ime a eas, i e s and lakes, bu also in d y a eas, no ably in A ica
(S ock, 1976; Cli e al., 1986).
T ansmission modes o he disease o he han con amina ed wa e include con amina ed
ood, omi es (inanima e objec s such as clo hes ha ha e been exposed o he in ec ion) as
well as pe son- o-pe son ansmission. All hese ansmission channels in e ac wi h one
ano he unde a ious wea he condi ions o sp ead he disease. In ac , in he 19 h cen-
u y, ansmission was e y common along a el ou es as well as du ing une al wakes
when mou ne s ouched he body and he clo hes o he dead. This is possible because he
chole a bac e ium can emain ali e du ing six o se en weeks on d y clo hes which we e
con amina ed when hey we e damp and swea y.
Fu he mo e, he ecu en seasonal pa e n has also been shown o di e ac oss a i-
ous a eas o he wo ld. Fo ins ance, in he 320km-long Cheaseapeake Bay on he US
eas coas , i has been obse ed ha wa me summe empe a u es en ail a esu gence o
he chole a and ha he bac e ium is mo e p e alen in he no h o he Bay (whe e em-
pe a u es a e sligh ly lowe han in he sou h) because o he di e ence in he salini y and
7 The analysis is es ic ed o mainland F ance and excludes Co sica whe e no dea h om chole a was
eco ded in 1832 and 1849, and whe e he e we e only 220 chole a dea hs ou o 236,251 inhabi an s in
1854 (0.09% o he popula ion). Mo eo e , h ee new depa men s (Alpes-Ma i imes, Hau e-Sa oie and
Sa oie) we e added o F ance in 1860. Since hey we e no pa o F ance du ing he 1832, 1849 and 1854
pandemics, hey a e excluded om he analysis.
548
Jou nal o Economic G ow h (2024) 29:543–583
1 3
humidi y le els (Colwell, 2004; S .Lau en e al., 2021). As we discuss in he nex sec ion,
his is a simila pa e n o he incidence o chole a in 19
h
c. F ance, whe eby chole a was
mo e p e alen in he no h han in he sou h du ing he summe .
2.1.2 Chole a in19 h cen u y F ance
To build he main explana o y a iable on he in ensi y o chole a ou b eaks in 1832, 1849
and 1854, he s udy uses he o icial s a is ics p o ided by he F ench go e nmen on he
sha e o chole a dea hs wi hin he popula ion o each depa men (F ance, 1862). As can
be seen in Fig.1, he h ee chole a pandemics mainly a ec ed he no h o F ance and he
A lan ic Coas . The sou h o F ance was only hi ha shly in 1854.8 Only 11 depa men s
loca ed in he hin e land sou h-wes o he F ench e i o y we e spa ed in he h ee chole a
ou b eaks (Can al, Co èze, C euse, Do dogne, Ge s, Landes, Lo , Lozè e, Hau es-Py é-
nées, Vienne and Hau e-Vienne).
He e wo ema ks a e impo an . Fi s i mus be no ed ha be o e 1855, he ansmis-
sion mode o he chole a had no been conclusi ely es ablished. A a ime whe e basic
knowledge abou mic obes was jus being disco e ed, some scien is s we e mis akenly
a guing ha he e we e ai bo ne “miasmas”which explained he di usion o he disease. As
such, a oiding pollu ed wa e sou ces, as well as p ope hygiene and social dis ancing, did
no play a ole in he beha io o indi iduals: since no-one knew how he disease sp ead,
i was no e en clea ha unning away om a eas a ec ed by he chole a could o e any
p o ec ion.9 Second, he disease was a p oblem o he cen al S a e, he local go e nmen-
al au ho i ies, he Chu ch as well as he local associa ions. Howe e he e was no heal h
policy which any go e nmen o o ganiza ion could implemen o s op he disease.
As can be seen in Table1, he dis ibu ion o chole a dea hs wi hin he popula ion
o each depa men is skewed: he 25 h pe cen ile is equal o 0, he median 0.057%, he
75 h pe cen ile 0.30% and he 99 h pe cen ile 2.84%. This e lec s he ac ha he disease
eached mos depa men s a leas once in ei he 1832, 1849 and 1854, bu only a ew we e
hi ha shly. None heless, 20 depa men s los mo e han 1% o hei popula ion in a leas
one o he h ee ou b eaks.
Tables B.1 and B.2 p o ide addi ional desc ip i e s a is ics and es s ega ding he sha e
o chole a dea hs in he popula ion. Table B.1 dis inguishes be ween he gende and age
o he ic ims du ing he 1854 pandemic while Table B.2 ocuses on he sha e o ic ims
by dis inguishing depa men s by hei mean and median popula ion in each o he h ee
pandemics.
The es s o means epo ed in bo h Tables B.1 and B.2 a e ne e signi ican , he eby
alle ia ing conce ns ha some sec ions o he popula ion would be mo e (o less) likely
o die om exposu e o he chole a. In pa icula , he es s in Table B.1 sugges ha ou
esul s canno be d i en by he gende and/o age o he chole a ic ims wi hin he popula-
ion o he depa men s hi by he chole a while hose in Table B.2 indica e ha hey canno
be d i en by he size o he depa men al popula ion and hence by he p opensi y o he
8 Anecdo al e idence sugges s ha each ime, he chole a came by boa om England. I only sp ead o he
sou h-eas o F ance in 1854 because o he F ench soldie s who emba ked om he sou he n ha bo s o
Toulon and Ma seille o igh he wa in C imea.
9 The F ench popula ion soon came o e e o he chole a as he “blue ea ”(peu bleue) because o he
blue colo a ion ha he aces o sick indi iduals would ake jus be o e dying. The exp ession “peu
bleue”is s ill commonly used in F ench and e e s o some hing which is e i ying.
549
Jou nal o Economic G ow h (2024) 29:543–583
1 3
ic ims o inhabi u ban o u al depa men s. The esul s a e no su p ising and in line
wi h he discussion in he p e ious sub-sec ion ha pe son- o-pe son ansmission was a
majo cause o he sp ead o he chole a in he 19 h cen u y, and was as common as con-
amina ion h ough pollu ed wa e . I is he e o e no su p ising ha he p e alence o he
chole a is no ela ed o popula ion densi y and social s a us.
I is wo h no ing ha he e we e addi ional chole a ou b eaks in 19 h cen u y F ance,
i.e., in 1884 and 1892. Howe e , hey occu ed a e 1855, when he ansmission mode
o he chole a had been inally es ablished by Snow (1855). As a esul , i is p e e able o
es ic he main analysis o he p e-1855 chole a ou b eaks: his a oids endogenei y con-
ce ns ha speci ic a eas migh become mo e e icien han o he s in p e en ing he sp ead
o he disease once he mode o con agion was known. In his espec , we show in Table
B.3 ha he sp ead o chole a be o e 1855 was no co ela ed wi h i s sp ead in 1884 and
1892 whose consequences we e mo e limi ed because local au ho i ies hen unde s ood
and could p e en he di usion o he disease. Table B.4 u he shows ha he 1832, 1849
and 1854 chole a pandemics we e no co ela ed wi h he a ious causes o dea hs in each
depa men in 1855. Mo eo e , Table B.5 shows ha he chole a pandemics in 1832, 1849
and 1854 a e no co ela ed wi h he sp ead o illnesses be o e he 19 h cen u y inso a as
he e is no co ela ion wi h he numbe o owns hi by he sp ead o he plague in he 18 h
c. in each depa men .
2.2 Summe empe a u e in19 h cen u y F ance
As es ablished by mode n esea ch (e.g., Glass & Black, 1992), he Vib io Chole ae Bac-
e ium quickly sp eads in humid en i onmen s whe e empe a u es a e abo e 15 deg ees
Celsius. This implies wo p edic ions o he di usion o chole a in F ance. Fi s , chol-
e a mainly sp eads du ing he summe because his is he season when empe a u es in
F ance a e abo e 15 deg ees Celsius o a long ime pe iod. Second, chole a is mo e likely
o sp ead in he No h han in he Sou h o F ance because ela i e humidi y is always
highe in no he n a eas whe e empe a u es a e always ela i ely lowe . While his sec-
ond poin migh seem sligh ly coun e -in ui i e o he eade because humans eel humidi y
mo e accu a ely (and hence expe ience mo e discom o ) a highe le els o empe a u e, i
is ac ually he case ha ela i ely lowe empe a u es en ail mo e ela i e humidi y because
hey enable o less wa e e apo a ion (Wallace & Hobbs, 1977; Lu gens & Ta buck, 2015).
In he case o F ance, he eg ession esul s in Table C.1 use mode n wea he da a om 42
wea he s a ions in 2018 and es ablish ha lowe empe a u es a e indeed associa ed wi h
highe ela i e humidi y, accoun ing o wea he s a ion ixed e ec s as well as mon h-,
day- and hou - ixed e ec s.10
Gi en he p ope ies o he Vib io Chole ae Bac e ium and he his o ical con ex , ou
iden i ica ion s a egy p edic s ha (1) empe a u es in he summe o 1832, 1849 and
1854, and no in any o he season o in any o he yea , a e signi ican ly co ela ed wi h he
sp ead o chole a because his is he only ime pe iod whe e empe a u es emain abo e
10 The nega i e co ela ion be ween empe a u e and ela i e humidi y is no speci ic o F ance. Fo
ins ance, (2019, Table1) epo ha in China, whe e empe a u es in he No h a e lowe han in he Sou h,
he e is a nega i e co ela ion be ween mean empe a u e and ela i e humidi y h oughou he yea ha is
only signi ican a he 5% le el du ing he summe . Fo he sake o he a gumen , i should also be no ed
ha he Saha a dese is loca ed o he Sou h o he Medi e anean sea and ha his dese ic a ea is d ye
han he coas al Medi e anean a eas o No h A ica.
550
Jou nal o Economic G ow h (2024) 29:543–583
1 3
15 deg ees Celsius and ha (2) summe empe a u e le els in 1832, 1849 and 1854 would
be nega i ely co ela ed wi h he sp ead o chole a because no he n F ench depa men s
expe ienced ela i ely lowe empe a u es, and hence mo e ela i e humidi y, han sou he n
depa men s. Anecdo al e idence on he mon hly sp ead o chole a in 1854 seems o sup-
po his p edic ion: Fig.2 shows ha he disease sp ead om he no h o he coun y and
claimed he highes numbe o ic ims in July, Augus and Sep embe .11
Ou s udy elies on he his o ical wea he da a o Lu e bache e al. (2004), Lu e bache
e al. (2006) and Pauling e al. (2006). These da a we e econs uc ed using a ious sou ces
such as lake sedimen s and ee ings as well as his o ical eco ds o e e y season o e
he 1500–1900 pe iod a a esolu ion o 0.5 by 0.5 decimal deg ees. The e a e he e o e
conce ns abou measu emen e o and he in e pola ion o clima ic da a o e depa men s,
i.e., wo cells pe depa men on a e age. S ill Lu e bache e al. (2004), Lu e bache e al.
(2006) and Pauling e al. (2006) show ha he quali y o he da a imp o e o e ime, espe-
cially om he end o he 18
h
c. onwa d. Figu e3 maps hose da a o he summe s o
Fig. 1 Sha e o chole a dea hs ou o depa men al popula ion, 1832, 1849 and 1854. No e The sou ce o
he map laye is Daudin e al. (2019).
Table 1 The dis ibu ion o he
pe cen age o chole a dea hs in
he popula ion ac oss F ench
depa men s in 1832, 1849 and
1854
This able epo s desc ip i e s a is ics o he pe cen age o chole a
dea hs in he popula ion ac oss he 85 F ench depa men s in 1832,
1849 and 1854. The o al F ench popula ion amoun ed o 32,443,430
inhabi an s in 1832, 36,910,360 in 1849 and 35,782,708 in 1854.
Mean 25
h
50
h
75
h
90
h
99
h
1832 0.26 0 0.01 0.26 0.86 2.35
1849 0.20 0 0.02 0.22 0.88 1.70
1854 0.46 0.009 0.16 0.61 1.36 4.20
All Yea s
Combined
0.31 0 0.06 0.30 0.90 2.84
11 Mo e gene ally, i mus be ha acknowledged ha wa e bodies could ha e inc eased he ansmission
o he chole a. Howe e , since he empi ical s a egy which we discuss in Sec .3 uses depa men ixed
e ec s, i is unlikely ha wa e bodies can sys ema ically explain a ia ions in he sp ead o he chole a in
1832, 1849 and 1854.
557
Jou nal o Economic G ow h (2024) 29:543–583
1 3
3.2 Summe empe a u es andchole a dea hs in hepopula ion: i s ‑s age
eg ession esul s and es s o p e‑ ends
3.2.1 Fi s ‑s age eg ession esul s
In line wi h he his o ical e idence on he sp ead o chole a in 19 h cen u y F ance, whe e
he disease mainly hi no he n depa men s du ing he summe s o 1832, 1849 and 1854,
Table 3 shows ha he summe empe a u e ins umen has a nega i e and signi ican
e ec on he sha e o chole a dea hs in he popula ion ( he comple e speci ica ions wi h
he con ol a iables a e shown in Table D.1). In all he speci ica ions using obus clus-
e ed s anda d e o s a he depa men le el, his nega i e e ec is signi ican a he 1%
le el. To ensu e he obus ness o ou esul s, we also compu e he s anda d e o s wi h he
maximum likelihood es ima ion s a egy o Acemoglu e al. (2020) ha co ec s o meas-
u emen e o and geog aphic co ela ion in ain all measu emen . These s anda d e o s
a e epo ed in cu ly b acke s in Table3: hey con i m he signi ican and nega i e e ec o
summe empe a u e on he sha e o chole a dea hs in he popula ion.
The es ima e in Column 1 o Table3 sugges s ha a 1% dec ease in summe empe a-
u e le els inc eased he sha e o chole a dea hs in he popula ion by 11.8%. Hence, o a
depa men expe iencing a dec ease in empe a u e om he 75 h pe cen ile o summe
empe a u e (18.10 deg ees Celsius) o he 50 h pe cen ile (i.e., 17.38 deg ee Celsius), his
4.03% dec ease in empe a u e would en ail 0.6% mo e in he sha e o chole a dea hs in he
popula ion, i.e., a decline equal o one s anda d de ia ion. Thus, in line wi h he his o ical
e idence, hese compu a ions sugges ha he successi e chole a pandemics en ailed a sub-
s an ial loss o popula ion.
3.2.2 Falsi ica ion es s and obus ness checks o p e‑ ends
To enhance he c edibili y o ou iden i ica ion s a egy, we p esen se e al alsi ica ion
es s and obus ness checks o p e- ends. They show ha nei he summe empe a u es
no chole a dea hs a e co ela ed wi h po en ially omi ed a iables pe aining o he p e-
exis ing cha ac e is ics o he depa men s ha could d i e hei ulne abili y o he chole a
epidemics and hei subsequen adop ion o echnology.
No e ha we al eady discussed he ollowing obus ness checks in Sec .2: (i) Tables B.1
and B.2 show ha all popula ion g oups (dis inguished by age o gende , u ban o u al)
we e equally a ec ed by he chole a; (ii) Tables B.3 and B.4 show ha he numbe s o ic-
ims in he 1832, 1849 and 1854 chole a pandemics we e no co ela ed wi h he numbe s
o ic ims om a ious causes o dea h in each depa men in 1855 o wi h he numbe s o
ic ims in he mino chole a ou b eaks in 1884 and 1892 (which occu ed a e he ans-
mission mode o he disease was unde s ood); (iii) Table B.5 shows ha he di usion o
chole a pandemics in 1832, 1849 and 1854 is no co ela ed wi h he numbe o owns hi
in each depa men by he sp ead o he plague in he 18
h
cen u y and (i ) Table B.7 shows
ha he e a e no signi ican di e ences in he p ices o impo ed machine y in ag icul u e
and indus y ha could po en ially d i e he esul s.

558
Jou nal o Economic G ow h (2024) 29:543–583
1 3
In wha ollows, we summa ize he addi ional alsi ica ion es s which we ca y ou in
suppo o ou iden i ica ion s a egy. In he Appendix, we p esen he da a sou ces and
epo he eg ession esul s.
Chole a, empe a u es and ain all. Because wea he da a a e co ela ed o e ime, a
po en ial conce n ega ding he iden i ica ion s a egy is ha he signi ican e ec o sum-
me empe a u e le els on chole a dea hs in he yea o each pandemic can be a ibu ed
o he gene al e ec o summe empe a u es in o he yea s, and is co ela ed wi h em-
pe a u es in o he seasons and wi h ain all. Reassu ingly, he sha e o chole a dea hs is no
co ela ed wi h summe empe a u es in he yea s jus be o e o a e he chole a ou b eaks
in Table D.2. Mo eo e , in he yea s o chole a ou b eaks, he sha e o chole a dea hs in
he popula ion is no co ela ed wi h empe a u es in sp ing, all and win e in Table D.3,19
wi h summe empe a u e shocks in Table D.4 and wi h ain all in sp ing, all and win e
in Table D.5.20 In Table D.5 howe e , summe ain all is signi ican ly co ela ed wi h he
Table 3 Summe empe a u e le els and sha e o chole a dea hs in he popula ion
This able epo s he i s s age es ima es ela ing summe empe a u e le els o he sha e o chole a dea hs
in he popula ion in 1832, 1849 and 1854. Geog aphic con ols o depa men s, which a e in e ac ed wi h
yea - ixed e ec s, include hei land sui abili y, hei sha e o ca boni e ous a ea and dummies o bo de
and ma i ime depa men s. Cons an no epo ed. All a iables a e in loga i hm. Robus s anda d e o s
clus e ed a he depa men le el a e epo ed in b acke s. Robus s anda d e o s clus e ed a he depa e-
men le el using he Maximum Likelihood app oach o Acemoglu e al. (2020) a e epo ed in cu ly b ack-
e s.
∗∗∗ p
<
0.01,∗∗ p
<
0.05,∗p
<
0.1
(1) (2) (3)
Fi s s age: he ins umen ed a iable is Sha e o Chole a Dea hs in Popula ion
Summe empe a u e
−
0.118***
−
0.141***
−
0.140***
[0.0271] [0.0303] [0.0308]
{0.044}∗∗∗
{0.058}∗∗
{0.061}∗∗
Mean dep. a 0.0031 0.0031 0.0031
1s s age F-s a 19.012 21.652 20.788
Mo an I
−
0.008
−
0.008
−
0.008
Mo an I p- alue 0.212 0.209 0.210
Depa men and yea ixed e ec s Yes Yes Yes
De ia ion om summe ain all No Yes Yes
Geog aphic con ols * yea ixed e ec s No Yes Yes
GDP pe capi a No No Yes
Clus e s 85 85 85
Obse a ions 255 255 255
19 As can be seen in Table D.3, he e a e speci ica ions whe e empe a u es in o he seasons a e some imes
signi ican ly, bu no sys ema ically, co ela ed wi h he sp ead o he chole a. In o he wo ds, hey may no
indica e a signi ican impac on he sp ead o he chole a so much as a empo al co ela ion wi h summe
empe a u es. Thus hese eg essions sugges ha i is bes o use summe empe a u e as he sole ins u-
men , which is signi ican h oughou , ins ead o che y-picking he da a and use di e en seasonal empe a-
u es in di e en eg essions.
20 The esul s in Table D.4 p o ide suppo o he alidi y o he exclusion es ic ion: di e en empe a-
u es explain he di e en ial impac o he pandemics ac oss he F ench depa men s, bu hey we e no
ou lie s and he e o e we e unlikely o impac o he a iables, mos no ably ag icul u al yields. In Sec .4
559
Jou nal o Economic G ow h (2024) 29:543–583
1 3
sp ead o he chole a in he i s s age eg essions, al hough i s e ec is quan i a i ely small.
Since summe ain all is no sys ema ically signi ican in many 2nd s age and educed o m
eg essions, i seems ha i s limi ed impac is cap u ed by summe empe a u e.
Fu he mo e, Table D.6 shows ha he e is a s a is ically signi ican ela ionship
be ween summe empe a u es in 1832, 1849 and 1854 in log and le el, and he sha e o
chole a dea hs in 1832, 1849 and 1854, bu no be ween he la e and squa ed summe
empe a u es in le el (which gi es equal weigh o abno mal ain all and d ough s) in hose
yea s. Finally Table D.7 shows ha he e is a s a is ically signi ican ela ionship be ween
summe empe a u es in 1832, 1849 and 1854 in log and le el, and he sha e o chole a
dea hs in 1832, 1849 and 1854, bu no a sys ema ically signi ican ela ionship be ween
he la e and seasonal ain all, e en when he in e ac ion a iables be ween summe em-
pe a u es and seasonal ain all a e included.
P e-pandemic ade and indus y. A po en ial conce n ega ding he exogenei y o he
ela ionship be ween summe empe a u e and chole a dea hs pe ains o ade and indus-
y. In pa icula , i is possible ha he anspo o goods wi hin F ance, and he associ-
a ed ci cula ion o people, would be co ela ed wi h wea he condi ions and would ha e
an impac on he sp ead o he pandemic. Reassu ingly, bo h Tables D.8 and D.9 show ha
he e is no ela ionship be ween in e nal ade and empe a u e as well as be ween in e nal
ade and he sp ead o chole a.
In addi ion, Table D.10 shows ha summe empe a u e and echnology adop ion in
indus y we e no co ela ed be o e he i s chole a pandemic in 1832. Namely, in 1789,
1811 and 1815, summe empe a u es had no signi ican impac on he numbe s o i on
o ges and mechanical mills in he co on indus y.
En y poin s and di usion hubs I is possible ha he local sha e o chole a dea hs could
be co ela ed wi h he ini ial poin o en y o he disease o wi h a speci ic hub o di usion
such as a majo ci y. Un o una ely, no-one knows wi h absolu e ce ain y wha he dis-
ease’s poin s o en y we e. Each ime, he chole a mos likely came om England. A bes ,
i can be said ha he i s cases o chole a we e no iced in he Manche depa men in 1832,
in he Aisne depa men in 1849 and in he No d depa men in 1854. The Manche and
No d depa men s a e loca ed on he English Channel bu he Aisne depa men is a land-
locked a ea, making i e en less likely o de e mine he poin o en y in 1849 (Bou delais
& Raulo , 1987; F ance - Minis è e de l’ag icul u e, 1862). Table D.11 es s he hypo hesis
ha he in ensi y o he chole a pandemics would be co ela ed wi h London o wi h po en-
ial poin s o en y and di usion hubs such as Pa is, ha bo s like Rouen and Ma seille, o
F esnes-su -Escau , he mining illage in he No d depa men whe e he i s s eam engine
was used o indus ial pu poses in F ance in he 18 h cen u y. The esul s show ha he
dis ance be ween each o hese hubs and he main adminis a i e cen e o each depa men
has no impac on he signi ican e ec o summe empe a u e on he local sha e o chole a
dea hs in 1832, 1849 and 1854.
P e-pandemic cha ac e is ics o he popula ion. Table D.12 shows ha he i s s age
ela ionship is no in luenced by omi ed a iables linking summe empe a u es and he
numbe o dea hs in each depa men o e ime. Fu he mo e, Tables D.13 and D.14 show
ha summe empe a u es and chole a dea hs we e no co ela ed wi h he numbe and
below, we p o ide e idence o he lack o a pe sis en impac o he pandemics on ag icul u al yields and
land en s.
Foo no e 20 (con inued)
560
Jou nal o Economic G ow h (2024) 29:543–583
1 3
densi y o inhabi an s as well as wi h he age s uc u e o each depa men p io o he
1832, 1849 and 1854 chole a pandemics.21
P e-pandemic human capi al and weal h. I could be conjec u ed ha he sha e o chol-
e a dea hs in he popula ion was co ela ed wi h he ela i e p esence o poo / ich indi idu-
als o o educa ed/uneduca ed indi iduals. While he e is no his o ical e idence sugges ing
ha he chole a ic ims we e cha ac e ized by speci ic social s a uses o income le els,
Tables D.15, D.16, D.17 a e mean o assuage conce ns ega ding a possible link be ween
chole a dea hs, educa ion and weal h.
Thus, in line wi h he his o ical e idence, Table D.15 shows ha he chole a claimed
ic ims among di e en occupa ional g oups, whe he ich (e.g., shipowne s), poo (e.g.,
enan a me s) o educa ed (e.g., cle gymen, p o esso s & eache s).22 Fu he mo e, Table
D.16 shows ha he e is no signi ican ela ionship be ween he sha e o chole a dea hs in
he popula ion, he p obabili y ha he dead le an inhe i ance as well as he alue o he
inhe i ance. Finally, Table D.17 shows ha he chole a pandemics we e no co ela ed wi h
human capi al as p oxied by he likelihood ha indi iduals bo n one o 20 yea s be o e
each pandemic could sign hei wedding license (as opposed o ma k i wi h a c oss).
4 Resul s: Sho ‑ e m e ec s on echnology adop ion andinno a ion
This sec ion explo es he e ec o he chole a pandemics on echnology adop ion and inno-
a ion in ag icul u e and indus y. The eg ession esul s in Tables4, 5, 6 and 7 sugges ha
he chole a epidemics had sho - e m and quan i a i ely small e ec s on echnology adop-
ion and inno a ion (Appendix E epo s he eg ession esul s wi h he ull se o con ols).
These e ec s we e conduci e o echnology adop ion in ag icul u e bu no in he indus ial
sec o . In addi ion, he esul s sugges ha he chole a epidemics en ailed labo ealloca ion
om he ag icul u al o he indus ial sec o s. They also indica e ha he nega i e e ec s o
labo sca ci y las ed longe in he ex ile sec o han in he indus ial sec o . The esul s a e
obus o he inclusion o con ol a iables, including GDP pe capi a, he eby making i
unlikely ha hey a e d i en by sho - e m nega i e income e ec s.
In ou esul s, ou IV es ima es o he e ec o he chole a epidemics on echnology
adop ion a e wo o h ee imes la ge han he OLS coe icien s. A possible in e p e a-
ion o hese indings is ha ou eg essions su e om e o s in a iables and a enua-
ion bias: while he e is no e idence ha he local ci il se an s who collec ed da a on he
numbe o chole a dea hs sough o minimize o in la e he impac o he epidemics, some
migh ha e collec ed da a mo e diligen ly han o he s. Ano he explana ion is ha ou IV
es ima es e lec he expec a ions o indi iduals ega ding he consequences o he chole a
22 The size o he coe icien s in Table D.15 is no he same o all occupa ions. Howe e , i is p obably bes
no o p o ide an in e p e a ion o he size o he coe icien s. I is possible o specula e ha ex ile wo ke s
we e mo e nega i ely a ec ed han blacksmi hs o p o esso s because hei sec o en ailed mo e ade and
in e ac ion. Howe e , he coe icien eg essions also sugges ha anspo wo ke s and membe s o he
cle gy we e less a ec ed by he chole a in 1854 han ex ile wo ke s, e en hough hey also in e ac ed wi h
many sec ions o he popula ion.
21 The eg ession in Column 1 o Table D.13 has one ewe obse a ion han he o he eg essions (254
ins ead o 255) because he Ta n-e -Ga onne depa men was only c ea ed in 1808. I s u u e e i o y was
spli be ween neighbou ing depa men s (mos ly Lo and Hau e-Ga onne). We chose no o adjus he GDP
pe capi a o hese depa men s o his alsi ica ion es because i is unclea how we would accoun o he
exac income di e ences be ween he a ious a eas o Lo and Hau e-Ga onne.
561
Jou nal o Economic G ow h (2024) 29:543–583
1 3
Table 4 The e ec s o he chole a in 1849 and 1854 on he Numbe and ho se powe o machines pe wo ke in he mining indus y one yea a e each pandemic
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
OLS OLS OLS 2SLS 2SLS OLS OLS OLS 2SLS 2SLS
A e age numbe o s eam-powe ed machines A e age ho se powe o s eam-powe ed machines
pe Wo ke Yea +1 pe Wo ke Yea +1
Sha e o chole a dea hs in popula ion
−
24.33***
−
30.79***
−
28.51***
−
75.52**
−
64.49**
−
32.98**
−
37.52**
−
34.61**
−
104.7**
−
91.55**
[8.538] [9.632] [8.720] [31.71] [28.28] [13.66] [15.18] [14.38] [46.60] [43.80]
Depa men and yea ixed e ec s Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
De ia ion om summe ain all No Yes Yes Yes Yes No Yes Yes Yes Yes
Geog aphic con ols No Yes Yes Yes Yes No Yes Yes Yes Yes
GDP pe capi a No No Yes No Yes No No Yes No Yes
Wi hin R2 0.174 0.228 0.255 0.123 0.137 0.155
Mean dep. a 38.709 38.709 38.709 38.709 38.709 401.809 401.809 401.809 401.809 401.809
Clus e s 85 85 85 85 85 85 85 85 85 85
Obse a ions 170 170 170 170 170 170 170 170 170 170
Fi s s age: he ins umen ed a iable is sha e o chole a dea hs in popula ion
Summe Tempe a u e
−
0.180***
−
0.179***
−
0.180***
−
0.179***
[0.0471] [0.0485] [0.0471] [0.0485]
1s s age F-s a 14.602 13.577 14.602 13.577
Reduced o m: he dependen a iable is
A e age numbe o s eam-powe ed machines A e age ho se powe o s eam-powe ed machines
Pe Wo ke Yea +1 Pe Wo ke Yea +1
Summe Tempe a u e 13.60** 11.53** 18.86** 16.37**
[5.977] [5.357] [8.808] [8.172]
562
Jou nal o Economic G ow h (2024) 29:543–583
1 3
This able p esen s OLS and IV eg essions ela ing he sha e o chole a dea hs in each depa men o he numbe and ho se powe o s eam-powe ed machines pe wo ke as
well as o he numbe o boile s and s eam gene a o s pe wo ke in he mining sec o in he yea a e each chole a ou b eak. Geog aphic con ols o depa men s, which a e
in e ac ed wi h yea - ixed e ec s, include hei land sui abili y, hei sha e o ca boni e ous a ea and dummies o bo de and ma i ime depa men s. Cons an no epo ed.
Robus s anda d e o s clus e ed a he depa men le el.
∗∗∗ p
<
0.01,∗∗ p
<
0.05,∗p
<
0.1
Table 4 (con inued)
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
OLS OLS OLS 2SLS 2SLS OLS OLS OLS 2SLS 2SLS
A e age numbe o s eam gene a o s A e age numbe o boile s
Pe Wo ke Yea +1 Pe Wo ke Yea +1
Sha e o Chole a Dea hs in Popula ion
−
21.29**
−
27.51***
−
24.89***
−
86.21**
−
74.20**
−
13.64
−
20.14*
−
20.51*
−
90.47**
−
99.37**
[8.887] [10.31] [9.427] [34.05] [30.86] [10.50] [10.35] [10.36] [38.90] [41.65]
Depa men and yea ixed e ec s Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
De ia ion om summe ain all No Yes Yes Yes Yes No Yes Yes Yes Yes
Geog aphic con ols No Yes Yes Yes Yes No Yes Yes Yes Yes
GDP pe capi a No No Yes No Yes No No Yes No Yes
Wi hin R2 0.131 0.182 0.215 0.211 0.329 0.330
Mean dep. a 40.091 40.091 40.091 40.091 40.091 45.667 45.667 45.667 45.667 45.667
Clus e s 85 85 85 85 85 85 85 85 85 85
Obse a ions 170 170 170 170 170 170 170 170 170 170
Fi s s age: he ins umen ed a iable is sha e o chole a dea hs in popula ion
Summe empe a u e
−
0.180***
−
0.179***
−
0.180***
−
0.179***
[0.0471] [0.0485] [0.0471] [0.0485]
1s s age F-s a 14.602 13.577 14.602 13.577
Reduced o m: he dependen a iable is
A e age Numbe o S eam Gene a o s pe Wo ke Yea +1 A e age numbe o boile s pe wo ke yea +1
Summe empe a u e 15.53** 13.26** 16.30** 17.76**
[6.325] [5.718] [7.479] [7.237]

563
Jou nal o Economic G ow h (2024) 29:543–583
1 3
Table 5 The e ec s o he chole a in 1849 and 1854 on employmen , wages and p oduc ion in he wake o each pandemic
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
OLS OLS OLS 2SLS 2SLS OLS OLS OLS 2SLS 2SLS
A e age Numbe o Wo ke s Yea +1 A e age Wage pe Wo ke Yea +1
Sha e o chole a dea hs in popula ion 3.960 9.768 8.128 38.65 30.41
−
24.46
−
11.48
−
12.78 99.52 101.0
[7.685] [9.736] [8.917] [33.55] [31.01] [22.63] [22.71] [22.23] [93.35] [92.91]
Depa men and yea ixed e ec s Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
De ia ion om summe ain all No Yes Yes Yes Yes Yes Yes Yes Yes No
Geog aphic con ols No Yes Yes Yes Yes Yes Yes Yes Yes No
GDP pe capi a No No Yes No Yes No Yes No Yes No
Wi hin R2 0.064 0.326 0.341 0.009 0.130 0.131
Mean dep. a 176.404 176.404 176.404 176.404 176.404 66.535 66.535 66.535 66.535 66.535
Clus e s 85 85 85 85 85 85 85 85 85 85
Obse a ions 170 170 170 170 170 170 170 170 170 170
Fi s s age: he ins umen ed a iable is sha e o chole a dea hs in popula ion
Summe empe a u e
−
0.180***
−
0.179***
−
0.180***
−
0.179***
[0.0471] [0.0485] [0.0471] [0.0485]
1s s age F-s a 14.602 13.577 14.602 13.577
Reduced o m: he dependen a iable is
A e age numbe o wo ke s yea +1 A e age wage pe wo ke yea +1
Summe Tempe a u e
−
6.961
−
5.436
−
17.93
−
18.06
[6.389] [5.804] [16.01] [15.70]
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
OLS OLS OLS 2SLS 2SLS OLS OLS OLS 2SLS 2SLS
A e age Value o Ex ac ed Coal ( +2)-( +3) A e age Value o Ex ac ed Pea ( +2)-( +3)
Sha e o Chole a Dea hs in Popula ion
−
2.765
−
2.409 3.615
−
2.625 2.732
−
9.316**
−
11.37**
−
10.98**
−
25.95**
−
24.71*
[2.242] [1.832] [6.429] [1.867] [6.724] [4.672] [5.090] [5.254] [12.06] [13.79]
Depa men - and yea ixed e ec s Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
De ia ion om summe ain all No Yes Yes Yes Yes No Yes Yes Yes Yes
564
Jou nal o Economic G ow h (2024) 29:543–583
1 3
This able p esen s OLS and IV eg essions ela ing he sha e o chole a dea hs in each depa men o he numbe and wage o wo ke s in he mining sec o in he yea a e
each chole a ou b eak and o he alues o ex ac ed coal and pea wo and h ee yea s a e each chole a ou b eak. Geog aphic con ols o depa men s, which a e in e ac ed
wi h yea - ixed e ec s, include hei land sui abili y, hei sha e o ca boni e ous a ea and dummies o bo de and ma i ime depa men s. Cons an no epo ed. Robus s and-
a d e o s clus e ed a he depa men le el.
∗∗∗ p
<
0.01,∗∗ p
<
0.05,∗p
<
0.1
Table 5 (con inued)
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
OLS OLS OLS 2SLS 2SLS OLS OLS OLS 2SLS 2SLS
A e age Value o Ex ac ed Coal ( +2)-( +3) A e age Value o Ex ac ed Pea ( +2)-( +3)
Geog aphic con ols No Yes Yes Yes Yes No Yes Yes Yes Yes
GDP pe capi a No No Yes Yes No No No Yes No Yes
Wi hin R2 0.091 0.157 0.162 0.367 0.464 0.467
Mean dep. a 0.527 0.527 0.527 0.527 0.527 0.234 0.234 0.234 0.234 0.234
Clus e s 85 85 85 85 85 85 85 85 85 85
Obse a ions 170 170 170 170 170 170 170 170 170 170
Fi s s age: he ins umen ed a iable is sha e o chole a dea hs in popula ion
Summe empe a u e
−
0.180***
−
0.179***
−
0.180***
−
0.179***
[0.0471] [0.0485] [0.0471] [0.0485]
1s s age F-s a 14.602 13.577 14.602 13.577
Reduced o m: he dependen a iable is
A e age alue o ex ac ed coal ( +2)-( +3) A e age alue o ex ac ed pea ( +2)-( +3)
Summe empe a u e
−
0.651
−
0.488 4.675** 4.417
[1.157] [1.213] [2.330] [2.660]
565
Jou nal o Economic G ow h (2024) 29:543–583
1 3
Table 6 The e ec s o he chole a in 1849 and 1854 on he numbe o mechanized ploughs and animal-powe ed h eshing machines pe day labo e and on he numbe and
wage o ag icul u al day labo e s in 1852 and 1862
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
OLS OLS OLS 2SLS 2SLS OLS OLS OLS 2SLS 2SLS
Mechanized ploughs pe day labo e Animal-powe ed h eshing machines pe day labo e
Sha e o chole a dea hs in popula ion 67.29** 55.09* 58.35* 323.6*** 369.9*** 18.90* 18.32** 18.58** 2.708 3.002
[28.36] [29.67] [29.65] [117.4] [135.1] [9.529] [8.501] [8.516] [7.922] [7.940]
Depa men - and yea ixed e ec s Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
De ia ion om summe ain all No Yes Yes Yes Yes No Yes Yes Yes Yes
Geog aphic con ols No Yes Yes Yes Yes No Yes Yes Yes Yes
GDP pe capi a No No Yes No Yes No No Yes No Yes
Wi hin R2 0.615 0.672 0.674 0.354 0.488 0.49
Mean dep a 2.80 2.81 2.82 2.83 2.84 0.1 0.1 0.1 0.1 0.1
Clus e s 85 85 85 85 85 85 85 85 85 85
Obse a ions 170 170 170 170 170 170 170 170 170 170
Fi s s age: he ins umen ed a iable is sha e o chole a dea hs in popula ion
Summe empe a u e
−
0.180***
−
0.179***
−
0.180***
−
0.179***
[0.0471] [0.0485] [0.0471] [0.0485]
1s s age F-s a 14.602 13.577 14.602 13.577
Reduced o m: he dependen a iable is
Mechanized ploughs pe day labo e Animal-powe ed h eshing machines pe day labo e
Summe empe a u e
−
58.29***
−
66.12***
−
0.488
−
0.537
[16.49] [18.28] [1.531] [1.549]
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
OLS OLS OLS 2SLS 2SLS OLS OLS OLS 2SLS 2SLS
Numbe o day labo e s A e age wage o day labo e s
Sha e o chole a dea hs in popula ion
−
15.56***
−
12.39**
−
11.32**
−
43.19***
−
38.86*** 0.0072 0.0051 0.0039 0.0353** 0.0304*
[5.449] [5.467] [5.452] [14.68] [14.63] [0.005] [0.006] [0.006] [0.018] [0.018]
Depa men - and yea ixed e ec s Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
566
Jou nal o Economic G ow h (2024) 29:543–583
1 3
This able p esen s OLS and IV eg essions ela ing he sha e o chole a dea hs in each depa men o he numbe o mechanized ploughs and animal-powe ed h eshing
machines pe day labo e . Geog aphic con ols o depa men s, which a e in e ac ed wi h yea - ixed e ec s, include hei land sui abili y, hei sha e o ca boni e ous a ea and
dummies o bo de and ma i ime depa men s. Cons an no epo ed. Robus s anda d e o s clus e ed a he depa men le el.
∗∗∗ p
<
0.01,∗∗ p
<
0.05,∗p
<
0.1
Table 6 (con inued)
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
OLS OLS OLS 2SLS 2SLS OLS OLS OLS 2SLS 2SLS
Numbe o day labo e s A e age wage o day labo e s
De ia ion om summe ain all No Yes Yes Yes Yes No Yes Yes Yes Yes
Geog aphic con ols No Yes Yes Yes Yes No Yes Yes Yes Yes
GDP pe capi a No No Yes No Yes No No Yes No Yes
Wi hin R2 0.924 0.934 0.936 0.464 0.576 0.593
Mean dep. a 26661.65 26661.65 26661.65 26661.65 26661.65 0.02 0.02 0.02 0.02 0.02
Clus e s 85 85 85 85 85 85 85 85 85 85
Obse a ions 170 170 170 170 170 170 170 170 170 170
Fi s s age: he ins umen ed a iable is sha e o chole a dea hs in popula ion
Summe empe a u e
−
0.180***
−
0.179***
−
0.180***
−
0.179***
[0.0471] [0.0485] [0.0471] [0.0485]
1s s age F-s a 14.602 13.577 14.602 13.577
Reduced o m: he dependen a iable is
Numbe o day labo e s A e age wage o day labo e s
Summe empe a u e 7.780*** 6.947***
−
0.00637**
−
0.00543*
[2.428] [2.596] [0.00315] [0.00324]
573
Jou nal o Economic G ow h (2024) 29:543–583
1 3
Table 8 The e ec s o he chole a in 1832, 1849 & 1854 on he signa u es o wedding licenses by spouses bo n one o 20 yea s a e each chole a pandemic
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)
OLS OLS OLS 2SLS 2SLS 2SLS OLS OLS OLS 2SLS 2SLS 2SLS
Signa u e o wedding license o
All indi iduals Indi iduals wo king in ag icul u e
Bo n 1 o 20 yea s a e each epidemic
Sha e o chole a dea hs
in popula ion
6.754*** 5.983*** 5.554*** 27.89*** 25.69*** 28.13*** 5.991* 4.962 3.454 39.37*** 42.48** 40.72**
[1.324] [1.379] [1.685] [4.940] [5.950] [6.729] [3.329] [3.462] [4.002] [14.40] [17.21] [16.12]
Male
−
0.00912
−
0.00906
−
0.00905
−
0.00932
−
0.00905
−
0.00915
−
0.0126
−
0.0129
−
0.0130
−
0.0114
−
0.0117
−
0.0117
[0.00694] [0.00693] [0.00693] [0.00698] [0.00696] [0.00697] [0.0145] [0.0145] [0.0145] [0.0144] [0.0143] [0.0143]
Depa men - and yea
ixed e ec s
Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
De ia ion om summe
ain all
No Yes Yes No Yes Yes No Yes Yes No Yes Yes
Geog aphic con ols No Yes Yes No Yes Yes No Yes Yes No Yes Yes
GDP pe capi a No No Yes No No Yes No No Yes No No Yes
Mean dep. a 0.797 0.797 0.797 0.797 0.797 0.797 0.731 0.731 0.731 0.731 0.731 0.731
Mo an I
−
0.0002
−
0.0002
−
0.0002
−
0.0002
−
0.0002
−
0.0002 0.002 0.002 0.002
−
0.001
−
0.001
−
0.001
Mo an I p- alue 0.246 0.246 0.246 0.246 0.246 0.246 1.000 1.000 1.000 0.242 0.243 0.243
Clus e s 3085 3085 3085 3085 3085 3085 1744 1744 1744 1744 1744 1744
Obse a ions 11,953 11,953 11,953 11,953 11,953 11,953 3,224 3,224 3,224 3,224 3,224 3,224
Fi s s age: he ins umen ed a iable is Sha e o Chole a Dea hs in Popula ion
Summe empe a u e
−
0.0825***
−
0.0708***
−
0.0628***
−
0.0510***
−
0.0443***
−
0.0472***
[0.00571] [0.00545] [0.00460] [0.00531] [0.00516] [0.00497]
1s s age F-s a 208.975 168.945 186.706 92.480 73.729 89.878
Reduced o m: he dependen a iable is
Signa u e o wedding license o

574
Jou nal o Economic G ow h (2024) 29:543–583
1 3
Table 8 (con inued)
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)
OLS OLS OLS 2SLS 2SLS 2SLS OLS OLS OLS 2SLS 2SLS 2SLS
Signa u e o wedding license o
All indi iduals Indi iduals wo king in ag icul u e
Bo n 1 o 20 yea s a e each epidemic
All indi iduals Indi iduals wo king in ag icul u e
Bo n 1 o 20 yea s a e each epidemic
Summe empe a u e
−
2.301***
−
1.818***
−
1.766***
−
2.009***
−
1.883**
−
1.920***
[0.400] [0.414] [0.416] [0.725] [0.745] [0.741]
This able p esen s OLS and IV eg essions ela ing he sha e o chole a dea hs o he abili y o b ides and g ooms bo n one o 20 yea s a e each ou b eak o sign hei wed-
ding license. Geog aphic con ols o depa men s, which a e in e ac ed wi h yea - ixed e ec s, include hei land sui abili y, hei sha e o ca boni e ous a ea and dummies o
bo de and ma i ime depa men s. Cons an no epo ed. Robus s anda d e o s clus e ed a he yea -depa men le el.
∗∗∗ p
<
0.01,∗∗ p
<
0.05,∗p
<
0.1
575
Jou nal o Economic G ow h (2024) 29:543–583
1 3
numbe o pa icipan s in cou ses o male adul s and app en ices as well as public spend-
ing on hese cou ses. Howe e , labo sca ci y nei he had a signi ican e ec on he numbe
o cou ses o emale adul s and app en ices no on he numbe o pa icipan s in hese
cou ses. A po en ial explana ion o his esul is ha ag icul u al mechaniza ion mainly
educed he demand o male labo , he eby leading men o immedia ely in es mo e in
human capi al and seek wo k in indus y whe e li e acy skills we e necessa y (e.g. F anck
& Galo , 2022).
Table I.5 shows ha he impac o he chole a pandemics in 1832, 1849 and 1854 on he
p ima y school a endance a e o boys and gi ls ou o he popula ion age 5–15 in 1837,
1851 and 1856 is posi i e bu no signi ican in all he speci ica ions. Mo eo e , Tables I.6
and I.7 assess he e ec o he chole a on public spending by he h ee ie s o he F ench
go e nmen (i.e., he cen al s a e, he depa men s and he communes) on p ima y school-
ing.29 Whe he we conside o al educa ion spending o educa ion spending pe inhabi an ,
he esul s sugges ha he pandemic had a nega i e impac on he depa men s’ spending
bu none on ha o he communes and o he cen al s a e, and o e all, no e ec on o al
public spending on p ima y schooling.
Those esul s should be pu in he gene al con ex o 19 h cen u y F ench educa ion.
The e we e o cou se p ima y schools in F ance be o e he i s chole a pandemic in 1832
(Mayeu , 2003). Mo eo e , a e he 28 June 1833 law (known as he “Loi Guizo ” a e
he hen Minis e o Educa ion), all communes had o hos a p ima y school in hei ju is-
dic ion. Tha school could be p i a ely o publicly unded, and un by a secula eache
paid by municipali y o by he local p ies (o nun). Thus, we may no ind any signi i-
can impac o he chole a on school spending because s a e in e en ion in educa ion was
al eady aking place in F ance a he ime, independen ly o he pandemics.
As such, in line wi h ou analysis ha iews labo and echnology as complemen a y
ac o s o p oduc ion in indus y and subs i u e in ag icul u e, labo sca ci y en ailed a ise
in human capi al in he a e ma h o he chole a pandemics. This inc ease did no s em
om he ising impo ance o s a e- unded p ima y schooling. Ins ead i esul ed om
p i a e in es men s made by pa en s in hei own human capi al as well as ha o hei
child en.
5.3 Al e na i e explana ions
O he han he inc ease in human capi al, ac o s such as mig a ion, u baniza ion, e il-
i y, age a ma iage, eligiosi y o local inancial in e media ion, could p o ide al e na i e
explana ions o ou main esul s.
In his sec ion, we b ie ly p esen he es s which we ca y ou o assess he impo ance
o such ac o s and p o ide mo e de ailed explana ions, including he da a sou ces, in he
Appendix. Reassu ingly, ou es s show ha hese ac o s we e no co ela ed wi h he
sp ead o chole a o wi h summe empe a u es in 1832, 1849 and 1854.
Mig a ion and u baniza ion. 19 h cen u y F ance was cha ac e ized by a high a e o
in e nal mig a ion (Daudin e al., 2019) bu no his o ical e idence connec s mig a ion and
u baniza ion o he chole a epidemics. I any hing, he po en ial e ec s o labo sca ci y on
mig a ion and u baniza ion a e no s aigh o wa d. Labo sca ci y en ails highe wages and
may a ac immig an s bu he adop ion o new echnology may lowe wages and hence
29 Because o da a limi a ions, Tables I.6 and I.7 only ocus on he impac o he 1854 chole a pandemic.
576
Jou nal o Economic G ow h (2024) 29:543–583
1 3
Table 9 Chole a in 1832, 1849 and 1854: numbe o pa icipan s in cou ses o male and emale adul s and app en ices
This able p esen s OLS and IV eg essions ela ing he sha e o chole a dea hs in each depa men o he numbe o pa icipan s in cou ses o male and emale adul s and
app en ices. Geog aphic con ols o depa men s, which a e in e ac ed wi h yea - ixed e ec s, include hei land sui abili y, hei sha e o ca boni e ous a ea and dummies o
bo de and ma i ime depa men s. Cons an no epo ed. Robus s anda d e o s clus e ed a he depa men le el.
∗∗∗ p
<
0.01,∗∗ p
<
0.05,∗p
<
0.1
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
OLS OLS OLS 2SLS 2SLS OLS OLS OLS 2SLS 2SLS
Numbe o pa icipan s in cou ses o
Male adul s and app en ices 1837–1850-1863 Female adul s and app en ices 1850–1863
Sha e o chole a dea hs in popula ion 26.49 19.20 21.42 126.0** 130.6** 29.80 34.15 32.28
−
3.720
−
19.63
[23.36] [24.45] [24.46] [57.31] [59.70] [27.62] [29.86] [28.32] [88.75] [89.00]
Depa men - and yea ixed e ec s Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
De ia ion om summe ain all No Yes Yes Yes Yes No Yes Yes Yes Yes
Geog aphic con ols No Yes Yes Yes Yes No Yes Yes Yes Yes
GDP pe capi a No No Yes No Yes No No Yes No Yes
Wi hin R2 0.404 0.430 0.438 0.023 0.057 0.059
Clus e s 85 85 85 85 85 85 85 85 85 85
Obse a ions 255 255 255 255 255 170 170 170 170 170
Fi s s age: he ins umen ed a iable is Sha e o Chole a Dea hs in Popula ion
Summe empe a u e
−
0.141***
−
0.140***
−
0.180***
−
0.179***
[0.0303] [0.0308] [0.0471] [0.0485]
1s s age F-s a 21.652 20.788 14.602 13.577
Reduced o m: he dependen a iable is
Numbe o pa icipan s in cou ses o
Male adul s and app en ices 1837–1850-1863 Female adul s and app en ices 1850–1863
Summe Tempe a u e
−
17.75**
−
18.32** 0.670 3.509
[7.848] [7.826] [16.31] [15.97]
577
Jou nal o Economic G ow h (2024) 29:543–583
1 3
Table 10 The e ec s o he chole a in 1832, 1849 and 1854 on Spending on cou ses o male adul s and app en ices and he numbe o cou ses o male and emale adul s and
app en ices in 1837, 1850 sand 1863
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15)
OLS OLS OLS 2SLS 2SLS OLS OLS OLS 2SLS 2SLS OLS OLS OLS 2SLS 2SLS
Spending on cou ses o Numbe o cou ses o Numbe o cou ses o
Male adul s and app en ices Male adul s and app en ices Female adul s and app en ices
1837–1850–1863 1837–1850–1863 1850–1863
Sha e o
chole a
dea hs in
popula-
ion
57.53*** 53.71** 54.65** 148.6 150.9 0.765
−
5.898
−
5.073 42.42 44.19 9.979 13.94 13.12 7.052 0.957
[20.98] [22.16] [21.69] [98.38] [98.31] [15.05] [15.24] [15.29] [40.81] [41.06] [11.81] [12.02] [11.27] [29.05] [30.16]
Depa -
men - and
yea ixed
e ec s
Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
De ia ion
om
summe
ain all
No Yes Yes Yes Yes No Yes Yes Yes Yes No Yes Yes Yes Yes
Geog aphic
con ols
No Yes Yes Yes Yes No Yes Yes Yes Yes No Yes Yes Yes Yes
GDP pe
capi a
No No Yes No Yes No No Yes No Yes No No Yes No Yes
Wi hin R2 0.607 0.615 0.616 0.416 0.438 0.441 0.009 0.054 0.058
Mean dep.
a
41.2 41.2 41.2 41.2 41.2 6669.5 6669.5 6669.5 6669.5 6669.5 2.04 2.04 2.04 2.04 2.04
Mo an I
−
0.008
−
0.008
−
0.008
−
0.008
−
0.008
−
0.007
−
0.007
−
0.007
−
0.007
−
0.007
−
0.012
−
0.012
−
0.012
−
0.012
−
0.0122
−
0.012
Mo an I
p- alue
0.239 0.236 0.235 0.237 0.237 0.259 0.257 0.257 0.253 0.255 0.207 0.210 0.209 0.210 0.209
578
Jou nal o Economic G ow h (2024) 29:543–583
1 3
Table 10 (con inued)
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15)
OLS OLS OLS 2SLS 2SLS OLS OLS OLS 2SLS 2SLS OLS OLS OLS 2SLS 2SLS
Spending on cou ses o Numbe o cou ses o Numbe o cou ses o
Male adul s and app en ices Male adul s and app en ices Female adul s and app en ices
1837–1850–1863 1837–1850–1863 1850–1863
Clus e s 85 85 85 85 85 85 85 85 85 85 85 85 85 85 85
Obse a-
ions
255 255 255 255 255 255 255 255 255 255 170 170 170 170 170
Fi s s age: he ins umen ed a iable is Sha e o Chole a Dea hs in Popula ion
Summe
empe a-
u e
−
0.141***
−
0.140***
−
0.141***
−
0.140***
−
0.180***
−
0.179***
[0.0303] [0.0308] [0.0303] [0.0308] [0.0471] [0.0485]
1s s age
F-s a
21.652 20.788 21.652 20.788 14.602 13.577
Reduced o m: he dependen a iable is
Spending on cou ses o Numbe o cou ses o Numbe o cou ses o
Male adul s and app en ices Male adul s and app en ices Female adul s and app en ices
1837–1850–1863 1837–1850–1863 1850–1863
Summe
empe a-
u e
−
20.94
−
21.16
−
5.976
−
6.198
−
1.270
−
0.171
[14.69] [14.60] [5.988] [5.957] [5.454] [5.553]
This able p esen s OLS and IV eg essions ela ing he sha e o chole a dea hs in each depa men o spending on cou ses o male adul s and app en ices and he num-
be o cou ses o male and emale adul s and app en ices. Geog aphic con ols o depa men s, which a e in e ac ed wi h yea - ixed e ec s, include hei land sui abil-
i y, hei sha e o ca boni e ous a ea and dummies o bo de and ma i ime depa men s. Cons an no epo ed. Robus s anda d e o s clus e ed a he depa men le el.
∗∗∗ p
<
0.01,∗∗ p
<
0.05,∗p
<
0.1

579
Jou nal o Economic G ow h (2024) 29:543–583
1 3
igge emig a ion (e.g., Fadinge & May , 2014). I may also be he case ha indi iduals
would lea e a eas hi by he chole a o escape dea h and would no come back. Tables J.1
and J.2 show ha mig a ion and u baniza ion we e no co ela ed wi h he sp ead o chol-
e a and canno he e o e d i e ou main esul s (i none heless bea s poin ing ou ha bo h
Tables do no ule ou ha mig a ion and u baniza ion could ha e played a ole in echnol-
ogy adop ion and inno a ion).
Religiosi y. To accoun o esea ch highligh ing he link be ween na u al disas-
e s (such as pandemics) and eligiosi y (e.g., Ben zen, 2019), we explo e whe he he
chole a ou b eaks could be co ela ed wi h changes in eligiosi y and po en ially wi h
a deepe cul u al shi ha could delay o accele a e echnology adop ion and inno a-
ion. Table J.3 shows ha he pandemics had a posi i e and signi ican bu quan i a i ely
small e ec on he sha e o semina ians in he popula ion, and no signi ican impac
on he sha e o eligious communi y membe s in he popula ion. O e all, hese esul s
sugges ha eligiosi y was no a ec ed by he chole a pandemics and canno he e o e
explain hei impac on echnology adop ion.
Fe ili y and nup iali y. Mo ali y shocks igge ed by pandemics could ha e an
impac on op imal e ili y beha io (Boucekkine e al., 2009; Siuda & Sunde, 2021).
Howe e , gi en ha he e ili y decline in F ance had begun in he la e 18 h cen u y
(e.g., Galo , 2011; Daudin e al., 2019; Blanc & Waczia g, 2020), i is no clea whe he
he sp ead o chole a could ha e an impac on e ili y a es and on he age a ma iage.
Tables J.4 and J.5 show ha indeed, he chole a epidemics had no sys ema ic signi ican
e ec on e ili y and nup iali y pa e ns, he eby sugges ing ha hose channels did no
a ec ou esul s.
Local inancial in e media ion. Because o he ela ionship be ween inancial in e -
media ion, economic g ow h and inno a ion (e.g., Go odnichenko & Schni ze , 2013;
Gennaioli e al., 2014), we examine whe he labo sca ci y os e ed echnological adop-
ion h ough he p esence o local banks. Table J.6 epo s he impac o he chole a
pandemics on he amoun o deposi s pe capi a in he sa ings banks o each depa men
a e aged o e he i e-yea pe iod which ollowed each pandemic. The e ec is insigni -
ican in all he speci ica ions. These esul s hus sugges ha local inancial de elopmen
was no co ela ed wi h he chole a ou b eaks and canno he e o e d i e ou esul s pe -
aining o echnology adop ion and inno a ion.
6 Conclusion
This pape examines he impac o labo sca ci y en ailed by he chole a epidemics in
1832, 1849 and 1854 in F ance on subsequen echnology adop ion and inno a ion. The
esul s show ha in he sho - un, labo sca ci y had a posi i e and signi ican impac
on echnology adop ion and inno a ion in ag icul u e while i had a nega i e impac on
echnology adop ion in indus y. This nega i e impac las ed longe in he ex ile indus-
y han in he mining sec o .
As labo sca ci y inc eased he expec ed e u ns o human capi al, indi iduals
in es ed mo e in hei own li e acy: his inc ease in he sha e o li e a e indi iduals in
he popula ion canceled ou he nega i e e ec o he popula ion loss on echnology
adop ion. Mo eo e , menial ag icul u al wo k became less appealing o li e a e wo ke s,
he eby leading o mo e echnology adop ion and inno a ion in ag icul u e.
580
Jou nal o Economic G ow h (2024) 29:543–583
1 3
The e a e h ee main implica ions o his s udy. Fi s , i sugges s ha in he 19 h cen-
u y, labo and echnology we e subs i u e ac o s o p oduc ion in ag icul u e bu com-
plemen a y in indus y. Second, i p o ides some suppo o he no ion ha ag icul u al
mechaniza ion in 19 h cen u y F ance was pa ly os e ed by labo sca ci y. Thi d, i
p o ides a mode a e iew on he e ec s o epea ed pandemics on economic g ow h.
No wi hs anding he human losses, he economic consequences o pandemics in socie-
ies ha escaped he Mal husian ap appea quan i a i ely limi ed in he sho - un and
disappea in he mid- o long- un.
Supplemen a y In o ma ion The online e sion con ains supplemen a y ma e ial a ailable a h ps:// doi.
o g/ 10. 1007/ s10887- 024- 09241-3.
Funding Open access unding p o ided by Heb ew Uni e si y o Je usalem.
Open Access This a icle is licensed unde a C ea i e Commons A ibu ion 4.0 In e na ional License,
which pe mi s use, sha ing, adap a ion, dis ibu ion and ep oduc ion in any medium o o ma , as long
as you gi e app op ia e c edi o he o iginal au ho (s) and he sou ce, p o ide a link o he C ea i e Com-
mons licence, and indica e i changes we e made. The images o o he hi d pa y ma e ial in his a icle
a e included in he a icle’s C ea i e Commons licence, unless indica ed o he wise in a c edi line o he
ma e ial. I ma e ial is no included in he a icle’s C ea i e Commons licence and you in ended use is no
pe mi ed by s a u o y egula ion o exceeds he pe mi ed use, you will need o ob ain pe mission di ec ly
om he copy igh holde . To iew a copy o his licence, isi h p://c ea i ecommons.o g/licenses/by/4.0/.
Re e ences
Ab ami zky, R., Dela ande, A., & Vasconcelos, L. (2011). Ma ying up: The ole o sex a io in asso a-
i e ma ching. Ame ican Economic Jou nal: Applied Economics, 3(4), 124–157.
Ab ami zky, R., Age , P., Bous an, L. P., Cohen, E., & Hansen, C. W. (2023). The e ec s o immig a ion
on he economy: Lessons om he 1920s bo de closu e. Ame ican Economic Jou nal: Applied
Economics, 15(1), 164–191.
Acemoglu, D. (2007). Equilib ium bias o echnology. Econome ica, 75(5), 1371–1409.
Acemoglu, D. (2010). When does labo sca ci y encou age inno a ion? Jou nal o Poli ical Economy,
118(6), 1037–1078.
Acemoglu, D., & Finkels ein, A. (2008). Inpu and echnology choices in egula ed indus ies: E idence
om he heal h ca e sec o . Jou nal o Poli ical Economy, 116(5), 837–880.
Acemoglu, D., de Feo, G., & Luca, G. D. D. (2020). Weak s a es: Causes and consequences o he Sicil-
ian Ma ia. Re iew o Economic S udies, 87, 537–581.
Acemoglu, D., & Res epo, P. (2022). Demog aphics and au oma ion. Re iew o Economic S udies,
89(1), 1–44.
Adda, J. (2016). Economic ac i i y and he sp ead o i al diseases: E idence om high equency da a.
Qua e ly Jou nal o Economics, 131(2), 891–941.
Aghion, P., & Howi , P. (1992). A model o g ow h h ough c ea i e des uc ion. Econome ica, 60(2),
323–51.
Agulhon, M., Dése , G., & Specklin, R. (2003). His oi e de la F ance Ru ale Vol. 3. Apogée de la ci ili-
sa ion paysanne. Edi ions du Seuil
Akcigi , U., G isby, J., & Nicholas, T. (2017). Immig a ion and he ise o Ame ican ingenui y. Ame i-
can Economic Re iew, 107(5), 327–331.
Albanesi, S., & Kim, J. (2021). E ec s o he COVID-19 ecession on he US labo ma ke : Occupa ion,
amily, and gende . Jou nal o Economic Pe spec i es, 35(3), 3–24.
Alesina, A., Ba is i, M., & Zei a, J. (2018). Technology and labo egula ions: Theo y and e idence.
Jou nal o Economic G ow h, 23(1), 41–78.
Amb us, A., Field, E., & Gonzalez, R. (2020). Loss in he ime o chole a: Long- un impac o a disease
epidemic on he u ban landscape. Ame ican Economic Re iew, 110(2), 475–525.
581
Jou nal o Economic G ow h (2024) 29:543–583
1 3
Ang is , J. D., & Pischke, J.-S. (2008). Mos ly ha mless econome ics: An empi icis ’s companion.
P ince on Uni e si y P ess.
Annuai e S a is ique De La F ance (1878–1939) , Imp ime ie Na ionale.
Ansa , S., Pela , C., Boelle, P. Y., Ca a , F., Flahaul , A., & Valle on, A. J. (2009). Mo ali y bu den
o he 1918–1919 in luenza pandemic in Eu ope. In luenza and O he Respi a o y Vi uses, 3(3),
99–106.
Ash a , Q., & Galo , O. (2011). Dynamics and s agna ion in he Mal husian epoch. Ame ican Economic
Re iew, 101(5), 2003–41.
A ack, Je emy, Ma go, Robe A., & Rhode, Paul W. (2019). “Au oma ion’’ o manu ac u ing in he la e
nine een h cen u y: The hand and machine labo s udy. Jou nal o Economic Pe spec i es, 33(2),
51–70.
Ba o, R.J., U súa., J.F. & Weng, J. (2020). The co ona i us and he g ea in luenza pandemic: Lessons
om he “Spanish lu” o he co ona i us’s po en ial e ec s on mo ali y and economic ac i i y,
Wo king Pape 26866, NBER.
Bazo , G. (2014). In e egional inequali ies, con e gence, and g ow h in F ance om 1840 o 1911. Annals
o Economics and S a is ics, 113–114, 309–345.
Belo i, F., Hughes., G. & Mo a i, A.P. (2013). XSMLE: S a a module o spa ial panel da a models es i-
ma ion, S a is ical So wa e Componen s S457610, Bos on College.
Ben zen, J. S. (2019). Ac s o God? Religiosi y and na u al disas e s ac oss subna ional wo ld dis ic s. Eco-
nomic Jou nal, 129(622), 2295–2321.
Blanc, G., & Waczia g, R. (2020). Change and pe sis ence in he age o mode niza ion: Sain -Ge main-
d’Anxu e, 1730–1895. Explo a ions in Economic His o y, 78, 101352.
Bodinie , B., & Teyssie , É. (2000). L’é énemen le plus impo an de la Ré olu ion: la en e des biens
na ionaux (1789–1867) en F ance e dans les e i oi es annexés. Socié é des é udes obespie is es,
Edi ions du CTHS.
Bonneuil, N. (1997). T ans o ma ion o he F ench demog aphic landscape, 1806–1906. Cla endon P ess.
Boucekkine, R., Desbo des, R., & La ze , H. (2009). How do epidemics induce beha io al changes? Jou nal
o he Economic G ow h, 14(3), 233–264.
Bou delais, P., & Raulo , J.-Y. (1987). Une peu bleue. His oi e du cholé a en F ance 1832–1854, Payo ,
Pa is.
Bou dieu, J., Kesz enbaum, L., & Pos el-Vinay, G. (2014). L’enquê e TRA, his oi e d’un ou il, ou il pou
l’his oi e. INED.
B occhi, A. (1834). Pompe amé icaine impo ée en F ance pa M. Fa co , ingénieu mécanicien. Re ue
comme ciale e indus ielle eu opéenne, 1(1), 15–19.
Came on, R., & Neal, L. (2015). A concise economic his o y o he wo ld om Paleoli hic imes o he
p esen (5 h ed.). Ox o d Uni e si y P ess.
Cap e ini, B., & Vo h, H.-J. (2020). Rage agains he machines: Labo -sa ing echnology and un es in
indus ializing England. Ame ican Economic Re iew: Insigh s, 2(3), 305–320.
Chak abo y, S., Papageo giou, C., & Sebas ián, F. P. (2010). Diseases, in ec ion dynamics and de elop-
men . Jou nal o Mone a y Economics, 57(7), 859–872.
Chanu , J.-M., He e , J., Mai esse, J., & Pos el-Vinay, G. (2000). L’indus ie ançaise au milieu du 19e
siècle. Les enquê es de la S a is ique Géné ale de la F ance. EHESS
Châ elain, A. (1977). Les mig an s empo ai es en F ance de 1800 à 1914. Publica ions de l’Uni e si é
de Lille.
Clemens, M. A., Lewis, E. G., & Pos el, H. M. (2018). Immig a ion es ic ions as ac i e labo ma -
ke policy: E idence om he Mexican B ace o exclusion. Ame ican Economic Re iew, 108(6),
1468–1487.
Cli , J. L., Zinkin, P., & Ma ell, A. (1986). A hospi al ou b eak o chole a in Mapu o, Mozambique.
T ansac ions o he Royal Socie y o T opical Medicine and Hygiene, 80, 473–476.
Colella, F., Lali e, R., Sakalli, S.O. & Ma hias, T. (2020). In e ence wi h a bi a y clus e ing, Discus-
sion Pape 12584, IZA.
Colwell, R. R. (2004). In ec ious disease and en i onmen : Chole a as pa adigm o wa e nbo ne dis-
ease. In e na ional Mic obiology, 7, 285–289.
C épin, A. (2009). His oi e de la consc ip ion. Gallima d, Folio His oi e.
Daudin, G., F anck, R., & Rapopo , H. (2019). Can in e nal mig a ion os e he con e gence in egional
e ili y a es? E idence om nine een h cen u y F ance. Economic Jou nal, 129(620), 1618–1692.
de Chaisema in, C., & D’Haul oeuille, X. (2020). Two-way ixed e ec s es ima o s wi h he e ogeneous
ea men e ec s. Ame ican Economic Re iew, 110(9), 2964–2996.
de la C oix, D., & Pe in, F. (2018). How a can economic incen i es explain he F ench e ili y and
educa ion ansi ion? Eu opean Economic Re iew, 108, 221–245.
582
Jou nal o Economic G ow h (2024) 29:543–583
1 3
Dechezlep ê e, A., Hémous, D., Olsen, M., & Ca lo, Z. (2019). Au oma ing labo : e idence om i m-
le el pa en da a, Discussion Pape 24421, CEPR.
Delapo e, J. (1986). Disease and ci iliza ion. The chole a in Pa is 1832. MIT P ess.
Dése , G. (1984). Machinisme e ag icul u e dans la F ance du XIXe siècle. His o ical Pape s, 19(1),
185–216.
Dup é, P. (1993). Les indus iels du Cal ados e la apeu . Cahie des Annales de No mandie, 25, 35–124.
Fadinge , H., & May , K. (2014). Skill-biased echnological change, unemploymen , and b ain d ain.
Jou nal o he Eu opean Economic Associa ion, 12(2), 397–431.
Fe nihough, A., & O’Rou ke Ke inH. (2014). Coal and he Eu opean indus ial e olu ion, Wo king
Pape 19802, NBER.
Fe nihough, A., & O’Rou ke, K. H. (2021). Coal and he Eu opean indus ial e olu ion. Economic
Jou nal, 131(635), 1135–1149.
F ance - Minis è e de l’ag icul u e. (1862). Documen s s a is iques e adminis a i s conce nan
l’épidémie de cholé a de 1854, compa ée aux p écéden es épidémies cholé iques qui on sé i en
F ance. Imp ime ie Impé iale.
F anck, R., & Johnson, N. D. (2016). Can public policies lowe eligiosi y? E idence om school choice
in F ance, 1878–1902. Economic His o y Re iew, 69(3), 915–944.
F anck, R., & Galo , O. (2022). Technology-skill complemen a i y in ea ly phases o indus ializa ion.
Economic Jou nal, 132(642), 618–643.
Galo , O. (2011). Uni ied g ow h heo y. P ince on Uni e si y P ess.
Gennaioli, N., Shlei e , A., & Vishny, R. (2014). Finance and he p ese a ion o weal h. Qua e ly
Jou nal o Economics, 129(3), 1221–1254.
Glass, R. I., & Black, R. E. (1992). The epidemiology o chole a. In D. Ba ua & W. B. G eenough
(Eds.), Chole a (pp. 129–154). Sp inge .
Go odnichenko, Y., & Schni ze , M. (2013). Financial cons ain s and inno a ion: Why poo coun ies
don’ ca ch up. Jou nal o he Eu opean Economic Associa ion, 11(5), 1115–1152.
Habakukk, H. J. (1962). Ame ican and B i ish echnology in he nine een h cen u y. Camb idge Uni e -
si y P ess.
Hanlon, W. W. (2015). Necessi y is he mo he o in en ion: Inpu supplies and di ec ed echnical
change. Econome ica, 83(1), 67–100.
Hassle , J., K usell, P., & Olo sson, C. (2021). Di ec ed echnical change as a esponse o na u al esou ce
sca ci y. Jou nal o Poli ical Economy, 129(11), 3039–3072.
Ho nbeck, R., & Naidu, S. (2014). When he le ee b eaks: Black mig a ion and economic de elopmen in
he Ame ican Sou h. Ame ican Economic Re iew, 104(3), 963–990.
Howi , P. (1999). S eady endogenous g ow h wi h popula ion and R &D inpu s g owing. Jou nal o Poli i-
cal Economy, 107(4), 715–730.
Huang, D., Yan, P., Xiao, X., Zhu, J., Tang, X., Huang, A., & Cheng, J. (2019). The i-pole ela ion among
daily mean empe a u e, a mosphe ic mois u e and p ecipi a ion in ensi y o e China. Global and
Plane a y Change, 179, 1–9.
Jedwab, R., Johnson, N.D., & Koyama, M. (2019). Pandemics, places, and popula ions: E idence om he
Black Dea h, Wo king Pape 7524, CESi o.
Jo dà, Ò., Singh, S. R., & Taylo , A. M. (2020). Longe - un economic consequences o pandemics. Co id
Economics Ve ed and Real-Time Pape s, 1, 1–15.
Juhász, R. (2018). Tempo a y p o ec ion and echnology adop ion: E idence om he Napoleonic blockade.
Ame ican Economic Re iew, 108, 3339–3376.
Ka z, L. F., & Ma go, R. A. (2014). Human capi al in his o y: The Ame ican eco d. In L. P. Bous an, C.
F ydman, & R. A. Ma go (Eds.), Technical change and he ela i e demand o skilled labo : The
Uni ed S a es in his o ical pe spec i e (pp. 15–57). Uni e si y o Chicago P ess.
Kelly, M. (2019). The s anda d e o s o pe sis ence, Discussion Pape 13783, CEPR.
Kelly, M., Moky , J., & ÓG áda, C. (2014). P ecocious Albion: A new in e p e a ion o he B i ish indus-
ial e olu ion. Annual Re iew o Economics, 6(1), 363–389.
Knoblach, Michael, & S öckl, Fabian. (2019). Wha de e mines he elas ici y o subs i u ion be ween capi al
and labo ?. Discussion pape , DIW: A li e a u e e iew.
Koch, R. (1884). Sechs e Be ich de Lei e s de Deu schen wissenscha lichen Commission zu
E o schung de Chole a. Deu sche Medizinische Wochensc i , 10(12), 191–192.
K eme , M. (1993). Popula ion g ow h and echnological change: One million b.c. o 1990. The Qua e ly
Jou nal o Economics, 108(3), 681–716.
La o une, J., Tessada, J., & González-Velosad, C. (2015). Mo e hands, mo e powe ? Es ima ing he impac
o immig a ion on ou pu and echnology choices using ea ly 20 h cen u y us ag icul u e. Jou nal o
In e na ional Economics, 97(2), 339–358.