Kayongo, Pa ick Sunday
A icle — Published Ve sion
Re-e alua ing he Re u ns o Labou in Mic oen e p ises:
A S a is ical Replica ion and C i ical Re iew o de Mel e al.
(2019)
The Indian Jou nal o Labou Economics
P o ided in Coope a ion wi h:
Sp inge Na u e
Sugges ed Ci a ion: Kayongo, Pa ick Sunday (2025) : Re-e alua ing he Re u ns o Labou in
Mic oen e p ises: A S a is ical Replica ion and C i ical Re iew o de Mel e al. (2019), The Indian
Jou nal o Labou Economics, ISSN 0019-5308, Sp inge India, New Delhi, Vol. 68, Iss. 3, pp.
1113-1132,
h ps://doi.o g/10.1007/s41027-025-00583-z
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/330914
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/4.0/
Vol.:(0123456789)
The Indian Jou nal o Labou Economics (2025) 68:1113–1132
h ps://doi.o g/10.1007/s41027-025-00583-z
ISLE
RESEARCH NOTE
Re‑e alua ing heRe u ns oLabou inMic oen e p ises:
AS a is ical Replica ion andC i ical Re iew o de Mel e al.
(2019)
Pa ickSundayKayongo1
Recei ed: 26 Feb ua y 2024 / Accep ed: 14 Augus 2025 / Published online: 8 Oc obe 2025
© The Au ho (s) 2025
Abs ac
This pape eplica es de Mel e al. (Am Econ J Appl Econ 11(1):202–235, 2019), a
ield expe imen in S i Lanka e alua ing he e ec s o wage subsidies on mic oen-
e p ise employmen . While he o iginal s udy ound no las ing impac on p o i s o
i m scale—in e p e ing his as e idence o diminishing labou e u ns— his epli-
ca ion con i ms he empi ical pa e ns bu challenges ha conclusion. I a gues ha
sho -li ed gains likely e lec deepe s uc u al and beha iou al cons ain s, such as
limi ed manage ial capaci y, in o mal labou dynamics, and ins i u ional unce ain y.
Beyond e i ica ion, he pape ex ends he o iginal he e ogenei y analysis, e eal-
ing ha i ms wi h p e-exis ing employees we e signi ican ly mo e esponsi e o he
subsidy, while solo i ms we e no —shi ing he ocus om sec o al e ec s o base-
line employmen s uc u e. This e inemen o e s new insigh s o a ge ing wage
subsidies. The eplica ion also aises e hical conce ns abou empo a y hi ing, which
may e ode wo ke mo ale and i m s abili y. By ein e p e ing null esul s as symp-
oms o policy design gaps a he han i m-le el ine iciency, he s udy unde sco es
he need o bundled, ecosys em-le el suppo . I d aws conc e e lessons o India’s
mic oen e p ise schemes, a guing ha wage subsidies alone a e insu icien o du a-
ble ans o ma ion. Replica ion, in his ligh , becomes a ool o heo y building and
policy diagnosis—no jus empi ical sc u iny.
Keywo ds Replica ion· Ma ginal e u ns o labou · Mic oen e p ises· Wage
subsidies· S i Lanka· India
JEL Classi ica ion C93· J22· J38· L26· J38· O15· O17
* Pa ick Sunday Kayongo
pa ick.kayongo@s ud.uni-due.de; kay[email p o ec ed]
1 Chai o Public Finance, Facul y o Business andEconomics, Uni e si y o Duisbu g-Essen,
Essen, Ge many
1114
The Indian Jou nal o Labou Economics (2025) 68:1113–1132
ISLE
1 In oduc ion
Mic oen e p ises a e o en ega ded as engines o economic g ow h, pa icula ly
in de eloping economies whe e hey domina e he in o mal sec o and p o ide
li elihoods o millions (Muske & Woods 2016). Howe e , hei capaci y o
scale—especially h ough hi ing addi ional labou — emains a subjec o ongoing
deba e in bo h economic heo y and policy. A cen al ques ion in his li e a u e
is whe he hi ing an addi ional wo ke yields posi i e and sus ained e u ns in
mic oen e p ise se ings, o whe he such i ms ope a e unde condi ions o apid
diminishing ma ginal e u ns o labou (de Mel e al. 2019).
Neoclassical models ypically p edic ha mic oen e p ises a e op imally small,
exhibi ing s eep declines in ma ginal p oduc i i y when labou inpu s inc ease
beyond a minimal scale (Lucas 1978). This heo y has guided much o he skep i-
cism a ound subsidised employmen in e en ions a ge ed a small i ms. Howe e ,
ecen empi ical esea ch has challenged his assump ion, a guing ha unde ce ain
condi ions—such as imp o ed in o ma ion lows, educed ic ions, o suppo i e
ins i u ional con ex s—addi ional labou may lead o meaning ul and sus ained p o-
duc i i y gains (B yan e al. 2014; Beaman e al. 2014; Ha dy & McCasland, 2023).
These compe ing pe spec i es e eal an un esol ed heo e ical and empi ical ension
ega ding he na u e o labou cons ain s in mic oen e p ises.
Replica ion s udies ha e eme ged as a i al ool in na iga ing his ension. As
emphasised by Clemens (2017) and McKenzie (2012), eplica ions play a c i ical
ole no only in e i ying empi ical esul s bu also in challenging hei heo e ical
in e p e a ion and policy ele ance. Pa icula ly in de elopmen economics—whe e
con ex sensi i i y and measu emen limi a ions a e common— eplica ion enhances
he c edibili y and ex e nal alidi y o esea ch indings. As Du lo e al. (2007)
a gue, he capaci y o ep oduce esul s using anspa en and s anda dised me hods
is ounda ional o building cumula i e knowledge in expe imen al economics.
To con ibu e o his deba e, De Mel e al. (2019) conduc ed a landma k ield
expe imen in S i Lanka, in which wage subsidies we e o e ed o a andomised
g oup o male-owned mic oen e p ises. The aim was o es whe he empo a y
educ ions in hi ing cos s would induce i ms o expand employmen —and, c u-
cially, whe he hese e ec s would pe sis beyond he subsidy pe iod. The s udy
ound ha while ea ed i ms hi ed mo e wo ke s and we e mo e likely o su i e
du ing he in e en ion, he e we e no las ing imp o emen s in employmen , p o i s,
o business scale a e he subsidies ended. These indings appea ed o ein o ce he
canonical iew ha mic oen e p ises a e no labou cons ained in he long un.
In his pape , I unde ake a s a is ical eplica ion o de Mel e al. (2019) using
hei publicly a ailable eplica ion package. My p ima y objec i e is o e i y he
ep oducibili y o he s udy’s indings using upda ed analy ical ools. Howe e ,
I go beyond me e duplica ion. I c i ically engage wi h he design, con ex , and
in e p e a ion o he expe imen , add essing b oade me hodological and heo e i-
cal conce ns aised in he li e a u e on mic oen e p ise de elopmen .
My eplica ion con i ms he main empi ical esul s o he o iginal s udy, wi h
only mino nume ical di e ences. Ye , I a gue ha he in e p e a ion o hese
1115
The Indian Jou nal o Labou Economics (2025) 68:1113–1132
ISLE
indings equi es mo e nuance. In pa icula , I highligh sho - e m employmen
gains as po en ially meaning ul in con ex s whe e labou ma ke sca ing and i m
in o mali y a e p e alen . I also discuss he limi s o andomised in e en ions in
cap u ing complex cons ain s acing mic oen e p ises, such as beha iou al ine -
ia, helack o manage ial capi al, and ins i u ional unce ain y.
This pape is guided by he ollowing hypo hesis:
Tempo a y wage subsidies can induce sho - e m employmen and su i al
gains in mic oen e p ises, bu he absence o long- e m p o i abili y imp o e-
men s may be a ibu able o deepe s uc u al and beha io al cons ain s—
no necessa ily o diminishing ma ginal e u ns alone.
The es o he pape is o ganised as ollows: Sec ion 2 de ails he eplica ion
s a egy and empi ical esul s. Sec ion 3 o e s a c i ical assessmen o he o igi-
nal expe imen ’s me hodology and heo e ical aming. Sec ion4 discusses policy
implica ions, pa icula ly in he con ex o Indian and Global Sou h mic oen e p ise
policy. Sec ion5 concludes wi h ecommenda ions o u u e esea ch and expe i-
men al design.
2 Replica ion andResul s
2.1 Replica ion S a egy
Gi en he ongoing deba e a ound labou cons ain s in mic oen e p ises, I eplica e
he expe imen al analysis o de Mel e al. (2019) using he o iginal eplica ion pack-
age hey publicly p o ided.1 The da ase and S a a.do iles include all a iables and
code used in he published pape , which enables a s ep-by-s ep e i ica ion o hei
es ima ion p ocedu es.2
The expe imen was conduc ed in S i Lanka in 2009 and in ol ed 1533 male-
owned mic oen e p ises. A andomised subse o hese i ms ecei ed a wage sub-
sidy in ended o o se he cos o hi ing a wo ke o up o 8mon hs. The p ima y
ou come a iables included i m su i al, employmen , p o i s, and capi al s ock.
Da a we e collec ed o e 12 su ey ounds, wi h he i s wo ounds se ing as
baseline.
I begin by inspec ing and e- unning he o iginal code o ep oduce he esul s
exac ly as hey appea ed in he published pape .3 A e con i ming consis ency, I
e-es ima e he key speci ica ions using S a a 18, ensu ing compa ibili y wi h cu -
en s a is ical packages and applying clus e ing and mul iple hypo hesis co ec ions
1 The au ho s o de Mel e al. (2019) p o ide a publicly accessible eplica ion package o whoe e is
in e es ed in eplica ing he s udy a : h ps:// www. openi cps . o g/ openi cps / p oje c / 113732/ e si on/ V1/
iew.
2 Fo a de ailed s ep by s ep desc ip ion o he s udy’s es ima ion p ocedu es, see Sec ion II o he
esea ch pape : h ps:// doi. o g/ 10. 1257/ app. 20170 497; pp. 208–215.
3 Fo comple e and mo e comp ehensi e esul s, see Sec ion IV o he o iginal esea ch pape : h ps://
doi. o g/ 10. 1257/ app. 20170 497; pp. 216–227.
1116
The Indian Jou nal o Labou Economics (2025) 68:1113–1132
ISLE
whe e ele an . The eg ession models ollow he o iginal s udy’s e en - ime ame-
wo k, es ima ing ea men e ec s o e dis inc pe iods: p e- ea men , du ing- ea -
men , and pos - ea men yea s 1 h ough 4.
The baseline es ima ion model is speci ied as:
whe e
Y is he ou come o i m i in pe iod = 3, …, 12. T ea is a dummy a iable indi-
ca ing whe he he i m ecei ed he wage subsidy ea men . P e deno es he p e-
ea men pe iod, speci ically pos -baseline ound 3. Du ing e e s o su ey ounds
4 and 5, when he wage subsidy was in e ec . Yea 1, Yea 2, and Yea 3 o 4 co -
espond o su ey ounds conduc ed app oxima ely 1yea , 2yea s, and 3–4yea s
pos -in e en ion, espec i ely. 1( = s) ep esen s a se o su ey ound ime dum-
mies, while X includes con ols o he andomisa ion s a a and o he baseline
a iables used in he andomisa ion p ocess. The e o e m, ε < sub > i, < /sub > ,
is clus e ed a he i m le el. Coe icien s β₂ h ough β₅ a e he main pa ame e s o
in e es , cap u ing he ajec o y o ea men e ec s o e ime. To accoun o mul-
iple hypo hesis es ing ac oss pe iods, he equali y β₂ = β₃ = β₄ = β₅ is es ed, as well
as he join null hypo hesis β₂ = β₃ = β₄ = β₅ = 0, o assess he s abili y o ea men
e ec s. Finally, β₁ se es as a placebo es , as i co esponds o he p e- ea men
pe iod. As in he o iginal s udy, I es he join signi icance o ea men coe icien s
(β₂ h oughβ₅), and assess s abili y o e ime. I also in e p e β1 as a placebo check,
since i cap u es any spu ious ea men e ec s be o e in e en ion. The ou pu
ables p esen ed in Sec s.2.2 h ough 2.4 a e d awn di ec ly om he S a a ou pu .
Whe e necessa y, I ha e e o ma ed hese o cla i y.
2.2 Fi m Su i al andEmploymen E ec s
The eplica ion con i ms ha wage subsidies inc eased i m su i al and employ-
men du ing he in e en ion pe iod, bu hese e ec s weakened o e ime. T ea ed
i ms we e mo e likely o emain ope a ional up o 1-yea pos - ea men , bu he
ad an age dissipa ed by yea s 3–4. Employmen e ec s peaked du ing he subsidy
phase, wi h modes spillo e s in o yea one be o e ading he ea e (Table1).
2.2.1 Employmen E ec s
The numbe o paid wo ke s inc eased signi ican ly du ing he subsidy pe iod.
Fi ms we e mo e likely o hi e a leas one paid wo ke du ing ounds 4–5, when
he subsidy was in e ec . Howe e , hese employmen gains we e no sus ained
Yi, =𝛼+𝛽
1
T ea
i
×P e
+𝛽2T ea i×Du ing +∑12
s=4𝛿s1( =s)+𝜃�Xi+𝜀i,
+𝛽3T ea i×Yea 1
+𝛽4T ea i×Yea 2
+𝛽
5T ea i
×
Yea
3
o
4
1117
The Indian Jou nal o Labou Economics (2025) 68:1113–1132
ISLE
Table 1 Rep oduced ea men e ec s on i m su i al
Sel -employed1 and 2 e e o sel -employed–assuming ha i ms which close and a e ne e obse ed again ha e s ayed closed; and sel -employed, assuming ha all
a i o s a e closed, espec i ely
Sou ce: Au ho ’s compu a ions based on de Mel e al. (2019)’s eplica ion package
Pa ame e Panel A: Sel -employed in su ey ound Panel B: Sel -
employed 1
Panel C: Sel -employed 2 Panel D: Su i al
as pe Fi m Dea h
Pape
Panel E: Employed
a all
P e- ea men − 0.006 − 0.006 − 0.008 0.012 − 0.007
(0.023) (0.023) (0.023) (0.023) (0.023)
Du ing- ea men − 0.009 − 0.010 − 0.012 − 0.006 − 0.006
(0.018) (0.018) (0.018) (0.021) (0.013)
Pos - ea men _
yea 1
0.058*** 0.056*** 0.055** 0.066** 0.034**
(0.021) (0.021) (0.022) (0.026) (0.015)
Pos - ea men
_yea 2
0.082*** 0.075*** 0.064** 0.079** 0.003
(0.025) (0.026) (0.028) (0.032) (0.019)
Pos - ea men
_yea 3
0.054** 0.054* 0.055* 0.074** 0.013
(0.027) (0.029) (0.030) (0.034) (0.021)
Sample size 5055 5092 5185 5057 5185
P- alue 1 0.001 0.003 0.009 0.004 0.022
P- alue 2 0.002 0.006 0.018 0.010 0.046
Con ol mean P e 0.927 0.927 0.927 0.920 0.927
Con ol mean du ing 0.958 0.958 0.958 0.939 0.971
Con ol mean Yea 1 0.885 0.885 0.875 0.839 0.927
Con ol mean Yea 2 0.850 0.847 0.825 0.773 0.933
Con ol mean Yea 3 0.831 0.815 0.788 0.729 0.906
S anda d e o s in
pa en heses
= *p < 0.10 **p < 0.05 ***p < 0.01”
1118
The Indian Jou nal o Labou Economics (2025) 68:1113–1132
ISLE
beyond he i s pos - ea men yea . Table 2 epo s he ea men e ec s on
employmen ou comes, including bo h he numbe o paid wo ke s (Panel A) and
he p obabili y o hi ing any paid wo ke (Panel B). As in he o iginal s udy, hese
e ec s peak du ing he subsidy pe iod and ade he ea e .
These esul s align closely wi h hose o de Mel e al. (2019), alida ing hei
co e inding ha sho - un employmen gains a e no accompanied by long- un
shi s in i m size o s uc u e. This dynamic is u he illus a ed in Fig.1, which
plo s he a e age numbe o paid wo ke s o e ime o ea ed and con ol i ms.
A sha p di e gence is isible du ing he subsidy pe iod (Rounds 4–5), ollowed
by con e gence in la e ounds, isually ein o cing he sho -li ed na u e o he
ea men e ec .
2.3 P o i abili y andRe u ns oLabou
I eplica e he p o i impac models using bo h uncondi ional and condi ional p o i
measu es. The o iginal s udy epo s ha he subsidy had no signi ican e ec on
p o i s du ing o a e he in e en ion, despi e he empo a y inc ease in employ-
men . My es ima es co obo a e his esul . I also es ima e he e u n o hi ing an
addi ional wo ke by eg essing p o i s on he numbe o paid wo ke s, con olling
o ime e ec s and i m-le el co a ia es. Wi hou i m ixed e ec s, he ma ginal
e u n o an ex a wo ke appea s s a is ically signi ican , wi h ea ed i ms ea ning
app oxima ely 6214 LKR mo e pe week. Howe e , when ixed e ec s a e included,
his alls o a ound 4902 LKR, below he subsidy alue (4000 LKR), implying lim-
i ed cos -e ec i eness (Table3).
Howe e , as poin ed ou by de Mel e al. (2019), his es ima e is no ee om
unobse ed i m cha ac e is ics o example i m-le el own p oduc i i y, which may
join ly in luence p o i s. Fo his eason, ime-in a ian i m beha iou is con olled
by adding i m-le el ixed e ec s o Eq.(2). This educes he ma ginal labou e u ns
o 4902.61 LKR (14% weekly p o i inc emen ) (Table4).
2.4 Capi al S ock andCon ounding E ec s
Finally, I eplica e he capi al s ock eg essions o e i y whe he capi al-labou
in e ac ions con ound he obse ed ea men e ec s. Like he o iginal s udy, I ind
no s a is ically signi ican change in capi al accumula ion a ibu able o he wage
subsidy. This suppo s he conclusion ha he obse ed employmen e ec s a e no
mechanically d i en by concu en inc eases in i m capi al (Table5).
2.4.1 Summa y o Replica ion Resul s
O e all, he eplica ion exe cise con i ms he obus ness o he o iginal s udy’s empi i-
cal indings. The sho - e m gains in employmen and su i al a e eal bu empo a y.
The e a e no sus ained imp o emen s in p o i abili y, and ma ginal e u ns o labou a e
modes a bes . The in e p e a ion o hese esul s—and whe he hey imply inhe en
1119
The Indian Jou nal o Labou Economics (2025) 68:1113–1132
ISLE
ine iciency in subsidising mic oen e p ise labou —will be discussed c i ically in he
nex sec ion.
2.5 Compa a i e Summa y wi hde Mel e al. (2019)
To syn hesise he eplica ion esul s p esen ed in Sec s.2.2 h ough 2.4, Table6 p o-
ides a side-by-side compa ison o key es ima es om he o iginal s udy and his
eplica ion. The Table summa ises ea men e ec s ac oss employmen , su i al,
p o i abili y, and e u ns o labou . The eplica ion aligns closely wi h he o iginal ind-
ings in bo h di ec ion and s a is ical signi icance. Mino di e ences in e ec size a e
obse ed—likely due o ounding, upda ed s a is ical so wa e, o ea men o missing
alues—bu none al e he subs an i e in e p e a ion.
No ably, he eplica ion ea i ms he empo a y na u e o employmen gains and
he absence o long- e m imp o emen s in p o i abili y o capi al accumula ion. I also
Table 2 Rep oduced ea men e ec s on i m employmen
Sou ce:Au ho ’s compu a ions based on de Mel e al. (2019)’s eplica ion package
Pa ame e Panel A Numbe o paid
wo ke s
Panel B Any paid wo ke
P e- ea men − 0.073 − 0.012
(0.081) (0.035)
Du ing- ea men 0.194** 0.138***
(0.075) (0.035)
Pos - ea men _yea 1 0.126 0.111***
(0.077) (0.034)
Pos - ea men _yea 2 0.051 0.028
(0.079) (0.035)
Pos - ea men _yea 3 − 0.024 − 0.005
(0.083) (0.032)
Base paid wo ke s 0.000 0.000
(.) (.)
Base any paid 0.251***
(0.087)
Sample size 4879 4879
P- alue 1 0.057 0.000
P- alue 2 0.038 0.000
Con ol mean P e 0.48 0.25
Con ol mean du ing 0.48 0.27
Con ol mean Yea 1 0.45 0.24
Con ol mean Yea 2 0.50 0.28
Con ol mean Yea 3 0.60 0.29
S anda d e o s in pa en heses
= ”*p < 0.10 **p < 0.05 ***p < 0.01”
1120
The Indian Jou nal o Labou Economics (2025) 68:1113–1132
ISLE
ep oduces he ma ginal labou e u n es ima es wi h ema kable p ecision. This con-
sis ency s eng hens he c edibili y o he o iginal s udy and con i ms he eplicabili y
o i s empi ical claims.
2.6 Subg oup He e ogenei y inEmploymen E ec s
de Mel e al. (2019) explo ed he e ogenei y in he wage subsidy’s e ec s by in e ac -
ing ea men s a us wi h a ious baseline cha ac e is ics, including i m age, p io
hi ing his o y, managemen p ac ices, capi al s ock, household weal h, and sec o
(Table 7). While hei indings sugges ed ha i ms wi h be e baseline business
p ac ices and hose in manu ac u ing we e somewha mo e likely o hi e, mos sub-
g oup di e ences we e s a is ically insigni ican , and ea men e ec s dissipa ed
o e ime ac oss all ca ego ies.
This eplica ion ex ends and pa ially e ines ha analysis by examining he e o-
genei y in employmen esponses along wo key dimensions: i m sec o (manu ac-
u ing s. non-manu ac u ing) and ini ial i m size, de ined by whe he he i m had
any paid employees a baseline. These c i e ia align closely wi h hose used in he
o iginal s udy, hough ou app oach simpli ies he analysis by ocusing on bina y
subg oup spli s and es ima es he a e age ea men e ec on he p obabili y o hi -
ing any paid wo ke o e he panel.
The esul s, p esen ed in Table7, ea i m he o iginal s udy’s key indings while
o e ing new p ecision:
Fig. 1 G aph o employmen ends o e ime, ea men s. con ol. Sou ce: Au ho ’s compu a ions
based on de Mel e al. (2019)’s eplica ion package
1127
The Indian Jou nal o Labou Economics (2025) 68:1113–1132
ISLE
con inue beyond he subsidy pe iod can os e job insecu i y and educe wo ke
mo ale. Economic heo y and empi ical e idence sugges ha pe cei ed employ-
men ins abili y may lead wo ke s o unde in es in i m-speci ic human capi al
and disengage om asks ha deli e long- un p oduc i i y gains (Becke 1962;
Kahn 2007). These beha iou s e lec a ional esponses o insecu e employmen
en i onmen s and a e u he co obo a ed by indings in labou economics ha
link job insecu i y o declines in wo ke e o , o ganisa ional commi men , and
p oduc i i y (G eenhalgh & Rosenbla 1984; Böcke man e al. 2012).
While he o iginal s udy does no speci y whe he wo ke s we e in o med
abou he empo a y na u e o hei employmen , e hical s anda ds o ield expe -
imen s—such as hose ou lined by he Ame ican Economic Associa ion (AEA,
2019)—sugges ha ull disclosu e should ha e been p o ided. I his was indeed
he case, he absence o sus ained p o i gains could pa ially e lec an icipa o y
educ ions in wo ke engagemen s emming om an awa eness o hei sho job
enu e. These po en ial beha iou al esponses highligh he impo ance o inco -
po a ing e hical design p inciples and managing wo ke expec a ions in expe i-
men al labou ma ke in e en ions.
4 Policy Implica ions andRele ance oIndia
The indings o his eplica ion s udy—while oo ed in S i Lanka’s policy con-
ex —hold signi ican implica ions o India, whe e mic o, small, and medium
en e p ises (MSMEs) cons i u e o e 90% o all businesses and con ibu e
app oxima ely 30% o GDP and 45% o expo s (Minis y o MSME, 2023). As
in S i Lanka, India’s mic oen e p ises ope a e unde subs an ial cons ain s: lim-
i ed access o o mal c edi , low manage ial capaci y, and in o mal labou ma -
ke s (Maheshka & Soni 2022). These cha ac e is ics complica e he e alua ion
o employmen -based in e en ions such as wage subsidies o hi ing incen i es.
4.1 Implica ions o Indian MSME Policy
The Indian go e nmen has deployed a ange o employmen -linked schemes o
encou age mic oen e p ise g ow h (Di ec o a e Gene al o Employmen , 2025),
including:
• P ime Minis e ’s Employmen Gene a ion P og am (PMEGP): O e s inancial
suppo o new and exis ing mic oen e p ises o c ea e employmen oppo uni-
ies.
• MUDRA scheme: P o ides mic oloans o non-co po a e, non- a m small busi-
nesses, many o whom use he unds o hi ing in o mal labou .
• Aa mani bha Bha a Rozga Yojana (ABRY): O e s EPF subsidy suppo o incen-
i ise o mal hi ing in MSMEs pos -COVID.
1128
The Indian Jou nal o Labou Economics (2025) 68:1113–1132
ISLE
These schemes sha e a common logic wi h he in e en ion s udied by de Mel e al.
(2019): he idea ha educing labou cos s will induce small i ms o hi e mo e wo k-
e s and, ideally, g ow. Howe e , as my eplica ion con i ms, such in e en ions end
o p oduce sho - e m gains in employmen and i m su i al, wi hou co esponding
long- e m imp o emen s in p o i abili y o i m scale. This inding should no be in e -
p e ed, as de Mel e al. sugges , as con i ma ion ha mic oen e p ises a e no labou
cons ained. Ins ead, i may e lec he s uc u al and beha iou al limi a ions o i ms
ha p e en hem om p oduc i ely abso bing labou , especially once subsidies end.
Indian empi ical s udies poin in his di ec ion.
Fo ins ance, Bane jee and Du lo (2011) show ha many Indian mic oen e p ises
unde in es in bo h labou and capi al due o in o ma ional asymme ies, c edi con-
s ain s, and low en ep eneu ial capaci y— a he han because o apidly diminishing
e u ns. Simila ly, Pa wa dhan and Tascio i (2022) ind ha wage subsidy schemes
unde he Maha ma Gandhi Na ional Ru al Employmen Gua an ee Ac (MGNREGA)
had sho -li ed e ec s on u al labou alloca ion, wi h li le e idence o las ing en e -
p ise c ea ion. In a closely ela ed domain, Field e al. (2013) showed ha business
aining and pee men o ing among emale en ep eneu s in u ban India can subs an-
ially imp o e business ou comes—while access o c edi alone has li le e ec . These
indings sugges ha wage subsidies o inancial suppo schemes a e mo e likely o
succeed when complemen ed wi h manage ial o o ganisa ional suppo ha enables
i ms o p oduc i ely deploy esou ces o e ime.
My eplica ion hus ein o ces a c i ical insigh o Indian MSME policy: labou
subsidies, when implemen ed in isola ion, may empo a ily alle ia e hi ing ic ions
bu a e unlikely o yield s uc u al ans o ma ion. The obse ed sho - e m gains in
employmen and i m su i al a e no i ial—bu wi hou complemen a y suppo in
a eas like manage ial aining, access o wo king capi al, and o malisa ion incen i es,
such gains end o dissipa e. This calls o a policy shi away om s andalone hi ing
incen i es owa d bundled, con ex -awa e in e en ions ailo ed o he pe sis en con-
s ain s aced by India’s mic oen e p ises. In a landscape whe e public p og ammes like
PMEGP, MUDRA, and ABRY con inue o p io i ise employmen gene a ion, my ind-
ings unde sco e he need o mo e igo ous e alua ion amewo ks—ones ha assess
no only immedia e job c ea ion bu also he condi ions equi ed o du able en e p ise
g ow h.
4.2 Beyond Subsidies: Towa ds In eg a ed andCon ex ‑Sensi i e In e en ions
The e idence om my eplica ion sugges s ha empo a y wage suppo alone is insu -
icien o p oduce las ing imp o emen s in i m p o i abili y o scale—especially in
en i onmen s ma ked by in o mali y, low manage ial capaci y, and weak demand link-
ages. These insigh s echo a g owing body o wo k in India ha ques ions he s andalone
e icacy o c edi and employmen subsidies o mic oen e p ise de elopmen .
P og ammes such as PMEGP and MUDRA ha e achie ed subs an ial ou each
bu ha e shown mixed esul s in e ms o en e p ise sus ainabili y and job c ea ion.
In esponse, policy discussions ha e inc easingly ocused on in eg a ed suppo
models— hose ha combine access o inance wi h capaci y-building, ma ke access
1129
The Indian Jou nal o Labou Economics (2025) 68:1113–1132
ISLE
acili a ion, and digi al onboa ding. Recen assessmen s by he Wo ld Bank (2018)
indica e ha add essing India’s MSME inancing gap is necessa y bu no su icien .
Wi hou complemen a y in e en ions—such as suppo o digi al adop ion (Pa a-
meswa an & Kadam 2025), wo k o ce skilling, and supply chain in eg a ion—mos
mic oen e p ises emain s uck in low-p oduc i i y equilib ia (Wo ld Bank, 2018).
Inco po a ing beha iou al insigh s in o p og amme design is equally c i ical.
Resea ch by Ka lan e al. (2019) has shown ha small i m owne s o en unde in-
es no because o poo e u ns, bu due o mispe cep ions abou isk, ime incon-
sis ency, o limi ed cogni i e bandwid h. These ic ions a e no add essed by wage
subsidies alone. In such cases, pee men o ing, planning ools, o ligh - ouch mana-
ge ial aining—as es ed by Field e al. (2013) in India—can signi ican ly imp o e
i m ou comes, pa icula ly among women and in o mal en ep eneu s.
This shi in unde s anding has also shaped gende - ocused MSME policy. The
IFC (2019) inds ha inance-only p og ams ha e limi ed impac o women-led
businesses unless accompanied by non- inancial se ices such as business de elop-
men suppo , digi al access, and aining in bookkeeping and compliance.
India’s digi al in as uc u e o e s p omising a enues o scaling such in eg a ed
app oaches. Pla o ms like he Udyam po al and GSTN p o ide an eme ging da a
backbone o acking en e p ise pe o mance o e ime (Alonso e al. 2023). These
ools could enable a mo e owa d adap i e policy designs, whe e suppo e ol es
based on obse ed i m beha iou , a he han ixed eligibili y ules.
Ul ima ely, he indings om his eplica ion sugges ha he p oblem is no
me ely one o i m size o capi al access, bu o policy agmen a ion. A wage sub-
sidy may enable a i m o hi e one addi ional wo ke . Bu unless ha i m is con-
nec ed o o mal ma ke s, equipped wi h basic manage ial skills, and suppo ed by
complemen a y in e en ions, he p oduc i i y gains will emain ansi o y. The uni
o in e en ion, he e o e, mus shi — om he i m in isola ion o he b oade eco-
sys em in which ha i m ope a es.
5 Conclusion
This pape has conduc ed a igo ous s a is ical eplica ion o de Mel e al. (2019),
alida ing hei key empi ical inding: empo a y wage subsidies o mic oen e -
p ises in S i Lanka p oduced sho - e m gains in employmen and i m su i al,
bu hese e ec s did no pe sis a e he subsidy ended. The eplica ion ep o-
duced ea men e ec s ac oss all majo ou comes—employmen , p o i s, capi al
s ock, and ma ginal e u ns o labou —wi h ema kable p ecision, con i ming he
empi ical obus ness o he o iginal s udy.
Howe e , his eplica ion goes beyond echnical e i ica ion. Fi s , i e is-
i s and expands he he e ogenei y analysis ini ially conduc ed by de Mel e al.
(2019), sha pening i s in e p e i e alue. While he o iginal pape ound sugges-
i e bu s a is ically weak e idence ha manu ac u ing i ms and hose wi h be e
baseline business p ac ices we e mo e likely o hi e, his eplica ion inds ha
i ms wi h paid employees a baseline we e signi ican ly mo e esponsi e o he
wage subsidy, whe eas solo i ms showed no such e ec . This nuance—absen
1130
The Indian Jou nal o Labou Economics (2025) 68:1113–1132
ISLE
om he o iginal emphasis—sugges s ha baseline employmen s uc u e, a he
han sec o pe se, may be a s onge p edic o o subsidy esponsi eness. This
insigh no only a i ms bu also ex ends he o iginal he e ogenei y indings, mak-
ing a no el con ibu ion o he li e a u e and o e ing p ac ical guidance o be e
a ge ing o wage subsidy p og ammes.
Second, he eplica ion challenges he o iginal s udy’s conclusion ha he null
long- un esul s con i m he neoclassical model o diminishing ma ginal e u ns
o labou . Ins ead, i a gues ha such ansience may be symp oma ic o deepe
s uc u al and beha iou al cons ain s—including limi ed manage ial capaci y,
in o mali y, low us in long- e m con ac s, and policy discon inui y. In ag-
ile o pos -con lic economies, e en empo a y hi ing gains can ha e non- i ial
wel a e bene i s, such as smoo hing income, p ese ing i m su i al, o educing
labou ma ke sca ing.
Finally, by si ua ing hese indings wi hin he con ex o India’s MSME sup-
po p og ams (e.g., PMEGP, MUDRA, and ABRY), he pape demons a es how
eplica ion can se e as a diagnos ic ool o policy e inemen , no me ely an
academic exe cise in e i ica ion. While India sha es S i Lanka’s emphasis on
employmen -based suppo , he eplica ion unde sco es ha wi hou comple-
men a y in es men s in i m capabili ies and ins i u ional sca olding, such p o-
g ammes a e unlikely o p oduce sus ainable en e p ise ans o ma ion.
In sum, his s udy a i ms he empi ical alidi y o de Mel e al. (2019) while
ecas ing i s in e p e a ion h ough a iche heo e ical and policy lens. The con-
ibu ions o his eplica ion lie no only in wha i con i ms, bu also in wha
i e ames. I ad oca es o an expanded iew o eplica ion—one ha includes
c i ical ein e p e a ion, nuanced subg oup analysis, and policy con ex ualisa-
ion—as an essen ial pilla o cumula i e knowledge in de elopmen economics.
Fu u e esea ch should p io i ise bundled in e en ions ha combine wage
subsidies wi h manage ial aining, access o inance, o echnology suppo o
es whe he sus ained p oduc i i y gains can be achie ed. Expe imen al designs
should also explo e longe -du a ion subsidies, as well as g adua ed phase-ou
mechanisms, o assess whe he a mo e g adual wi hd awal enhances e en ion
e ec s. Addi ionally, mixed-me hods app oaches—combining RCTs wi h quali-
a i e ools—may unco e mechanisms ha a e obscu ed in pu ely quan i a-
i e analyses, such as wo ke mo ale, manage ial in en ions, o in o mal labou
dynamics. Finally, eplica ion e o s should become a mo e ou ine componen
o empi ical economics, especially in con ex s whe e policy decisions hinge on
indings om a single s udy.
Funding Open Access unding enabled and o ganized by P ojek DEAL.
Da a a ailabili y The o iginal da a ha suppo he indings o his s udy a e openly a ailable a he
named da a sou ces, while he eplica ion da a and code a e a ailable on eques om he au ho .
Decla a ions
Con lic o in e es The au ho epo s no po en ial con lic o in e es .
1131
The Indian Jou nal o Labou Economics (2025) 68:1113–1132
ISLE
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 eco mmons. o g/ licen
ses/ by/4. 0/.
Re e ences
Alonso, C., T. Bhojwani, E. Haneda , D. P iha dini, G. Uña, and K. Zhabska. 2023. S acking up he
bene i s: Lessons om India’s digi al jou ney. IMF Wo king Pape s No. WP/23/78. In e na ional
Mone a y Fund. Accessed Augus 5, 2025 om h ps:// sdgkn owled gehub. undp. o g. in/ wp- con e n /
uploa ds/ 2022/ 07/ S ack ing- up- he- Bene i s. pd .
Ame ican Economic Associa ion (AEA). 2019. AEA Code o P o essional Conduc and Disclosu e Poli-
cies. Accessed Augus 7, 2025 om h ps:// www. aeaweb. o g/ e hics.
A kin, D., A. Chaudh y, S. Chaud y, A. K. Khandelwal, and E. Ve hoogen. 2017. O ganiza ional ba ie s
o echnology adop ion: E idence om socce -ball p oduce s in Pakis an. The Qua e ly Jou nal o
Economics 132 (3): 1101–1164. h ps:// doi. o g/ 10. 1093/ qje/ qjx010.
Bane jee, A. V., and E. Du lo. 2011. Poo economics, a adical e hinking o he way o igh global po -
e y. New Yo k: Public A ai s.
Beaman, L., J. Mag ude , and J. Robinson. 2014. Minding small change among small i ms in Kenya.
Jou nal o De elopmen Economics 108:69–86.
Becke , Ga y S.. 1962. In es men in human capi al: A heo e ical analysis. Jou nal o Poli ical Economy
70 (5): 9–49. h ps:// doi. o g/ 10. 1086/ 258724.
Bloom, N., B. Ei e , A. Mahajan, D. McKenzie, and J. Robe s. 2013. Does managemen ma e ? E i-
dence om India. Qua e ly Jou nal o Economics 128 (1): 1–51. h ps:// doi. o g/ 10. 1093/ qje/
qjs044.
Böcke man, P., E. Johansson, and A. Kauhanen. 2012. Inno a i e wo k p ac ices and sickness absence:
Wha does a na ionally ep esen a i e employee su ey ell? Indus ial and Co po a e Change 21
(3): 587–613. h ps:// doi. o g/ 10. 1093/ icc/ d 047.
B yan, G., S. Chowdhu , and A. M. Moba ak. 2014. Unde in es men in a p o i able echnology: The
case o seasonal mig a ion in Bangladesh. Econome ica 85 (5): 1671–1748.
Clemens, M. A. 2017. The meaning o ailed eplica ions: A e iew and p oposal. Jou nal o Economic
Su eys 31 (1): 326–342. h ps:// doi. o g/ 10. 1111/ joes. 12139.
de Mel, S., D. McKenzie, and C. Wood u . 2019. Labo d ops: Expe imen al e idence on he e u n
o addi ional labo in mic oen e p ises. Ame ican Economic Jou nal: Applied Economics 11 (1):
202–235.
Di ec o a e Gene al o Employmen . 2025. Employmen gene a ion schemes/p og ammes o Go e nmen
o India. Accessed Augus 5, 2025 om h ps:// dge. go . in/ dge/ schem es_ p og ammes.
Field, E., R. Pande, J. Papp, and N. Rigol. 2013. Does he classic mic o inance model discou age en e-
p eneu ship among he poo ? Expe imen al e idence om India.Ame ican Economic Re iew 103
(6): 2196–2226. h ps:// doi. o g/ 10. 1257/ ae . 103.6. 2196.
G eenhalgh, L., and Z. Rosenbla . 1984. Job insecu i y: Towa d concep ual cla i y. The Academy o
Managemen Re iew 9 (3): 438–448. h ps:// doi. o g/ 10. 2307/ 258284.
Ha dy, Ma. 2023. A e small i ms labo cons ained? Expe imen al e idence om Ghana. Ame ican Eco-
nomic Jou nal: Applied Economics 15 (2): 253–284.
IFC. 2019. Financial inclusion o woman-owned mic o, small & medium en e p ises (MSMEs) in India.
Accessed Augus 5, 2025 om h ps:// www. i c. o g/ con e n / dam/ i c/ doc/ mg / inan cial- inclu sion-
o - women- owned- msmes july- 31. pd .
Kahn, Lisa B. 2007. The long- e m labo ma ke consequences o g adua ing om college in a bad econ-
omy. Labou Economics 17 (2): 303–316. h ps:// doi. o g/ 10. 1016/j. labeco. 2009. 09. 002.
1132
The Indian Jou nal o Labou Economics (2025) 68:1113–1132
ISLE
Ka lan, D., and J. Zinman. 2011. Mic oc edi in heo y and p ac ice: Using andomized c edi sco ing o
impac e alua ion. Science 332 (6035): 1278–1284. h ps:// doi. o g/ 10. 1126/ scien ce. 12001 38.
Ka lan, D., S. Mullaina han, and B. N. Ro h. 2019. Deb aps? Ma ke endo s and moneylende deb in
India and he Philippines. Ame ican Economic Re iew: Insigh s 1 (1): 27–42.
Lucas, R. E., J . 1978. On he size dis ibu ion o business i ms. The Bell Jou nal o Economics
1:508–523.
Maheshka , C., and N. Soni. 2022. P oblems aced by Indian mic o, small and medium en e p ises
(MSMEs). SEDME (Small En e p ises De elopmen , Managemen & Ex ension Jou nal) 48 (2):
142–159. h ps:// doi. o g/ 10. 1177/ 09708 46421 10644 98.
McKenzie, D. 2012. Beyond baseline and ollow-up: The case o mo e T in expe imen s. Jou nal o
De elopmen Economics 99 (2): 210–221.
Minis y o MSME, Go e nmen o India. 2023. Annual Repo 2022–2023. h ps:// dcmsme. go . in/
Annua lRepo _ Tcs_ 2022_ 23. aspx.
Muske, G., and M. Woods. 2016. Mic o businesses as an economic de elopmen ool: Wha hey b ing
and wha hey need. Communi y Economic De elopmen 2016:111–130.
Pa ameswa an, S., and A. Kadam. 2025. Access o o mal c edi o women-owned en e p ises in India’s
uno ganized sec o : Does echnology adop ion help close he gende gap? Applied Economics
2025:1–14. h ps:// doi. o g/ 10. 1080/ 00036 846. 2025. 25268 53.
Pa wa dhan, S., and L. Tascio i. 2022. The e ec o he Maha ma Gandhi Na ional Ru al Employmen
Gua an ee Ac on he size o ou s anding deb s in u al India. Jou nal o De elopmen E ec i eness
15 (4): 353–372. h ps:// doi. o g/ 10. 1080/ 19439 342. 2022. 21031 69.
Wo ld Bank. 2018. Financing India’s MSMEs: Es ima ion o deb equi emen o MSMEs in India. In e -
na ional Finance Co po a ion. Accessed Augus 5, 2025 om h ps:// docum en s1. wo ld bank. o g/
cu a ed/ en/ 75926 15488 28982 149/ pd / 134150- WP- IN- Finan cing- India-s- MSMEs- Es im a ion- o -
Deb - Requi emen - o - MSMEs- PUBLIC. pd .
Du lo, E., Glenne s e , R., and K eme , M. 2007. Using andomiza ion in de elopmen economics
esea ch: A oolki . Handbook o de elopmen economics 4:3895-3962. h ps:// doi. o g/ 10. 1016/
S1573- 4471(07) 04061-2
Publishe ’s No e Sp inge Na u e emains neu al wi h ega d o ju isdic ional claims in published maps
and ins i u ional a ilia ions.
