Picci, Lucio
Wo king Pape
P oduc i i y and In as uc u e in he I alian Regions
Quade ni - Wo king Pape DSE, No. 230
P o ided in Coope a ion wi h:
Uni e si y o Bologna, Depa men o Economics
Sugges ed Ci a ion: Picci, Lucio (1995) : P oduc i i y and In as uc u e in he I alian Regions,
Quade ni - Wo king Pape DSE, No. 230, Alma Ma e S udio um - Uni e si à di Bologna,
Dipa imen o di Scienze Economiche (DSE), Bologna,
h ps://doi.o g/10.6092/unibo/amsac a/5102
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/159073
S anda d-Nu zungsbedingungen:
Die Dokumen e au EconS o dü en zu eigenen wissenscha lichen
Zwecken und zum P i a geb auch gespeiche und kopie we den.
Sie dü en die Dokumen e nich ü ö en liche ode komme zielle
Zwecke e iel äl igen, ö en lich auss ellen, ö en lich zugänglich
machen, e eiben ode ande wei ig nu zen.
So e n die Ve asse die Dokumen e un e Open-Con en -Lizenzen
(insbesonde e CC-Lizenzen) zu Ve ügung ges ell haben soll en,
gel en abweichend on diesen Nu zungsbedingungen die in de do
genann en Lizenz gewäh en Nu zungs ech e.
Te ms o use:
Documen s in EconS o may be sa ed and copied o you pe sonal
and schola ly pu poses.
You a e no o copy documen s o public o comme cial pu poses, o
exhibi he documen s publicly, o make hem publicly a ailable on he
in e ne , o o dis ibu e o o he wise use he documen s in public.
I he documen s ha e been made a ailable unde an Open Con en
Licence (especially C ea i e Commons Licences), you may exe cise
u he usage igh s as speci ied in he indica ed licence.
h ps://c ea i ecommons.o g/licenses/by-nc/3.0/
P oduc i i y and In as uc u e in he I alian Regions.
Lucio Picci1
Uni e si y o Bologna
Augus 1995
J.E.L.: H54, C80
Abs ac .
Weadd ess heissueo whe he publicin as uc u eplayanimpo an oleinde e mining
ac o p oduc i i y in I aly, and we show ha he e idence is mixed.
Publiccapi alissigni ican inexplainingou pu inmos cases.Howe e ,when hea en ion
is d awn on he long- un p ope ies o he da a, o when ca e is aken o ule ou
con empo aneous sho - un e ec s, hen public capi al esul s o be ei he non-signi ican ,
o signi ican bu o negligibleimpo ance.Weconclude ha hein luenceo in as uc u e
on ou pu is p obably due, o a g ea ex en , o sho - un demand-side phenomena.
1 Dipa imen o di Scienze Economiche, S ada Maggio e 45, I-40125 Bologna,
I aly. Tel. +39-51-6402608, Fax. +39-51-6402664. E-mail: [email p o ec ed]
Iwouldlike o hank o he aluablecommen sDouglasHol z-EakinandSe gioPas o ello.
1 In oduc ion.
The huge public deb a ec ing he I alian economy is an ob ious eason o conce n o
economis sandpolicy make salike. Jappelliand RipadiMeana(1990)a gue ha policies
whosegoalis o educe hedeb /ou pu a iocanno abs ac om he ole ha in as uc u e
playin heeconomy. I in as uc u edo in luence ou pu signi ican ly, hencu ing public
capi al in es men , by educing u u e po en ial ou pu , could e en wo sen he p oblem.
On he o he hand, i ha is no he case, hen policy make s, in a sense, could do wi hou
ca ing oo much on wha kind o budge cu s hey make, as long as hey make hem.
The in e es in he ela ionship be ween in as uc u e and ou pu o , mo e o he poin ,
ac o p oduc i i y,howe e ,isno limi ed o hisissue.In heUni edS a es,mucha en ion
has been dedica ed o he analysis o he causes o he p oduc i i y slowdown o he las
wo decades.A whole h ead o he li e a u e, s a ing om he seminal pape o Aschaue
(1989), has ied o impu e he decline o ac o p oduc i i y o he decline in public
in es men 2.
The oleplayed byin as uc u e in he economy has beenanalyzed by acoun less numbe
o wo ks. A a simple and in ui i e le el, we no e ha in us uc u e a ec ac o
p oduc i i y by "assis ing" public capi al: a gi en uck is much mo e p oduc i e on a
eeway, han on a coun y oad. Mo eo e , new in as uc u e embodies echnological
p og ess: o build a b and-new mass- ansi sys em is an amoun o he in oduc ion o a
new echnology in o a local economy. In as uc u e, also, allows o echnological
p og ess: we en’ oads a ailable a all, we wouldn’ ha e ucks. A he same ime, all
hesee ec sp obably ake ime o mani es hemsel es, since heyin ol ec ea iono new
business, eloca ion o exis ing ac i i y, and lea ning3.
The poin o in e es , ob iously, is no whe he public capi al is signi ican in de e mining
p oduc i i y,a hesis ha e y ew people,i anybody,wouldques ion.Wha isin e es ing
ounde s andis owha ex en in as uc u ea ec p oduc i i ybeyondi sdi ec p o ision
o ameni ies, on one side, and wha a e he mechanisms ha make hem in luence
p oduc ion.Bo h ques ions ha e impo an policy implica ions. Ananswe o he i s one
2 See Ebe s (1990), Munnell (1990), Hol z-Eakin (1994), and Hol z-Eakin and
Schwa z (1994).
3 Fo a e ue essay, also illumina ing o he poin s ouched by his wo k, see
G amlich (1994).
1
would helppolicy make sin choosing he op imalamoun o publicin es men ; ananswe
o he second one could p o ide use ul guidelines on how o make public in es men
succes ul.
Using I alian egional da a, his wo k ies o add ess he i s o hese wo ques ions. We
y o de e mine he impo ance o in as uc u e in de e mining p oduc ion in he I alian
egions4o e he pe iod 1970-1991. Only e y ecen ly he a ailabili y o he necessa y
da a has made his ype o s udy possible o he I alian economy. We conclude ha
in as uc u e a e signi ican in explaining egional ou pu , bu ha his is p obably due o
he p esence o sho - un demand-side, as opposed o supply-side, e ec s.
Sec ion 2, wi h e e ence o a ew ep esen i e wo ks, examines he exis ing empi ical
e idence on he issue, bo h o he U.S. and o I aly. Sec ion 3 deals wi h he analysis o
he empi ical esul s. Sec ion 4 concludes.
2 The Exis ing E idence.
Aschaue (1989) shows he esul s o a ious eg essions o di e en concep s o ac o
p oduc i i y on ac o inpu s, including public capi al, o he U.S. as a whole. The esul s
a e qui e as ounding: public capi al seems o be highly signi ican in in luencing
p oduc i i y,wi hposi i eelas ici ieso a ound0.35.Simila esul sa e oundbyMunnell
(1990a) who, again o he U.S. as a whole, inds elas ici ies o ou pu wi h espec o
public capi al o compa able magni udes.
Munnell(1990b)analyses heimpac o publiccapi alonp oduc i i yusingins eadapanel
o U.S. s a e da a. The esul s o he pooled OLS eg essions indica e a coe icien o
public capi al lowe (0.15) han in he ime se ies s udies men ioned abo e, bu s ill highly
signi ican .
4 I alyisdi idedin o20 egions, some imes, o desc ip i epu posesonly,lumped
in o i e mac o- egions.
The egions (and he mac o- egions) a e: Piedmon , Val d’Aos a, Ligu ia and Lumba dy
(No h-Wes ); T en ino-Al o Adige, Vene o, F iuli Venezia Giulia and Emilia Romagna
(No h-Eas ); Tuscany, Ma che, Umb ia and Lazio (Cen e ); Ab uzzo, Molise, Apulia,
Campania,Basilica aandCalab ia(Sou h);SicilyandSa dinia(Islands).Sou handIslands
oge he a e called he Mezzogio no o I aly.
2
The i s wa e o wo ks on he opic le he imp ession ha public capi al was indeed e y
impo an inde e miningp oduc i i y,and ha i sdec easedg ow h a eswe eap ominen
candida e o explain he p oduc i i y slowdown ha he U.S. economy has expe ienced in
he las decades.
A e a i s phase o en husiasm, howe e , some cau ion was asked o by o he
less-op imis ic esul s. Hol z-Eakin (1994) c i icizes he econome ic analysis ca ied ou
byMunnell(1990b),anda gues ha ,inac oss-s a eanalysiso p oduc i i y,s a especi ic
e ec s a e po en ially impo an . Hol z-Eakin hen ejec s he hypo hesis ha indi idual
s a e e ec s a e no ele an and inds e idence agains he hypo hesis ha public capi al
plays a ole in de e mining ou pu . The same esul also eme ges when di e en
econome ic echniques, such as ins umen al a iables panel es ima ion and es ima ion
using long di e ences o he da a, a e used.
Hol z-Eakin and Schwa z (1994) ake a di e en app oach, and de elop a neoclassical
g ow hmodel"à laSolow" ha inco po a esin as uc u eas oneo hep oduc ioninpu s.
They check whe he he da a con o m o he p edic ions o hei model, and again ind
essen ially no ole o public capi al.
E idence o I aly is scan , due a leas in pa o lack o he necessa y da a. B acalen e
and Di Palma (1982) compu e a se ies o indexes o measu e in as uc u e o he I alian
egions o he yea 1977. Usingbo h OLS analysis and o a ed- ac o eg ession analysis,
hey ind ha in as uc u e a e signi ican in explaining egional de elopmen . The
de e mina ion o he di ec ion o causali y, howe e , emains an open ques ion in ha
wo k.
Jappelliand Ripa di Meana (1990) es ima e ase ies o educed- o m equa ions o ou pu ,
whe e bo h p i a e consump ion and public in es men a e used as eg esso s. The
es ima ed coe icien s o public capi al a e signi ican and bigge han he coe icien s o
p i a econsump ion; he au ho sin e p e his esul as e idence in a o o he hypo hesis
ha he e ec o public capi al is impo an in de e mining ou pu .
Picci (1994) uses a da a se de eloped by Rossi e al. (1993) and by Picci (1995a) on
agg ega e public capi al o he pos WWII pe iod and o he be ween-wa pe iod, o
es ima e a numbe o eg essions simila o he ones conside ed by Aschaue (1989) and
Munnell (1990). The esul s gene ally indica e a signi ican ole o public in as uc u e,
wi h e y high ou pu elas ici ies. This ype o analysis pe mi s, o some ex en , o add ess
3

he ques ion o he di ec ions o he casual ela ion be ween in as uc u e and ou pu .
G ange -causa ion analysis be ween mul i ac o p oduc i i y (o "Solow esiduals") and
heg ow h a eo publiccapi algi eshowe e ambiguous esul s,dependingon hesample
pe iod conside ed.
3 The Analysis o Regional Da a.
In wha ollows, we analyze he incidence o in as uc u e on ou pu using a ecen ly
de eloped da ase on egional in as uc u e co e ing he pe iod om 1970 o 1991, o a
o al o 22 annual obse a ions. Regional public capi al s ocks ha e been compu ed using
he pe pe ual in en o y echnique. The necessa y egional public in es men ime se ies
ha e been ob ained by appo ioning he na ional agg ega e in Rossi e al. (1993) using he
yea lyda aon"publicwo ks"collec edbyIs a (Is a , a iousyea s). Thewholep ocedu e
is ai ly in ol ed; ull de ails a e in Picci (1995a).
P i a e capi al also has been aken om Picci (1995a), whe e i has been compu ed using
a benchma k o he egional capi al s ock o he census yea 1981 and egional g oss
p i a e in es men s o he emaining yea s. Ou pu is egional g oss p oduc , and labo
is egional uni s o labo (sou ce: SVIMEZ (1993); ISTAT (1990a, 1990b, 1992).
Table1 shows he a e ageg ow h a e o pe capi a ou pu ,labo uni s, p i a e and public
capi al, oge he wi h hei beginning- and end-o -sample le els, o I aly as a whole and
o i s i e mac o- egions. The Mezzogio no (comp ising he Sou he n and he Insula
egions) has a pe -capi a income much lowe han he na ional a e age. Con e gence has
no occu ed o e he wo decades conside ed, and he wo low income mac o- egions
ha e bo h sco ed below-a e age pe -capi a income g ow h a es.
Labo uni s inc eased by li le and a he e enly ac oss I aly5.
P i a e capi al has inc eased d ama ically, mos ly in he Mezzogio no. In compa ison,
publiccapi al has inc eased byless, whi hg ow h a es o a ound3.5% yea ly ac ossI aly.
A mo e de ailed b eak-up o he da a would show ha mos o he inc ease o he capi al
5 Howe e ,whilein heNo ho I aly hisg ow his e lec edbyasensibleinc ease
in he a io o labo uni s o e popula ion, wi h a mo e o less cons an popula ion, in he
Mezzogio no popula ion has also g own.
4
s ock - bo h p i a e and public - in he Mezzogio no occu ed du ing he i s pa o he
sample pe iod, while he huge e o o indus ialize he less-de eloped Sou he n and
Insula egions was s ill ongoing. Du ing he ’80’s, wi h he dismission o he "Cassa del
Mezzogio no"(a s a e agencyaimed a he de elopmen o he Mezzogio no), he massi e
in es men s in ha pa o I aly ha had cha ac e ized he p e ious decades came o an
end.
No e ha , acco ding o he da a, and con a y o in ui ion, he less-de elopedMezzogio no
has a s ock o in as uc u e bigge han he wo indus ialized No he n mac o- egions.
Picci (1995b) compa es measu es o egional in as uc u e endowmen compu ed using
pe pe ual in en o y da a - le ’s call hem, "PI" (pe pe ual in en o y) indexes -, wi h he
indexes in Biehl, B acalen e, Di Palma and Mazzio a (1990), based on he physical
consis ency o he capi al s ock - le ’s call hem "BBDPM" indexes -. Bo h indexes a e
ob ained by de la ing egional capi al s ocks ei he by egional popula ion - in he case o
"popula ion-se ing" in as uc u e, such as schools, hospi als, e c. - o by egional a ea -
in he case o "space-se ing" in as uc u e, such as oads and ail oads. A compa ison o
he indexes gi es some indica ion on he ela i e e iciency o he di e en egions in
p oducing in as uc u e: a PI index highe han a BBDPM index, would show ela i e
ine iciency in building in as uc u e, indica ing he p esence o li le in as uc u e in
ela ion o he amoun o esou ces spen o p oduce i , and ice- e sa.
Picci (1995b), while wa ning abou he isk o d awing has y conclusions, shows ha , in
his espec , he e is a e y sha p di e ence be ween he Mezzogio no and he es o I aly:
all he Mezzogio no egions ha e PI indexes gene ally much highe han he BBDPM
indexes.Thisispa icula ly ue o Campania, Calab ia andSicilia, he h ee big Sou he n
egions plagued by o ganized c ime. On he o he hand, only Ligu ia, among he No he n
egions, shows he same ca ach e is ic. This should come as no su p ise: Ligu ia, a
moun ainous and densely popula ed egion, is ce ainly cha ac e ized by highe cos s in
buildingin as uc u e.Fo all heo he No he nandCen al egions,PIindexesa elowe
hen BBDPM indexes. In o he wo ds, he e is a case o o e es ima ion o he public
capi al s ock compu ed using he pe manen in en o y echnique o Sou he n I aly in
gene al, e en mo e so o Campania, Calab ia, Sicily, and also o he No he n egion o
Ligu ia.Figu e1showsbo hindexes o all egions.The45deg eeslinedi ides"e icien "
egions (below) om "ine icien " ones (abo e).
5
The empi ical analysis is ca ied ou by es ima ing a p oduc ion unc ion whe e public
capi al is used in conjunc ion wi h p i a e capi al and labo inpu o assess i s impo ance
in de e mining ou pu .
The es ima ed equa ion is:
whe e is ou pu , is p i a e capi al, is labo , is public capi al, and deno es
logs. The and subsc ip indica e, espec i ely, egion and ime.
The e o e m has he ollowing s uc u e:
is a egion speci ic componen ; is a ime speci ic componen , and is aidiosinc a ic
i.i.d e o .
Di e en speci ica iono hemodels -and, aswe ha eseen, o en di e en esul s- ollow
om di e en assump ions abou he e o e m. No conside ing gi es pooled OLS, he
echnique used by Munnell (1990). Hol z-Eakin (1994), in his analysis on U.S. s a e
p oduc i i y, a gues ha s a e e ec s a e po en ially impo an . Once he conside s hem,
he o e u ns Munnell’s esul s.
I is impo an o s ess ha de ec ing asigni ican publiccapi alcoe icien does no imply
ha publiccapi alcausesou pu .Fi s , he ela ioncouldbeexplainedby e e se-causa ion,
wi h policy make s esponding wi h inc eased public in es men o be e economic
condi ions. Mo eo e , e en i in us uc u e do in luence ou pu , a dis inc ion be ween
"supply" and "demand" e ec s should be d awn. I could be ha he signi ican
in as uc u e coe icien in he es ima ed equa ion is due o i s e ec on he unde lying
de e minan s o p oduc i i y. On he o he hand, such a coe icien could esul om he
e ec s o he inc eased public expendi u e on demand. This ype o demand e ec could
bepa icula ly impo an in Sou he n I aly, whe e he sizeo public in es men , especially
du ing he i s pa o he ’70’s, was a conside able ac ion o ou pu (see igu e 2).
Is he e a way o disc imina e be ween hese wo ype o e ec s? I seems easonable o
ly = α0 + α1lk + α2llab + α3lkpub + ε ,(1)
y k lab kpub l
εi = + δ + η .
δ η
6
expec ha sho -li ed demand e ec s should be de ec ed by analysis ha ocus on he
sho - un ime se ies p ope ies o he da a, such as ixed-e ec s OLS eg essions. On he
o he hand, long- un supply e ec s should be de ec ed by ocusing on he long un
c oss-sec ional dimension o he da a, o example by es ima ing long-di e ences o he
da a.
The i s and second columns o able 2 show, espec i ely, esul s o pooled eg essions
and ixed-e ec s es ima es o egional p oduc ion unc ions o he I alian egions6. In he
pooledOLS eg ession esul s, hees ima edcoe icien o publiccapi alisemba assingly
nega i e and signi ican .
By es ima ing OLS eg essions sepa e ely o he 20 egions ( esul s no epo ed he e bu
a ailable om he au ho ), we ob ain gene ally posi i e and signi ican es ima es o he
public capi al coe icien s, as in Picci (1994), o he I alian economy, and in Aschaue
(1989), o heUS economy. The(unweigh ed) a e age o hese 20 es ima ed coe icien s,
ha end o be bigge o he Sou he n egions, is equal o 0.504.
The ixede ec ses ima es,unlike hepooledOLSes ima es,showsigni ican andposi i e
public capi al elas ici y. The F- es on he null hypo hesis ha all ixed e ec s a e equal
( ha is, ha he e a e no ixed e ec s) is s ongly ejec ed. As in Hol z-Eakin (1994) we
a e led o conclude ha OLSpooled eg essions a eno consis en . Unlike inHol z-Eakin,
assuming ixed-e ec sdoes no change he esul s ob ained wi h heagg ega e ime se ies
app oach. The es ima ed public capi al oup u elas ici ies is e y high, and compa able o
he esul s ob ained by unning sepa a e OLS eg essions. No e also ha he es ima ed
labo ou pu elas ici y is implausibly bigge han one.
Random e ec s es ima es, shown in column c o able 2, p o ide simila esul s. The
Hausman es ,howe e , ejec s henullhypo hesis ha he egionale ec sa eunco ela ed
wi h he igh -hand side a iables o he eg ession. In his case, andom e ec s es ima es
a e biased, while ixed e ec s es ima es a e s ill consis en .
Fixed e ec s es ima ion accoun s only o a ia ion in he ime se ies dimension o he
da a,andlea esno oom o c oss-sec ional a ia ion.A heopposi e endo hespec um,
OLS eg essions on long-di e ences o he da a conside only c oss sec ional a ia ion.
6 All panel da a eg essions include ime e ec s.
7
Table 1
Summa y s a is ics.
Ou pu pe capi a N. Wes N. Eas Cen e Sou h Islands I aly
A g. g ow h a e: 3.36% 4.04% 3.57% 3.49% 3.16% 3.52%
1970 12.20 10.63 10.36 6.75 7.06 9.66
1991 20.81 19.66 18.12 11.71 11.74 16.80
Pe capi a ou pu : million o ’85 li e.
Labo Uni s N. Wes N. Eas Cen e Sou h Islands I aly
A g. g ow h a e .38% 1.15% 1.12% .89% .89% .83%
1970 6243.0 3942.0 3829.3 4048.0 1887.0 19949.4
1991 6786.5 4891.5 4726.7 4806.5 2238.6 23449.8
P i a e Capi al N. Wes N. Eas Cen e Sou h Islands I aly
A g. g ow h a e 6.35% 5.38% 8.19% 11.78% 9.90% 7.34%
1970 164674.5 125344.0 76512.4 56237.5 35740.2 458508.7
1991 384359.3 267076.8 208139.1 195326.8 110079.0 1164981
Public Capi al N. Wes N. Eas Cen e Sou h Islands I aly
A g. g ow h a e 3.87% 3.86% 3.06% 3.44% 3.69% 3.56%
1970 72718.0 67924.9 72717.1 103456.1 52944.7 369760.9
1991 131819.2 122936.9 119383.7 178190.3 93955.5 646285.6
P i a e and public capi al: billion o 1985 li e.
14

Table 2
Dependen a iable: G oss Regional P oduc .
All a iables a e in logs.
I aly.
Va iable: a) OLS b) F.E. c) R.E. d) L.D.
Cons an 2.445 - .200 .453
(21.63) (1.28) (3.98)
P i a e Capi al .248 .097 .145 .011
(20.76) (7.81) (12.59) (.293)
Labo .863 1.080 .737 .807
(49.41) (18.79) (39.13) (3.56)
Public Capi al -.063 .430 .355 -.012
(-4.15) (15.51) (14.58) (-.08)
.993 .945 .948 .459
Tes : : no FE. P-Value: .000
Hausman Tes : FE s. RE. P-Value: .000
OLS: Pooled OLS;
F.E.: Fixed E ec s;
R.E.: Random E ec s;
L.D.: Long Di e ences.
-s a is ics a e be ween pa en heses.
R2
H0
15
Table 3
Dependen a iable: G oss Regional P oduc .
All a iables a e in logs.
I aly excl. Cal, Camp, Sic, Lig.
Va iable: a) OLS b) F.E. c) R.E. d) L.D.
Cons an 2.184 - .134 .512
(17.21) (.81) (2.95)
P i a e Capi al .224 .088 .126 .012
(18.46) (6.47) (9.98) (1.27)
Labo .849 1.22 .758 .789
(45.21) (18.58) (32.16) (2.39)
Public Capi al -.005 .413 .374 -.090
(-.26) (13.68) (14.04) (-.48)
.994 .945 .951 .438
Tes : : no FE. P-Value: .000
Hausman Tes : FE s. RE. P-Value: .000
OLS : Pooled OLS;
F.E.: Fixed E ec s;
R.E.: Random E ec s;
L.D.: Long Di e ences.
-s a is ics a e be ween pa en heses.
R2
H0
16
Table 4
Dependen a iable: G oss Regional P oduc .
All a iables a e in logs.
No h and Cen e excl. Lig.
Va iable: a) OLS b) F.E. c) R.E.
Cons an 1.151 - .197
(7.78) (1.52)
P i a e Capi al .200 .390 .418
(11.80) (16.67) (20.91)
Labo .700 .846 .588
(31.53) (13.48) (26.56)
Public Capi al .236 .171 .178
(9.37) (6.61) (7.27)
.997 .977 .991
Tes : : no FE. P-Value: .000
Hausman Tes : FE s. RE. P-Value: .019
OLS : Pooled OLS;
F.E.: Fixed E ec s;
R.E.: Random E ec s;
-s a is ics a e be ween pa en heses.
R2
H0
17
Table 5
Dependen a iable: G oss Regional P oduc .
All a iables a e in logs.
Public capi al is lagged one pe iod.
I aly I aly excl. Cal, No h and
Camp, Sic, Lig. Cen e excl. Lig.
Va iable: a) F.E. b) R.E. c) F.E. d) R.E. e) F.E. ) R.E.
Cons an - 1.623 - .117 - -.018
(13.42) (10.41) (-.143)
P i a e Capi al .246 .279 .102 .154 .418 .447
(17.62) (21.42) (6.25) (10.41) (16.32) (20.70)
Labo 1.342 .832 1.196 .738 .810 .553
(20.38) (38.83) (18.40) (31.69) (12.54) (24.29)
Lagged Public Capi al .013 .013 .410 .359 .179 .192
(3.26) (3.12) (12.77) (13.02) (6.66) (7.48)
.919 .944 .945 .953 .977 .981
Tes : : no FE. P-Value: .000 .000 .000
Hausman Tes : FE s. RE. .000 .000 .038
P-Value:
F.E.: Fixed E ec s;
R.E.: Random E ec s;
-s a is ics a e be ween pa en heses.
R2
H0
18
Table 6
Dependen a iable: G oss Regional P oduc .
All a iables a e i s di e ences o logs.
I aly I aly excl. Cal., Sou h and Cen e
Camp., Sic, Lig. exc. Lig.
Va iable: a) b) c) d) e) ) g) h) i)
OLS F.E. R.E. OLS F.E. R.E. OLS F.E. R.E.
.014 .007 - .008
Cons an .014 .158 .014 .015 - (4.17) (1.45) (1.60)
(4.34) (2.00) (4.16) (4.26)
P i a e Capi al .011 .013 .013 .017 .021 .017 .060 -.179 .011
(.55) (.486) (.558) (.80) (.77) (.81) (.74) (-1.46 (.12)
Labo .542 .524 .534 .529 .506 .526 .546 .487 .527
(9.79) (9.16) (9.97) (8.47) (7.90) (8.42) (6.47) (5.52) (6.20)
Public Capi al .120 .158 .136 .083 .129 .089 .073 .211 .108
(1.78) (2.00) (1.98) (1.20) (1.64) (1.27) (1.07) (2.56) (1.51)
.622 .623 .626 .659 .670 .661 .784 .797 .459
Tes : : no FE. P-Value: .568 .279 .072
H. Tes : FE s. RE. P-V: .999 .999 .999
OLS : Pooled OLS;
F.E.: Fixed E ec s;
R.E.: Random E ec s;
H. Tes : Hausman Tes .
-s a is ics a e be ween pa en heses.
R2
H0
19

Table 7
Dependen a iable: G oss Regional P oduc .
All a iables a e in logs.
Cons an e u ns o scale in he p i a e inpu s .
I aly I aly excl. Cal, No h and
Camp, Sic, Lig. Cen e excl. Lig.
Va iable: a) F.E. b) R.E. c) F.E. d) R.E. e) F.E. ) R.E.
Cons an - -.068 - .102 - .198
(-.404) (.584) (1.48)
P i a e Capi al ------
Labo .908 .844 .922 .853 .594 .585
(72.69) (76.92) (66.45) (70.84) (24.71) (33.21)
Public Capi al .472 .300 .484 .291 .203 .182
(19.37) (16.31) (17.86) (15.15) (8.09) (11.73)
.960 .947 .958 .944 .887 .893
Tes : : no FE. P-Value: .000 .000 .000
Hausman Tes : FE s. RE. .000 .000 .905
P-Value:
F.E.: Fixed E ec s;
R.E.: Random E ec s;
-s a is ics a e be ween pa en heses.
α1+α
2=1
R2
H0
20
Table 8
Dependen a iable: G oss Regional P oduc .
All a iables a e in logs.
Cons an e u ns o scale in he p i a e and public inpu s .
I aly I aly excl. Cal, No h and
Camp, Sic, Lig. Cen e excl. Lig.
Va iable: a) F.E. b) R.E. c) F.E. d) R.E. e) F.E. ) R.E.
Cons an - 2.07 - 2.190 - 1.425
(27.36) (25.76) (18.96)
P i a e Capi al ------
Labo .439 .593 .457 .623 .384 .420
(17.22) (27.52) (15.55) (25.57) (18.21) (20.02)
Public Capi al .436 .216 .433 .197 .138 .117
(13.47) (8.45) (11.73) (6.95) (4.86) (4.26)
.947 .930 .945 .953 .867 .849
Tes : : no FE. P-Value: .000 .000 .000
Hausman Tes : FE s. RE. .000 .000 .002
P-Value:
F.E.: Fixed E ec s;
R.E.: Random E ec s;
-s a is ics a e be ween pa en heses.
α1+α
2+α
3=1
R2
H0
21
Figu e 1
High In es men / Low Capi al S ock High In es men / High Capi al S ock
Low In es men / Low Capi al S ock Low In es men / High Capi al S ock
Public In es men and In as uc u e.
The egions a e:
No h-Wes : Piedmon (PIE), Val d’Aos a (VAA), Ligu ia (LIG) and Lumba dy (LOM);
No h-Eas :T en ino-Al o Adige(TAA), Vene o (VEN), F iuli-Venezia Giulia(FVG) and
Emilia-Romagna (ER);
Cen e : Tuscany (TOS), Ma che (MAR), Umb ia (UMB) and Lazio (LAZ);
Sou h: Ab uzzo (ABR), Molise (MOL), Apulia (PUG), Campania (CAM), Basilica a
(BAS) and Calab ia (CAL);
Islands: Sicily (SIC) and Sa dinia (SAR).
22
Figu e 2
NW: No h-Wes ;
NE: No h-Eas ;
C: Cen e ;
S: Sou h;
I: Islands.
23