weeks of the course ec2212 industrial growth and competitionsimonk/pdf/igc_lec1to10combo.pdf5....
Post on 21-Jun-2020
3 Views
Preview:
TRANSCRIPT
EC
2212
Ind
ustr
ial G
row
th a
ndC
ompe
titio
n•
The
rol
e of
tech
nolo
gy in
indu
stri
es, a
ndth
e ro
le o
f bo
th in
reg
iona
l/nat
iona
l gro
wth
•T
ake
cour
se n
otes
& s
emin
ar f
orm
fro
m th
e fr
ont
•L
ectu
res
star
t at 3
:05p
m
•St
udy
cour
se n
otes
bef
ore
sem
inar
s
•St
udy
othe
r re
adin
gs b
y ex
am te
rm
•Fi
ve a
ssig
nmen
ts le
ad to
pro
ject
s
•C
hoos
e pr
ojec
t top
ics
in n
ext w
eek’
s se
min
ars
Weeks of the Course1. Technology and growth.
Sources of successful technology:
2. Small firms (and regions).
3. Large firms.
Industry competition & technology:
4. Overview.
5. Market leadership change.
6. Shakeouts.
7. Sources of firm advantage.
8. Product differentiation.
9. Firm growth & technology.
10. Firm and national success.
EC
2212
Ind
ustr
ial G
row
than
d C
ompe
titio
n
Lec
ture
1
Tec
hnol
ogy
Is th
e Pr
imar
ySo
urce
of
Eco
nom
ic G
row
th
The
Man
ufac
ture
of
Pins
•A
dam
Sm
ith, T
he W
ealt
h of
Nat
ions
—ex
ampl
e of
div
isio
n of
labo
r
•Y
et te
chno
logy
, not
div
isio
n of
labo
r no
rsc
ale,
has
dri
ven
cost
of
pins
•17
70s:
ave
rage
wor
ker
4,80
0 pi
ns p
er d
ay
•19
70s:
ave
rage
wor
ker
800,
000
pins
per
day
•2.
6% a
nnua
l pro
duct
ivity
gro
wth
Tec
hnol
ogy
and
Gro
wth
•R
ober
t Sol
ow w
on N
obel
Pri
ze la
rgel
y fo
rsh
owin
g th
e im
port
ance
of
tech
nolo
gy in
econ
omic
gro
wth
•Pr
evio
us e
cono
mis
ts: c
apita
l, di
visi
on, …
•So
low
: 12.
5% o
f gr
owth
in o
utpu
t per
hou
r,no
n-fa
rm 1
909-
49, f
rom
cap
ital e
quip
men
t
•R
emai
ning
87.
5% a
ttrib
uted
to im
prov
edpr
oduc
tion
and
skill
s
Gro
wth
is M
ore
than
Fig
ures
Say
•N
ew p
rodu
cts
not i
n “b
aske
t” o
f go
ods
tom
easu
re c
onsu
mer
pri
ce in
dex
–Y
et th
ey h
ave
the
mos
t rap
id p
rice
dec
lines
–O
vere
stim
ates
infl
atio
n, u
nder
estim
ates
gro
wth
•Q
ualit
y im
prov
emen
ts n
ot r
efle
cted
inec
onom
ic g
row
th f
igur
es
…B
ecau
se o
f N
ew T
echn
olog
y
Wor
ldw
ide
Eco
nom
ic G
row
th
•G
ross
wor
ld p
rodu
ct (
GW
P):
–M
easu
red
in c
onst
ant 1
995
US
dolla
rs
–$4
.9 tr
illio
n in
195
0
–$2
6.9
trill
ion
in 1
995
–1.
6% a
vera
ge a
nnua
l gro
wth
–0.
9% a
vera
ge p
er p
erso
n
•O
utpu
t per
wor
ker
hour
has
ris
en 0
.9%
+an
nual
ly
Nat
iona
l Gro
wth
& T
echn
olog
y
•C
ompa
re g
row
th in
out
put p
er w
orke
r ho
uram
ong
lead
ing
indu
stri
al n
atio
ns–
Dif
fere
nces
bet
wee
n co
untr
ies
–D
iffe
renc
es b
etw
een
time
peri
ods
–C
onve
rgen
ce a
mon
g le
adin
g in
dust
rial
nat
ions
•L
ess-
indu
stri
aliz
ed c
ount
ries
–O
ften
dif
fere
nt p
atte
rns
(ins
titut
iona
l pro
blem
s)
–N
ot a
lway
s co
nver
genc
e
Gro
wth
Rat
es (
% p
er y
ear)
of
GD
P pe
r W
orke
r H
our
inSi
xtee
n In
dust
rial
Cou
ntri
es, 1
870-
1979
Sour
ce: W
illia
mso
n (1
991)
1870
-80
1880
-90
1890
-00
1900
-13
1913
-29
1929
-38
1938
-50
1950
-60
1960
-70
1970
-79
Aus
tral
ia1.
820.
37-0
.80
1.01
1.49
0.88
2.20
2.76
2.22
2.83
Aus
tria
1.50
1.98
1.93
1.50
0.72
0.21
1.61
5.69
5.90
4.32
Bel
gium
1.84
1.36
0.93
0.90
1.79
1.01
1.14
3.14
4.88
4.88
Can
ada
2.19
1.23
1.70
2.70
1.21
0.00
5.36
3.09
2.72
1.83
Den
mar
k1.
471.
951.
902.
212.
570.
431.
232.
974.
903.
06Fi
nlan
d1.
291.
143.
362.
421.
951.
892.
103.
966.
372.
60Fr
ance
2.32
0.90
2.02
1.82
2.34
2.83
0.75
4.39
5.38
4.09
Ger
man
y1.
502.
152.
421.
411.
402.
34-0
.40
6.64
5.29
4.50
Ital
y0.
220.
431.
202.
351.
922.
960.
564.
276.
693.
91Ja
pan
1.87
1.72
2.11
1.88
3.42
3.41
-3.2
05.
579.
965.
03N
ethe
rl.
1.44
1.26
0.98
1.07
2.44
-0.1
01.
933.
334.
934.
06N
orw
ay1.
391.
961.
172.
022.
782.
611.
884.
034.
523.
66Sw
eden
1.76
1.95
2.70
2.62
2.40
2.66
3.43
3.43
4.79
2.55
Switz
erl.
1.59
1.37
1.47
1.26
3.18
1.01
1.52
2.98
3.69
1.91
U.K
.1.
631.
201.
240.
901.
440.
872.
212.
193.
562.
77U
.S.
2.28
1.86
1.96
1.98
2.39
0.74
4.03
2.41
2.51
1.92
Tec
hnol
ogy
vs. C
ompe
titiv
enes
s
•E
cono
mis
ts c
are
abou
t ens
urin
g co
mpe
titio
n
•B
ut te
chno
logi
cal c
hang
e m
ore
impo
rtan
t in
med
ium
& lo
ng te
rm
•Su
ppos
e 10
% p
rice
dec
reas
e fr
om b
ette
rco
mpe
titio
n
•Sa
me
10%
dec
reas
e in
20
year
s by
0.5
%m
ore
prod
uctiv
ity g
row
th (
10.6
yea
rs b
y 1%
mor
e), &
the
chan
ge k
eeps
wor
king
Jose
ph S
chum
pete
r
•Po
inte
d to
pro
duct
ivity
gro
wth
, new
prod
ucts
, cau
sing
gro
wth
•L
arge
fir
ms
& m
onop
olie
s as
pri
mar
yso
urce
of
rapi
d gr
owth
•St
oppi
ng a
nti-
com
petit
ive
prac
tices
thus
may
irre
para
bly
dam
age
the
econ
omy!
•W
e w
ill d
iscu
ss m
ore
duri
ng th
e co
urse
UK Technological Progress
• Leadership to 1800 lost by 1900s
• Strong in services not manufacturing
• Strong in pharmaceuticals, militaryaerospace
• Large government budget to military– Benefits those industries
– But secrecy, use of good personnel
• Little education in eng., app. science
• Culture looks down on engineers
• Oxbridge old-boy network reinforces
• Modest government R&D funding,especially non-military
Walker (1993)E
cono
mic
Gro
wth
•y
= A
(t)
× ƒ(
k)–
y ou
tput
per
per
son
per
year
–A
(t)
tech
nolo
gy f
acto
r, ch
ange
s ov
er ti
me
t
–ƒ(
k) p
rodu
ctio
n fu
nctio
n, k
cap
ital p
er p
erso
n
•
–h
frac
tion
of o
utpu
t con
sum
ed, (
1-h)
inve
sted
–!
depr
ecia
tion
rate
dk dth
yk
=−
−(
)1
!
Prod
uctio
n ov
er T
ime
in G
row
th M
odel
with
No
Prod
uctiv
ity G
row
th
02
55
07
51
00
Yea
rs
05
10
Pro
du
ctio
n
Prod
uctio
n ov
er T
ime
in G
row
th M
odel
with
3%
Ann
ual P
rodu
ctiv
ity G
row
th
02
55
07
51
00
Yea
rs
05
10
Pro
du
ctio
n
0
91
4.0
6
18
27
.12
Pro
du
ctio
n
11
1
1
02
55
07
51
00
Yea
rs
Prod
uctio
n ov
er T
ime
in G
row
th M
odel
with
3%
Ann
ual P
rodu
ctiv
ity G
row
th
EC
2212
Ind
ustr
ial G
row
than
d C
ompe
titio
n
Lec
ture
2
Smal
l Fir
ms
Can
Be
Inno
vativ
e!
(and
Som
e W
ays
to H
elp)
Is N
ew T
echn
olog
y E
ver
from
Sm
all F
irm
s?•
Yes
:
•Je
wke
s, S
awer
s, &
Stil
lerm
an (
1959
), T
heSo
urce
s of
Inv
enti
on
•In
vent
ions
oft
en f
rom
sm
all f
irm
s,in
divi
dual
s
•Sm
all f
irm
s: c
ello
phan
e ta
pe, T
eryl
ene;
indi
vidu
als:
sel
f-w
indi
ng w
atch
, pen
icill
in
•T
rue
for
inve
ntio
n, le
ss f
or i
nnov
atio
n
Smal
l Fir
ms
Are
Inv
entiv
e
•A
bove
a m
inim
um s
ize,
larg
e an
d sm
all
firm
s ha
ve s
imila
r R
&D
spe
ndin
g pe
rem
ploy
ee
•Sm
all f
irm
s av
erag
e m
ore
pate
nts
per
R&
Ddo
llar
spen
t
•C
ohen
(19
95)
revi
ews
a la
rge
liter
atur
e on
the
subj
ect
Hel
ping
Sm
all F
irm
s In
nova
te
•U
se H
eter
ogen
ous
Skill
s
•Fi
t Ind
ustr
y-Sp
ecif
ic N
eeds
•L
ook
for
Inno
vativ
e Id
eas
Out
side
Fir
ms
•U
se R
egio
nal N
etw
orks
of
Inno
vato
rs
Het
erog
eneo
us S
kills
•Fi
rms
diff
er in
em
ploy
ee s
kills
, res
ourc
es
•H
arne
ss th
e sk
ills
& r
esou
rces
•E
.g.,
radi
o m
anuf
actu
rers
ent
erin
g T
V s
etpr
oduc
tion:
inno
vate
d 5+
x a
s m
uch,
60%
low
er a
nnua
l exi
t, hi
gher
mar
ket s
hare
(Kle
pper
& S
imon
s, 2
000
SMJ)
•E
.g.,
auto
mob
ile m
aker
s in
ear
ly 1
900s
Indu
stry
-Spe
cifi
c N
eeds
•A
mou
nts
& k
inds
of
R&
D n
eede
d va
ry b
yin
dust
ry
% o
f fi
rms
cond
uctin
g R
&D
(100
0-49
99 e
mpl
oyee
s)
Food
& k
indr
ed p
rodu
cts
50%
Petr
oleu
m69
.2%
Air
craf
t & p
arts
89.5
%
(Jew
kes,
Saw
ers,
Stil
lerm
an, 1
959,
p. 1
92)
Inno
vativ
e Id
eas
Out
side
Fir
ms
•In
nova
tive
idea
s of
ten
orig
inat
e ou
tsid
eco
mpa
nies
•Jo
hnst
on &
Gib
bons
(19
75),
von
Hip
pel
(198
8)
•E
.g.,
in N
MR
spe
ctro
met
ers
79%
of
maj
orin
nova
tions
fro
m u
sers
•O
ther
inno
vatio
ns f
rom
sup
plie
rs
Sour
ces
of I
nnov
atio
ns
Von
Hip
pel (
1988
, p. 4
)
Tec
hnol
ogy
Use
rM
anuf
.Su
ppli
erO
ther
Scie
ntif
ic in
stru
men
ts77
%23
%0%
0%S
emic
ondu
ctor
& P
CB
pro
cess
67%
21%
0%12
%Pu
ltrus
ion
proc
ess
90%
10%
0%0%
Tra
ctor
-sho
vel r
elat
ed6%
94%
0%0%
Eng
inee
ring
pla
stic
s10
%90
%0%
0%P
last
ics
addi
tives
8%92
%0%
0%In
dust
rial
gas
-usi
ng42
%17
%32
%8%
The
rmop
last
ics-
usin
g43
%14
%36
%7%
Wir
e te
rmin
atio
n eq
uipm
ent
11%
33%
56%
0%
Reg
iona
l Net
wor
ks•
Clu
ster
s of
fir
ms
in a
n in
dust
ry M
arsh
all
(189
0) p
oint
s ou
t cla
ssic
ben
efits
:–
inte
rmed
iate
goo
ds s
uppl
y
–la
bor
supp
ly
–kn
owle
dge
spill
over
•Pa
tent
s (e
tc.)
ref
lect
clu
ster
ing
bene
fits
•E
.g.,
tire
mak
ers:
89%
in A
kron
(66
% n
ot)
prod
uced
cor
d tir
es in
192
0, y
ield
ing
3x le
ssch
ance
of
exit
(Kle
pper
& S
imon
s, 2
000
JPE
)
Succ
essf
ul I
nnov
atio
n in
Agg
lom
erat
ions
Silic
on V
alle
y su
cces
ses
(Sax
enia
n, 1
994)
:
•In
terc
hang
e of
idea
s by
man
ager
s, e
ngin
eers
•R
apid
job
chan
ge
•Su
cces
sful
inve
stor
s fu
nd s
tart
-ups
•L
inks
with
uni
vers
ity r
esea
rch,
edu
catio
n
•L
ocal
gov
ernm
ent -
exe
cutiv
e co
oper
atio
n
•Fi
rms
shar
ing
reso
urce
s
Smal
l Fir
ms
Are
Inv
entiv
e(C
oncl
usio
n)•
Smal
l fir
ms
are
inve
ntiv
e–
Evi
denc
e fr
om c
ase
stud
ies,
cro
ss-s
ectio
n da
ta
–T
rue
for
inve
ntio
n, le
ss f
or in
nova
tion
•T
ailo
r R
&D
wor
k to
fir
m s
kills
, ind
ustr
ies
•G
et id
eas
from
sup
plie
rs a
nd c
usto
mer
s to
o
•U
se in
nova
tive
netw
orks
;m
ake
inno
vativ
e ne
twor
ks s
ucce
ssfu
l
EC
2212
Ind
ustr
ial G
row
than
d C
ompe
titio
n
Lec
ture
2 e
xtra
par
t
The
Eco
nom
y of
Citi
es:
the
Rol
e of
Inn
ovat
ion
EC
2212
Ind
ustr
ial G
row
than
d C
ompe
titio
n
Lec
ture
3
Lar
ge F
irm
s A
re I
nnov
ativ
e(a
nd w
hen)
Org
aniz
atio
n of
Lec
ture
1. L
arge
fir
ms
som
etim
es d
omin
ate
inno
vatio
nA
Sto
ry T
hat M
ust B
e To
ld:
2. S
chum
pete
r’s
view
: lar
ge b
ette
r
3. I
nves
tigat
ion
sugg
ests
: sm
all b
ette
r
4. T
heor
y re
inte
rpre
ts: l
arge
oft
en b
ette
r
5. L
arge
fir
ms
are
bette
r in
thes
e ci
rcum
stan
ces…
Lar
ge F
irm
s So
met
imes
Dom
inat
e In
nova
tion
% o
f M
ajor
Inn
ovat
ions
fro
m L
arge
st F
irm
s
Aut
os to
p 2
Tir
es to
p 4
prod
uct 4
3%pr
oduc
t 98%
proc
ess
95%
proc
ess
mos
t/all
Tel
evis
ions
top
5 Pe
nici
llin
top
4
prod
uct 8
4%pr
oduc
t 100
%
proc
ess
mos
t/all
proc
ess
mos
t/all
Sour
ce: D
ata
from
K. S
imon
s (s
ee K
lepp
er &
Sim
ons,
199
7).
Schu
mpe
ter
on I
nnov
atio
n (1
)
“[I]
n ca
pita
list r
ealit
y as
dis
tingu
ishe
d fr
om it
s te
xtbo
ok p
ictu
re, i
t is
not
[pri
ce]
com
petit
ion
whi
ch c
ount
s bu
t the
com
petit
ion
from
the
new
com
mod
ity, t
he n
ew te
chno
logy
, the
new
sou
rce
of s
uppl
y, th
e ne
w ty
peof
org
aniz
atio
n…—
com
petit
ion
whi
ch c
omm
ands
a d
ecis
ive
cost
or
qual
ity a
dvan
tage
and
whi
ch s
trik
es n
ot a
t the
mar
gins
of
the
prof
its a
ndth
e ou
tput
s of
the
exis
ting
firm
s bu
t at t
heir
fou
ndat
ions
and
thei
r ve
ryliv
es.
Thi
s ki
nd o
f co
mpe
titio
n is
as
muc
h m
ore
effe
ctiv
e th
an th
e ot
her
asa
bom
bard
men
t is
in c
ompa
riso
n w
ith f
orci
ng a
doo
r, a
nd s
o m
uch
mor
eim
port
ant t
hat i
t bec
omes
a m
atte
r of
com
para
tive
indi
ffer
ence
whe
ther
com
petit
ion
in th
e or
dina
ry s
ense
fun
ctio
ns m
ore
or le
ss p
rom
ptly
; the
pow
erfu
l lev
er th
an in
the
long
run
exp
ands
out
put a
nd b
ring
s do
wn
pric
esis
in a
ny c
ase
mad
e of
oth
er s
tuff
.”
(Cap
ital
ism
, Soc
iali
sm a
nd D
emoc
racy
, 3rd
ed.
, 195
0, p
p. 8
4-85
)
Schu
mpe
ter
on I
nnov
atio
n (2
)
Tra
ckin
g do
wn
sour
ces
of e
cono
mic
pro
gres
s, “
the
trai
l lea
ds…
pre
cise
lyto
the
door
s of
the
larg
e co
ncer
ns”
(195
0, p
. 82)
1. M
onop
oly
pric
es ju
stif
iabl
e as
insu
ranc
e ag
ains
t new
tech
nolo
gy.
2. M
onop
olis
tic
or o
ligop
olis
tic
beha
vior
not
so
bad
, vs.
ben
efit
fr
om te
chno
logi
cal c
hang
e. G
ood
(e.g
. sta
bilit
y in
rec
essi
ons)
.3.
Tec
hnol
ogy
dri
ves
dow
n pr
ices
any
way
; use
hed
onic
pri
ces.
4. F
irm
s ad
opt n
ew in
nova
tion
s d
espi
te a
ny “
loss
” in
cap
ital
.5a
. Tru
e m
onop
oly
is e
xcee
din
gly
rare
; lar
ge s
ize
is n
ot m
onop
oly.
5b. M
onop
olis
ts m
ay b
e m
ore
effi
cien
t, cr
eate
new
pro
duc
ts.
5c. S
hort
-run
mon
opol
y m
ore
com
mon
than
long
-run
mon
opol
y.6.
Rep
lace
sta
tic
perf
ect c
ompe
titi
on e
cono
mic
vie
w; n
eed
dyn
amic
u
nder
stan
din
g w
ith
tech
nolo
gy.
Tec
hnol
ogic
al c
hang
e ca
nnot
h
appe
n in
per
fect
com
peti
tion
, whi
ch d
isal
low
s pr
ofit
s fr
om
inno
vati
on.
Em
piri
cal S
tudi
es o
f In
nova
tion
R&
D in
tens
ity (
$/pe
rson
/yr.
)
firm
siz
e
R&
D s
pend
ing
($/y
r.) fir
m s
ize
R&
D e
ffici
ency
(va
lue/
R&
D$)
firm
siz
e
Fake
dat
a ill
ustr
ate
real
fin
ding
s:•
R&
D s
pend
ing
rise
s w
ith f
irm
siz
e•
R&
D in
tens
ity f
lat w
ith f
irm
siz
e
(Sc
here
r fi
nds
slig
ht in
vert
ed-U
)•
R&
D e
ffic
ienc
y fa
lls w
ith f
irm
siz
e•
I.e.
, sm
all f
irm
s sp
end
sam
e pe
r
em
ploy
ee a
nd g
et m
ore
resu
lts!
A T
heor
y of
R&
D a
nd F
irm
Siz
e
•Fa
cts
som
etim
es m
isle
ad, w
hen
not
info
rmed
by
corr
ect t
heor
y
•I
will
sho
w a
n ec
onom
ic th
eory
(si
mpl
eve
rsio
n of
Coh
en &
Kle
pper
, 199
6)
•A
ssum
ptio
ns o
f th
eory
fit
mos
t ind
ustr
ies
•T
he th
eory
fits
em
piri
cal f
acts
•T
he th
eory
impl
ies
larg
e fi
rms
bette
r
Ass
umpt
ions
of
The
ory
•Fi
rm s
pend
s R
on
rese
arch
& d
evel
opm
ent
•R
&D
incr
ease
s qu
ality
, low
ers
unit
cost
–(O
rex
pect
ed q
ualit
y an
d un
it co
st)
–I.
e., R
&D
incr
ease
s PC
= p
rice
- c
ost:
PC'(R
)>0
–D
imin
ishi
ng m
argi
nal r
etur
ns to
R: P
C''(
R)<
0
•Fi
rms
max
. pro
fit: #
= P
C(R
) ×
Q -
R
•Fu
ture
out
put Q
is f
ixed
; fir
m c
hoos
es R
Pric
e-C
ost M
argi
n ve
rsus
R
PC
rese
arch
spe
ndin
g R
pric
e -
cost
Why
Tw
o K
ey A
ssum
ptio
ns
•R
esea
rch
resu
lts c
anno
t be
sold
to o
ther
com
pani
es–
Ret
urns
to R
&D
rar
ely
can
be a
ppro
pria
ted
exce
pt th
roug
h a
com
pany
’s o
wn
sale
s
–Fi
rms
can
“inv
ent a
roun
d” p
aten
ts, u
se la
wsu
its
•Fu
ture
out
put Q
is a
fix
ed c
onst
ant
–E
.g.,
Qfu
ture
= Q
now
× k
–Fi
rm g
row
th k
is li
mite
d in
pra
ctic
e
Thr
ee I
mpl
icat
ions
of
the
The
ory
1. L
arge
fir
ms:
mor
e R
&D
than
sm
all f
irm
s, n
ot (
necc
.) m
ore
R&
D p
er u
nit o
f si
ze
2. L
arge
fir
ms:
hig
her
PC th
an s
mal
l fir
ms
–B
ette
r fo
r so
ciet
y (b
ette
r qu
ality
and
cos
t)
–B
ette
r co
mpe
titor
s (h
ighe
r pr
ofit
mar
gin)
3. L
arge
fir
ms:
less
R&
D r
esul
ts p
er R
&D
$–
Les
s w
aste
ful d
uplic
atio
n of
eas
y re
sear
ch
–D
o m
ore
chal
leng
ing
(mor
e m
argi
nal)
R&
D to
o
Proo
fs S
tep
0: A
Lem
ma
# # #=×
−
=$
×−
=
$=
=$$
×<
PCR
QR
d dRPC
RQ
PCR
Q
d dRPC
RQ
() (
)
()
()
10 1
02
2
i.e.,
The
fir
m m
axim
izes
#
by
choo
sing
R.
At t
he m
axim
um, d#
/dR
m
ust e
qual
0, a
nd d
2 #/d
R2
m
ust b
e ne
gativ
e.So
, the
fir
m c
hoos
es R
so
th
at th
e m
argi
nal i
ncre
ase
in
the
pric
e-co
st m
argi
n
fro
m a
noth
er $
of
R&
D
spe
ndin
g eq
uals
1/Q
;
that
is, P
C'(R
) =
1/Q
.
Proo
fs S
tep
1: T
heor
ems
1a &
1b
Let
:
The
n:
,
so
use
,
to
get
The
n: PC
Rk
R
PCR
kR
PCR
Q
k RQ
dR
QdQ
()
log(
)
()
/
()
/
(/
)
=
$=
$=
=
=1 1
0
PCR
Q
d dQPC
Rd dQ
Q
PCR
dR dQQ
dR dQQ
PCR
$=
$(
)=% &'
( )*
$$=−
=−$$
>
()
()
()
()1
1
1
10
2
2
Big
ger
firm
s do
mor
e R
&D
.
Big
ger
firm
s ar
e no
t nec
essa
rily
mor
e R
&D
inte
nsiv
e.
Proo
fs S
tep
2: T
heor
ems
2a &
2b
Sinc
e R
incr
ease
s w
ith f
irm
s
ize
(dR
/dQ
> 0
),an
d si
nce
PC in
crea
ses
with
R
,PC
mus
t inc
reas
e w
ith f
irm
s
ize: dP
CdQ
>0
Big
ger
firm
s ha
ve b
ette
rqu
ality
and
low
er u
nit c
ost.
Com
pare
siz
es Q
2 >
Q1.
Opt
imal
R&
D is
R2
> R
1.Su
ppos
e th
e la
rger
fir
m s
pend
s
onl
y R
1 on
R&
D.
The
n its
pro
fit p
er u
nit,
[
PC(R
1)×
Q2
– R
1] /
Q2
exce
eds
that
of
the
smal
ler
firm
,
[PC
(R1)
× Q
1 –
R1]
/ Q
1.C
hoos
ing
optim
al R
&D
R2
o
nly
incr
ease
s pr
ofit
per
unit.
Big
ger
firm
s ea
rn m
ore
prof
it pe
r un
it.T
hey
are
“fitt
er”
com
petit
ors.
Proo
f St
ep 3
: The
orem
3
For
any
conc
ave
incr
easi
ng
fun
ctio
n ƒ(
x), ∆
y/∆
x
is s
mal
ler
for
larg
er ∆
x.Si
nce
PC(R
) is
a c
onca
ve
incr
easi
ng f
unct
ion
(
PC'(R
) >
0, P
C''(
R)
< 0
),
dPC
RPC
RdR
()
()
/−
()
[] <
00
Big
ger
firm
s ac
com
plis
h le
ss u
nit-
cost
savi
ngs
per
aver
age
$ sp
ent o
n R
&D
.I.
e., t
hey
don’
t jus
t dup
licat
e sm
all f
irm
s’ R
&D
;th
ey d
o m
ore
chal
leng
ing
R&
D to
o.
ƒ(x)
x
{
{
∆x
∆y
Con
clus
ion
•Sc
hum
pete
r’s
stor
y ca
n fi
t the
evi
denc
e...
•W
hen:
–R
etur
ns to
R&
D a
re n
ot a
ppro
pria
ble
–Fu
ture
out
put i
s lim
ited
in p
ract
ice
–R
&D
is d
uplic
ated
acr
oss
firm
s
–I.
e., e
spec
ially
for
gra
dual
impr
ovem
ent,
not f
or p
aten
tabl
e ra
dica
l inv
entio
ns
•T
hen
larg
er f
irm
s do
mor
e in
tens
ive
R&
D,
are
fitte
r co
mpe
titor
s; s
ocie
ty b
enef
its m
ore
EC
2212
Ind
ustr
ial G
row
than
d C
ompe
titio
n
Lec
ture
4
Patte
rns
of I
ndus
try
Evo
lutio
n,
in p
art D
eter
min
ed b
y T
echn
olog
y
Def
initi
on o
f an
Ind
ustr
y
•D
iffe
rent
way
s to
def
ine
an in
dust
ry
•St
anda
rd in
dust
rial
cla
ssif
icat
ion
(SIC
)–
Cat
egor
ies
chos
en b
y go
vern
men
t
–U
sed
to a
naly
ze c
ensu
s da
ta o
n fi
rms
–SI
C c
odes
are
hie
rarc
hica
l
–D
ata
com
mon
ly a
t 2-
or 4
-dig
it le
vels
–Fi
rms
in S
IC c
lass
usu
ally
not
com
petit
ors
–So
com
petit
ive
dyna
mic
s ar
e hi
dden
SIC
Cod
e H
iera
rchy
•M
ajor
gro
ups:
–01
-09
Agr
icul
ture
, for
estr
y, f
ishi
ng
–20
-39
Man
ufac
turi
ng
–60
-67
Fina
nce,
insu
ranc
e, r
eal e
stat
e
•2-
digi
t gro
ups:
–20
Foo
d an
d ki
ndre
d pr
oduc
ts
–37
Tra
nspo
rtat
ion
equi
pmen
t
•4-
digi
t gro
ups:
–37
11 M
otor
e ve
hicl
es &
pas
seng
er c
ar b
odie
s
–37
24 A
ircr
aft e
ngin
es &
eng
ine
part
s
Com
petit
ive
Indu
stry
Def
initi
on
•Fi
rms
com
pete
: pro
duce
“sa
me”
pro
duct
•Pr
oduc
t var
iant
s su
bstit
utab
le f
or m
ost c
usto
mer
s
•E
xam
ple:
aut
omob
iles
–Sp
orts
car
s, s
mal
l car
s, 8
-sea
ters
— f
airl
y su
bstit
utab
le
–M
otor
cycl
es, t
ruck
s, b
uses
— s
erve
qui
te d
iffer
ent
need
s
•C
ould
def
ine
as, e
.g.,
subs
titut
abili
ty ≥
20%
•In
pra
ctic
e, g
uess
timat
e gi
ven
mar
ket k
now
ledg
e
Alte
rnat
ive
Indu
stry
Def
initi
ons
•In
dust
ries
def
ined
by:
–Su
bstit
utab
ility
(fo
r co
mpe
titio
n an
alys
is)
–Pr
ice
iden
tical
for
all
buye
rs (
for
mar
ket a
naly
sis)
–O
ligop
olie
s if
pre
sent
cou
ld in
flue
nce
pric
es (
for
antit
rust
ana
lysi
s)
–E
cono
mie
s of
sco
pe in
tech
nolo
gy o
f pr
oduc
t or
prod
uctio
n (f
or d
iver
sifi
catio
n an
alys
is)
•T
his
cour
se: m
ainl
y su
bstit
utab
ility
Patte
rns
of I
ndus
try
Evo
lutio
n
•T
raits
that
evo
lve:
sup
ply
and
dem
and
•Fo
cus
on s
uppl
y si
de
•T
ypol
ogy
of in
dust
ries
–C
lass
ify
indu
stri
es b
y ev
olut
iona
ry ty
pe?
–Sh
akeo
ut &
con
cent
ratio
n in
som
e in
dust
ries
;so
me
earl
y en
tran
ts d
omin
ate
–N
on-s
hake
out i
n ot
her
indu
stri
es;
som
etim
es m
arke
t lea
ders
are
rep
lace
d
–T
echn
olog
y: w
hat r
ole
does
it p
lay?
Supp
ly a
nd D
eman
d T
raits
•Su
pply
-sid
e–
Firm
s in
the
indu
stry
: ent
ry, e
xit
–T
echn
olog
y of
pro
duct
and
pro
duct
ion
–Fi
rm g
row
th, m
arke
t sha
re, c
once
ntra
tion
•D
eman
d-si
de–
Num
ber
of c
usto
mer
s &
indi
vidu
al d
eman
d
–T
ende
ncy
to s
tick
with
a b
rand
; sw
itchi
ng c
osts
–Sp
read
of
info
rmat
ion
and
fads
Shak
eout
in S
ome
Indu
stri
es
•G
ort &
Kle
pper
(19
82),
Kle
pper
& G
radd
y (1
990)
–47
new
pro
duct
s
–Fi
rst s
ever
al d
ecad
es o
r m
ore
of m
arke
t
–In
itial
ris
e in
num
ber
of p
rodu
cers
–T
hen
in m
ost p
rodu
cts
a sh
akeo
ut o
f pr
oduc
ers
–D
espi
te c
ontin
ued
grow
th in
out
put
•G
reat
er s
hake
out w
ith m
ore
tech
nolo
gica
l cha
nge
(Aga
rwal
, 199
8)
•So
me
earl
y en
tran
ts d
omin
ate
thro
ugh
tech
nolo
gy(K
lepp
er &
Sim
ons,
199
7, …
)
Industrial Growth and Competition K. Simons, 2001
69
Figure 6.1. Number of Manufacturers versus Time of Various US and UK Products
Fir
ms
Adding and Calculating Machines, US
0
10
20
30
40
50
60
1905 1925 1945 1965
Fir
ms
Adding and Calculating Machines, UK
0
10
20
30
40
50
60
1908 1928 1948 1968
Fir
ms
Television Receivers, UK
0
5
10
15
20
25
30
35
40
1935 1955 1975 1995
Fir
ms
Television Receivers, US
0
20
40
60
80
100
120
1947 1967 1987
Fir
ms
Pneumatic Automobile Tires, US
0
50
100
150
200
250
300
1905 1925 1945 1965
Fir
ms
Tires of All Types, UK
0
20
40
60
80
100
120
140
1906 1926 1946 1966 1986
Industrial Growth and Competition K. Simons, 2001
70
Figure 6.1 continued.
Fir
ms
Typewriters, UK
0
10
20
30
40
50
60
1903 1923 1943 1963 1983
Fir
ms
Typewriters, US
0
5
10
15
20
25
30
35
40
1905 1925 1945 1965
Fir
ms
Electric Blankets, UK
0
5
10
15
20
25
30
35
1946 1966 1986
Fir
ms
Electric Blankets, US
0
2
4
6
8
10
12
14
16
18
1915 1935 1955 1975
Fir
ms
Cathode Ray Tubes, UK
0
2
4
6
8
10
12
14
16
1950 1970
Fir
ms
TV Picture Tubes, US
0
10
20
30
40
50
60
70
1949 1969
Industrial Growth and Competition K. Simons, 2001
71
Figure 6.1 continued.F
irm
s
Ball-Point Pens, UK
0
5
10
15
20
25
30
1949 1969
Fir
ms
Ball-Point Pens, US
0
20
40
60
80
100
120
1946 1966
Fir
ms
Zippers, UK
0
10
20
30
40
50
60
1933 1953 1973
Fir
ms
Zippers, US
0
10
20
30
40
50
60
70
80
90
1928 1948 1968
Fir
ms
Photocopy Machinery, UK
0
10
20
30
40
50
60
70
1954 1974
Fir
ms
Photocopy Machinery, US
0
5
10
15
20
25
30
35
1939 1959 1979
Industrial Growth and Competition K. Simons, 2001
72
Figure 6.1 continued.
Fir
ms
Art ificial Christmas Trees, UK
0
1
2
3
4
5
6
1949 1969
Fir
ms
Art ificial Christmas Trees, US
0
5
10
15
20
25
1924 1944 1964
Fir
ms
Freezers, UK
0
1
2
3
4
5
6
1949 1969
Fir
ms
Freezers for Home and Farm, US
0
10
20
30
40
50
60
70
1944 1964
Fir
ms
Phonograph Records, UK
0
5
10
15
20
25
30
35
40
45
50
1912 1932 1952 1972
Fir
ms
Phonograph Records, US
0
10
20
30
40
50
60
1905 1925 1945 1965
No
Shak
eout
in S
ome
Indu
stri
es
•N
umbe
r of
fir
ms
may
fal
l non
e or
littl
e–
Bal
l-po
int p
ens
in th
e U
S
–Z
ippe
rs in
the
US
•N
o ea
rly-
mov
er a
dvan
tage
, unl
ike
insh
akeo
uts
•In
oth
er c
ases
, lat
e-m
over
adv
anta
ges
–M
echa
nica
l to
elec
tron
ic c
alcu
lato
rs
The
Rol
e of
Tec
hnol
ogy
•T
echn
olog
y is
a p
ower
ful c
ompe
titiv
e fo
rce
•If
(1)
larg
e fi
rms
have
mos
t R&
D in
cent
ive,
and
(2)
tech
nolo
gy f
its f
irm
s’ a
bilit
ies,
then
:te
chno
logy
rei
nfor
ces
the
lead
ing
firm
s
•If
(1)
R&
D m
atte
rs li
ttle
to c
ompe
titio
n, o
r(2
) ne
w f
irm
s be
tter
at r
elev
ant R
&D
, the
n:le
adin
g fi
rms
have
no
adva
ntag
e
A V
alid
ity C
heck
•D
o te
chno
logy
and
mar
ket t
raits
rea
lly d
eter
min
ein
dust
ry e
volu
tion?
•O
r ar
e ou
tcom
es ju
st r
ando
m?
•If
just
ran
dom
, sho
uld
ofte
n di
ffer
acr
oss
coun
trie
s
•Si
mon
s (2
000)
com
pare
s 18
pro
duct
s in
US
& U
K
•D
egre
e an
d tim
e of
sha
keou
t ver
y si
mila
r
•C
omm
on p
roce
ss a
t wor
k in
bot
h na
tions
•N
ot ju
st r
ando
m
Con
clus
ion
•T
echn
olog
y, o
ther
mar
ket t
raits
, aff
ect i
ndus
try
evol
utio
n
•T
ypes
of
indu
stry
evo
lutio
n:–
Mar
ket l
eade
rshi
p tu
rnov
er: a
ided
by
disr
uptiv
ete
chno
logy
–C
ontin
ued
entr
y &
exi
t pos
sibl
e: te
chno
logy
lead
ersh
ipis
uni
mpo
rtan
t
–Sh
akeo
ut: m
arke
t lea
ders
use
tech
nolo
gy to
dom
inat
e,no
n-di
srup
tive
tech
nolo
gy
–M
arke
t nic
hes
prot
ect f
irm
s
EC
2212
Ind
ustr
ial G
row
than
d C
ompe
titio
n
Lec
ture
5
Inno
vato
rs c
an u
se r
adic
al n
ewte
chno
logy
to s
eize
mar
kets
.Turnover of Corporate
Leadership
• New firms replace old as marketleaders.
• Entry and exit may occur.
• Major innovations can providethe means for market turnover.
A Technology Mudslide?Hypothesis: “Coping with the relentlessonslaught of technology change was akin totrying to climb a mudslide raging down a hill.You have to scramble with everything you’vegot to stay on top of it, and if you ever once stopto catch your breath, you get buried.” Leadingfirms frequently lose.
Reality: “The established firms were the leadinginnovators not just in developing risky, complex,and expensive component technologies…, but inliterally every… one of the sustaininginnovations in the industry’s history…. [T]hispattern of technology leadership… is stunninglyconsistent…. [T]here have been only a few ofthe other sort of technological change, calleddisruptive technologies…, that toppled theindustry’s leaders.”
– C. Christensen on hard disk drives
Examplescement - rotary kiln & coal dust ~1892,
suspension preheating 1972
minicomputers - solid state, ICs 1960s
container glass - semiautomation 1893,Owens machine 1903
flat glass - drawing machines 1917,continuous forming 1923, float glass 1963
aligners for semiconductor mfg.- contact / proximity / scanner /step &repeat
hard disk drives - 14", 8", 5.25", 3.5"
transistor - vs. vacuum tube
electronic calculator - vs. mechanical
Analyzing Turnover• Innovators: new firms or old?
E.g., powdered coal for cementNew firms - 4 of 5 innovatorsOld firms - 1 of 5
• Era of fermentProduct/process diversitySales growth, hard to forecast
• Entry & exit, market sharesEntry & exit rise, perhaps
entry/exit ratio risesDispersion in market shares,
turnover
• Tushman & Anderson article.
Why Incumbents Lose4 Explanations
• New technology makes obsoleteprevious core competency.
• Blinded by prior technologicalviews & organizational filters.
• Blinded by customer base.
• Differing incentives:incumbents & entrants,existing & new markets.
• See: Tushman & Anderson, Henderson &Clark, Christensen & Rosenbloom
Non-TechnologicalLeadership Turnover
• Non-technological reasons forturnover may also exist.
• Targeting a new market:UK potato crisps: Golden
Wonder.US ball-point pens: Bic.(Both in part involved technical
innovation, but redefinitionof the market was key.)
Opportunities to Imitate
• Small firms pioneer newmarkets- Beat them while they’re weak.
• Patents absent or can becircumvented- Show previous patents or usage- Come up with alternative design for same
purpose- Defensive & offensive use of previous
patents
• Related experience
• New market segments to create
Success with Imitation
• Anticipate unlikely threats
• Concurrent R&D
• Legal & regulatory challengesto the pioneer(s)
• Enter quickly after the markethas formed, not after 1st entry
• Don’t copy too closely
• Continuity for consumers, nottoo much change
Defense againstImitation
• Sell out
• Licensing & joint ventures
• Fight off copycats:- Legal measures (copyright /
patent)- Introduce low-end generics
(or focus on high-value end)- Perpetually innovate- Create a proprietary standard
You
Sho
uld
Kno
w•
Wha
t kin
ds o
f te
chno
logy
may
cau
se m
arke
tle
ader
ship
turn
over
? W
hy?
•E
xam
ples
•Pr
edic
tions
for
the
mar
ket (
Tus
hman
& A
nder
son)
•W
hy d
on’t
mar
ket l
eade
rs in
nova
te w
ith d
isru
ptiv
ete
chno
logi
es?
4 th
eori
es
•W
hat s
trat
egie
s ca
n ai
d su
cces
sful
inno
vatio
n?
•W
hat s
trat
egie
s he
lp d
efen
d m
arke
t lea
ders
hip?
EC
2212
Ind
ustr
ial G
row
than
d C
ompe
titio
n
Lec
ture
6
Sust
aini
ng te
chno
logy
lets
ski
lled
earl
y en
tran
ts d
estr
oy c
ompe
titor
s.
Shak
eout
s•
New
pro
duct
s: o
ften
ris
e th
en f
all i
n nu
mbe
r of
prod
ucer
s
•Fa
ll in
num
ber
of p
rodu
cers
oft
en c
alle
d a
“sha
keou
t”
•M
ost p
rodu
cts
have
sha
keou
ts, w
ithi
n 3+
dec
ades
of w
hen
the
mar
ket f
orm
s
•C
an b
e ve
ry d
ram
atic
: US
auto
mob
iles
wen
t fro
m27
3 pr
oduc
ers
to 5
•C
once
ntra
ted
mar
ket s
hare
s te
nd to
res
ult
1894 1912 1930 1948 1966
0
55
110
165
220
275firms
exit
entry
US Automobile Producers, 1896-1966
Why
Do
Shak
eout
s H
appe
n?
•Fo
cus
on th
e m
ain
caus
e
•…
in p
rodu
cts
wit
h se
vere
sha
keou
ts
•Y
ou w
ill s
ee–
Evi
denc
e on
ent
ry a
nd e
xit
–T
heor
y th
at b
est f
its th
e fa
cts
•ap
prox
imat
ely,
Kle
pper
(19
96, 2
001)
–E
vide
nce
on e
arly
-mov
ers,
tech
nolo
gy
•T
hen
disc
uss
ram
ific
atio
ns
1894 1912 1930 1948 1966
0
55
110
165
220
275
0%
32%
Automobiles
firms
exit
entry
%exit
5-yearmovingaverage
1905 1920 1935 1950 1965 1980
0
55
110
165
220
275
0%
25%
Tires
firms
exit
entry
%exit
5-yearmovingaverage
1945 1956 1967 1978 19890
22
44
66
88
110
0%
24%
Televisions
firms
exit
entry
%exit
5-yearmovingaverage
1942 1952 1962 1972 1982 1992
0
6
12
18
24
30
0%
29%
Penicillin
firms
exit
entry
%exit
5-yearmovingaverage
Firms, Entry, Exit in Four Products (US)
1945 1952 1959 1966 1973 1980
$0
$400
$800
$1,200
0
4
8
12
millionsB&W
priceproduced
Televisions
$0
$400
$800
$1,200
0
4
8
12millionsColor
priceproduced
0
2
4
6
1945 1955 1965 1975
0
4
8
12Penicillin
price (10 $/ lb.)3 prod’n (10 lbs.)6
$1,500
$3,000
$4,500
$0
1900 1910 1920 1930 1940
0
1.5
3
4.5Automobiles
price millionssold
0
100
200
300
1910 1919 1928 1937
0
20
40
60
80
price
millions
Tires
sold
index
Price and Output in the Four Products
Exp
lana
tion
of S
hake
outs
Part
1 o
f 3
Ent
rant
s an
d T
heir
Ski
lls
Mor
e Sk
illed
Fir
ms
Can
Ear
n M
ore
Prof
it
low
high
com
pete
nce
at R
&D
% o
fpo
ten
tial
entr
ant
s
Pote
ntia
l En
tran
ts i
n Ye
ar X S
hade
dre
gion
: fir
ms
have
eno
ugh
skill
to
ear
npr
ofit
> 0
aft
er
entr
y
Ent
ry &
Gro
wth
Dri
ve D
own
Pric
e
•L
imite
d nu
mbe
r of
fir
ms
have
ski
lls n
eede
dto
ent
er, a
t any
poi
nt in
tim
e
•E
ach
year
som
e nu
mbe
r of
fir
ms
can
ente
r
•Fi
rms
ente
r fa
irly
sm
all,
but t
hen
grow
•E
ntry
and
gro
wth
incr
ease
tota
l out
put
•M
ore
outp
ut, l
ower
pri
ce (
dem
and
curv
e)
Skill
Nee
ded
to E
nter
Ris
es o
ver
Tim
e
low
high
com
pete
nce
at R
&D
% o
fpo
ten
tial
entr
ant
s
Pote
ntia
l En
tran
ts i
n Ye
ar X S
hade
dre
gion
: fir
ms
have
eno
ugh
skill
to
ear
npr
ofit
> 0
aft
er
entr
y
With
low
er p
rice
, nee
dm
ore
skill
to e
arn
a pr
ofit
Ent
ry E
vent
ually
Sto
ps
low
high
skill
low
skill
high
low
high
skill
%Ent
rant
s ne
ed in
crea
sing
ski
ll to
ear
n pr
ofit
> 0
, sin
ce p
rice
falls
Tim
e1
23
EX
AM
PLE
:60
pot
entia
l firm
s40
ent
er30
0 po
tent
ial f
irms
70 e
nter
800
pote
ntia
l firm
s0
ente
r
May
be
mor
e po
tent
ial e
ntra
nts,
but
eve
ntua
lly n
o en
tran
ts
Exp
lana
tion
of S
hake
outs
Part
2 o
f 3
R&
D, S
ize,
and
Pro
fit
R&
D w
ith I
mita
tion
•R
&D
impr
oves
qua
lity,
low
ers
cost
•D
ecre
asin
g re
turn
s to
R&
D
•C
ost-
per-
unit-
of q
ualit
y c
= c
(R),
c'<
0,c'
'>0
•Fi
rms
bene
fit f
rom
R&
D d
urin
g 1
time
peri
od
•Fi
rms
imita
te a
ll pa
st in
nova
tions
in th
ene
xt p
erio
d
Firm
i’s
Prof
it at
Tim
e t
•p t
pri
ce p
er u
nit o
f qu
ality
, pt =
ƒ(∑
Qit)
•[c
t – s
ic(R
it)]
cost
per
uni
t pro
duce
d–
c t h
ighe
st p
ossi
ble
cost
giv
en im
itatio
n of
pas
t R&
D
–s i
fir
m i’
s sk
ill a
t R&
D
–c(
Rit)
cos
t dec
reas
es w
ith c
urre
nt R
&D
, c'<
0, c
''>0
•Q
itou
tput
pro
duce
d
•R
itsp
endi
ng o
n R
&D
•g(
Qit
– Q
it-1)
cos
t of
grow
th, g
'>0,
g''>
0
#it
tt
iit
itit
itit
pc
sc
RQ
Rg
=−
−[
](
)−
−−
−(
)(
)1
Impl
icat
ions
of
the
Prof
it Fu
nctio
n
•Fi
rms
choo
se R
it, Q
it to
max
imiz
e pr
ofit
•L
arge
r fi
rms
spen
d m
ore
on R
&D
–Sp
read
cos
t of
R&
D o
ver
mor
e ou
tput
–R
emem
ber
lect
ure
3
•G
row
th is
lim
ited
–Fi
rms
grow
eac
h pe
riod
–In
crea
sing
mar
gina
l cos
t lim
its
grow
th
•Si
ze (
Qit)
and
ski
ll (s
i) en
hanc
e pr
ofit
Exp
lana
tion
of S
hake
outs
Part
3 o
f 3
Exi
t (gi
ven
Size
and
Ski
ll)
Who
Exi
ts W
hen?
•Fi
rms
exit
if #
it <
0
•G
row
th c
ause
s ex
it at
eve
ry t
–G
row
th.
/Q
it.
pt.
prof
it
•E
xitin
g fi
rms
are
smal
lest
, lea
st-s
kille
d–
Sinc
e si
ze a
nd s
kill
enh
ance
pro
fit
•E
arlie
r en
tran
ts a
re la
rger
, cet
eris
par
ibus
–H
ave
had
mor
e tim
e an
d in
cent
ive
to g
row
•Sk
illed
ear
ly e
ntra
nts
are
long
-run
sur
vivo
rs
Sum
mar
y in
Cou
rse
Not
es, p
. 89
low
high
com
pet
ence
of
man
ager
s
% o
fpo
tent
ial
entr
ants
inco
mp
eten
t:hi
gher
cos
ts
com
pete
nt:
low
er
cost
s
(Pot
ent
ial)
Ent
rant
s 1
895
-19
04
low
high
com
pet
ence
of
man
ager
s
% o
fp
oten
tial
entr
an
ts
(Pot
enti
al)
Entr
ants
19
05
-19
09
low
high
com
pete
nce
of
man
ager
s
% o
fp
oten
tial
entr
an
ts
(Pot
enti
al)
Entr
an
ts1
91
0-1
91
6
(By
mid
-19
20
s,en
try
beco
mes
impo
ssib
le.)
Ho
wbi
gar
efi
rms
that
ente
red
in1
89
5-1
90
4?
How
big
... e
nter
ed i
n 1
905
-190
9?
How
big
... e
nter
ed i
n 1
91
0-1
91
6?
circ
a1
90
4
circ
a1
90
9
circ
a1
91
6
sma
ll
med
ium
(b
ut 8
0%
hav
e ex
ited
)
larg
e (b
ut 9
0%
hav
e ex
ited
)
smal
l
med
ium
(bu
t 8
0%
hav
e ex
ited
)sm
all
Firm
s al
way
s en
ter
at s
mal
l si
zes.
As
tim
e go
es o
n, s
urvi
ving
firm
s gr
ow
.A
t an
y po
int
in t
ime,
ear
lier
entr
ants
are
lar
ger
th
an l
ater
ent
ran
ts.
Siz
e an
d co
mp
eten
ce r
educ
e a
firm
's c
osts
. B
ecau
se o
fsu
rviv
al o
f th
e fi
ttes
t, f
irm
s in
eac
h g
rou
p ar
e fo
rced
out
unti
l on
ly c
ompe
tent
ear
ly e
ntra
nts
rem
ain.
Impl
icat
ions
of
the
The
ory
•S
hake
out
–E
ntry
eve
ntua
lly s
tops
–E
xit c
ontin
ues
fore
ver,
cau
sing
sha
keou
t
•E
arlie
r en
tran
ts h
ave
low
er c
hanc
e of
exi
t–
May
be n
ot a
t fir
st (
depe
nds
on s
kill
dist
ribu
tion)
–B
ut e
vent
ually
eve
n hi
gh-s
kille
d la
te e
ntra
nts
exit
•E
arlie
r en
tran
ts d
o m
ore
R&
D
•Fi
rms
succ
essf
ul a
t R&
D s
urvi
ve b
ette
r
% S
urvi
val b
y E
ntry
Dat
e of
Aut
omob
ile P
rodu
cers
01
42
84
25
67
0
100%
0.1
%
Yea
rs
of
Pro
duct
ion
Firm
s S
urvi
ving
Ent
rant
s in
18
95-1
904
19
05
-09
19
10
-66
31.6
%
10.0
%
3.2
%
1.0
%
0.3
%
S1
S2
014
2842
5670
100% 0.1%
Aut
omob
iles
Yea
rs o
f P
rodu
ctio
n
Firm
s S
urvi
ving
Ent
rant
s in
189
5-19
04
19
05
-09
19
10
-66
31.6
%
10.0
%
3.2%
1.0%
0.3%
S1
S2
020
4060
80
100%
31.6%
10.0% 3.2%
1.0%
Tire
s
Yea
rs o
f P
rodu
ctio
n
Firm
s S
urvi
ving
Ent
rant
s in
190
1-19
06
19
07
-16
19
17
-22
19
23
+
S1
S2
S3
09
1827
3645
100%
31.6%
10.0% 3.2%
1.0%
Tel
evis
ions
Yea
rs o
f P
rodu
ctio
n
Firm
s S
urvi
ving
Ent
rant
s in
194
6-19
48
19
49
-19
51
19
52
+
S2
S1
010
2030
4050
100%
31.6%
10.0% 3.2%
1.0%
Pen
icilli
n
Yea
rs o
f P
rodu
ctio
n
Firm
s S
urvi
ving
Ent
rant
s in
194
3
19
45
-52
19
53
+
S1
S2
% S
urvi
val b
y E
ntry
Dat
e in
the
Four
Pro
duct
s
07
14
21
28
35
10
0%
31
.6%
10
.0%
3.2
%
1.0
%
Pe
ns,
B
allp
oin
t
% S
urvi
val b
y E
ntry
Dat
e in
a N
on-S
hake
out P
rodu
ct
Ent
rant
s 19
46-5
0
1951
-54
1955
-67
1968
-79
Yea
rs o
f pr
od
uct
ion
Inno
vatio
n, %
Ado
ptio
n, b
y E
ntry
Tim
e in
the
Four
Pro
duct
s
Use
sam
e en
try-
time
coho
rts
as p
revi
ousl
y, b
ut d
ivid
e tir
es c
ohor
t 1
Prod
uct
Inno
vati
on ty
peC
ohor
t 1C
ohor
t 2C
ohor
t 3A
utom
obil
esPr
oduc
t9
21
Aut
omob
iles
Proc
ess
30
0.1
Tir
esPr
oduc
t1
00
Tir
esC
ord
1917
36%
8%T
ires
Cor
d 19
2010
0%73
%62
%T
ires
Bal
loon
192
363
%16
%7%
Tel
evis
ions
Prod
uct
21
0T
elev
isio
nsPr
oces
s63
70
Peni
cilli
nPr
oces
s5
00
Rel
ativ
e in
nova
tion
rate
s by
pro
duct
& in
nova
tion
type
— c
ompa
re c
ohor
ts
0
.28
.21
.14
.07
Au
tom
ob
ile
s
1895
1967
1949
1931
1913
Sm
oot
hed
Haz
ard
No
n-In
nova
tors
Inno
vato
rs
Inno
vatio
n an
d E
xit
0
.28
.21
.14
.07
Au
tom
ob
ile
s
1895
1967
1949
1931
1913
Sm
ooth
ed H
azar
d
No
n-I
nno
vato
rs
Inno
vato
rs
0
.24
Tir
es
1920
1980
Sm
ooth
ed H
azar
d
No
n-I
nno
vato
rs
Cor
d &
Bal
loon
T
ires
Co
rd T
ires
.06
.12
.18
1965
1950
1935
0
.24
1952
1980
Sm
oot
hed
Haz
ard
No
n-In
nova
tors
Te
lev
isio
ns
Inno
vato
rs
1966
1973
1959
0
.36
.18
.27
.09
Pe
nic
illi
n
1959
1992
Sm
ooth
ed H
azar
d
No
n-In
nova
tors
1983
.75
1975
.519
67.2
5
Sem
isyn
thet
ic P
rodu
cers
Inno
vatio
n an
d E
xit i
n th
e Fo
ur P
rodu
cts
Ram
ific
atio
ns o
f Sh
akeo
uts
•In
indu
stri
es w
ith s
tron
g su
stai
ning
R&
D
•H
igh-
skill
ed e
arly
ent
rant
s do
min
ate
•O
ther
fir
ms
may
pro
fit f
or a
whi
le–
But
eve
ntua
lly f
orce
d to
exi
t
•E
nter
ear
ly, k
eep
up w
ith R
&D
, to
surv
ive
•C
once
ntra
tion
is a
nat
ural
res
ult
–A
nti-
trus
t aut
hori
ties
oft
en in
vest
ifat
e
–B
ut e
xpec
t con
cent
rati
on w
ith
lega
l beh
avio
r
EC
2212
Ind
ustr
ial G
row
than
d C
ompe
titio
n
Lec
ture
7
Oth
er f
acto
rs m
ay c
ontr
ibut
e to
conc
entr
atio
n an
d sh
akeo
uts.
Forc
es C
ausi
ng C
once
ntra
tion
•Pr
evio
us le
ctur
e ex
plai
ned
mai
n ca
use
ofsh
akeo
uts
•O
ther
for
ces
may
be
at w
ork
sim
ulta
neou
sly
insh
akeo
uts,
or
othe
rwis
e af
fect
con
cent
rati
on
•W
hy k
now
abo
ut th
ese
othe
r fo
rces
?–
Com
mon
them
es; p
eopl
e ex
pect
you
to k
now
–A
ffec
t com
petit
ion
in c
erta
in s
ituat
ions
–U
nder
stan
d m
ore
of h
ow te
chno
logy
wor
ks
•Si
x fo
rces
•St
atic
long
-run
: cos
t-sp
read
ing
& c
once
ntra
tion
1. F
irst
- &
Ear
ly-M
over
Adv
anta
ge
•Pr
evio
us le
ctur
e: E
arly
ent
rant
s gr
ow la
rger
,sp
read
R&
D c
ost o
ver
mor
e un
its
•O
ther
pos
sibl
e ea
rly-
mov
er a
dvan
tage
s:–
Win
rac
e fo
r pa
tent
(s)
[But
sim
ple
race
s ar
e ra
re]
–R
eput
atio
n, c
usto
mer
loya
lty, s
witc
hing
cos
ts [
But
big
qual
ity/p
rice
dif
fere
nces
und
erm
ine;
lim
ited
outp
utm
akes
less
rel
evan
t]
–W
ill d
iscu
ss c
ost-
redu
ctio
n, lo
ck-i
n, n
etw
orks
sho
rtly
–…
. 2. E
ffic
ient
Pro
duct
ion
Scal
e
•L
owes
t uni
t cos
t at a
min
imum
eff
icie
nt s
cale
:ou
tput
of
1 m
ost-
effi
cien
t mac
hine
•Sa
me
cost
for
mor
e ou
tput
: 2+
mac
hine
s
•In
crea
sing
ret
urns
to s
cale
belo
w m
inim
um
•C
onst
ant r
etur
ns to
sca
leab
ove
min
imum
min
imum
effi
cien
t sca
le oupu
t
aver
age
unit
cost
Oth
er S
cale
/Sco
pe A
dvan
tage
s
•R
&D
cos
t-sp
read
ing
•A
dver
tisi
ng c
ost-
spre
adin
g
•D
istr
ibut
ion
netw
orks
•M
anag
eria
l eff
icie
ncy
and
inef
fici
ency
3. P
rogr
essi
ve C
ost R
educ
tion
•O
ften
cal
led
lear
ning
cur
ves
–O
rex
peri
ence
cur
ves
–Im
plie
s w
orke
rs le
arn
to b
e m
ore
effi
cien
t
–B
ut o
ften
wor
kers
bec
ome
skill
ed q
uick
ly (
2-4
wee
ks in
TV
set
ass
embl
y), c
an m
ove
betw
een
jobs
•St
udie
s of
pro
gres
sive
cos
t red
ucti
on d
isti
ngui
shsp
ecif
ic s
ourc
es
•R
&D
a k
ey s
ourc
e of
pro
gres
sive
cos
t red
ucti
on
4. T
echn
olog
y L
ock-
In
•O
ne te
chno
logi
cal s
tand
ard
inst
ead
of a
noth
er
•H
ard
to c
hang
e (u
ser
netw
orks
, dev
elop
men
t cos
t)
•M
ay b
e an
infe
rior
tech
nolo
gy
•E
xam
ples
:–
QW
ER
TY
type
wri
ter
keyb
oard
(vs
Dvo
rjak
)
–V
HS
vide
ocas
sette
rec
orde
rs (
vs B
etam
ax)
•A
rgum
ents
ove
r w
heth
er in
feri
or te
chno
logy
lock
ed in
–L
iebo
witz
& M
argo
lis c
ontr
adic
t Dav
id r
e. Q
WE
RT
Y
5. N
etw
ork
Eco
nom
ies
•N
etw
orks
of
user
s be
nefi
t fro
m e
ach
othe
r
•H
ard
to s
witc
h st
anda
rds
once
the
netw
ork
ises
tabl
ishe
d
•C
ause
s on
e ty
pe o
f lo
ck-i
n
•E
xam
ples
–Q
WE
RT
Y, V
HS
–M
icro
soft
Wor
d
–R
ailr
oad
trac
k w
idth
s
6. S
unk
Cos
ts
•E
nter
ing
an in
dust
ry h
as a
cos
t
•M
ay b
e fo
r te
chno
logy
dev
elop
men
t,pu
rcha
se o
f m
achi
nes,
etc
.
•Fe
w f
irm
s en
ter
if th
e su
nk c
ost i
s ve
ryla
rge
•B
ecau
se th
e co
st m
ust b
e sp
read
ove
r th
enu
mbe
r of
uni
ts p
rodu
ced
A S
tatic
Lon
g-R
un V
iew
•Su
tton
(199
1, 1
998)
•A
rgue
s w
e m
ust l
ook
acro
ss d
iffe
rent
com
petit
ive
mod
els:
–B
ertr
and
vs C
ourn
ot v
s m
onop
oly
–P
rodu
ct d
iffe
rent
iati
on
•Pu
tsbo
unds
on
com
petit
ive
outc
omes
,in
stea
d of
pre
dict
ing
the
outc
omes
With
out C
ost-
Spre
adin
g
conc
entr
atio
nhi
gh
(100
%by
1 f
irm
)
low
(ev
enly
dist
ribut
ed,
man
y fir
ms)
mar
ket s
ize
/ min
imum
effi
cien
t sca
le
indu
stry
con
cent
rati
on in
the
gre
y re
gio
n
With
Cos
t-Sp
read
ing
mar
ket s
ize
/ min
imum
effi
cien
t sca
le
indu
stry
con
cent
rati
on in
the
gre
y re
gio
n
conc
entr
atio
nhi
gh
(10
0%
by 1
fir
m)
low
(ev
enly
dist
ribu
ted,
man
y fir
ms)
EC
2212
Ind
ustr
ial G
row
than
d C
ompe
titio
n
Lec
ture
8
Prod
uct d
iffe
rent
iatio
n ca
n pr
otec
tag
ains
t com
peti
tive
dest
ruct
ion.
Prot
ectio
n by
Dif
fere
ntia
tion
•Fi
rms
prod
uce
som
ethi
ng d
iffe
rent
fro
m r
ival
s
•If
pro
duct
dif
fere
nt e
noug
h th
at c
ross
-pri
ceel
asti
city
of
dem
and
near
zer
o, n
o co
mpe
titio
n–
Cal
led
a di
ffer
ent p
rodu
ct in
dust
ry in
this
cou
rse
–E
.g.,
bicy
cles
vs.
aut
omob
iles
or v
s. a
pple
s
•D
iffe
rent
iati
on th
eori
es a
re f
or in
term
edia
te c
ases
–Fi
rms
com
pete
with
eac
h ot
her
–B
ut d
iffe
rent
iatio
n le
ssen
s co
mpe
titio
n
•T
echn
olog
y: o
ne w
ay to
ach
ieve
dif
fere
ntia
tion,
tech
. pro
gres
s m
ay d
iffe
r ac
ross
sub
-mar
kets
Eco
nom
ic T
heor
ies
ofD
iffe
rent
iatio
n•
Typ
es o
f di
ffer
entia
tion
:–
Hor
izon
tal:
diff
eren
t fea
ture
s
–V
erti
cal:
sam
e fe
atur
es, d
iffe
rent
qua
lity
leve
ls
•A
bstr
act a
way
fro
m te
chno
logi
cal c
hang
e–
We
just
stu
died
theo
ries
with
ver
tical
and
hor
izon
tal
diff
eren
tiatio
n ca
used
by
tech
nolo
gica
l cha
nge
–Fo
r no
w, a
naly
ze d
iffe
rent
iatio
n w
ith n
o te
ch. c
hang
e,or
with
fir
ms’
cur
rent
tech
nolo
gies
Eco
nom
ic T
heor
ies
Con
t.
•D
efin
e “p
rodu
ct d
iffe
rent
iatio
n sp
ace”
:–
Lin
e se
gmen
t, e.
g. s
elle
rs a
long
a b
each
, car
col
ors
–C
ircl
e, e
.g. b
each
aro
und
a la
ke, s
easo
ns o
f th
e ye
ar
–2-
D s
quar
e or
cir
cle,
e.g
. loc
atio
ns in
a c
ity, f
ertil
izer
s
–2-
D s
urfa
ce o
f a
sphe
re, e
.g. l
ocat
ion
on E
arth
•D
efin
e ho
w c
ompe
titio
n ha
ppen
s–
Cus
tom
ers
deci
de w
here
to g
o
–Pa
y a
“tra
nspo
rtat
ion
cost
” to
“ge
t the
re”
–M
ake
purc
hasi
ng d
ecis
ions
acc
ordi
ngly
Eco
nom
ic T
heor
ies
Con
t.
Dri
nks
stan
ds a
long
a b
each
Cus
tom
ers
even
ly s
prea
d ou
t alo
ng th
e be
ach
The
y w
alk
to n
eare
st s
tand
, if p
rice
is s
ame
Ass
ume
pric
e is
sam
e, f
or s
impl
icitl
yW
here
do
you
loca
te?
-
how
man
y fi
rms?
-
near
eac
h ot
her
or s
prea
d ap
art?
- le
apfr
oggi
ng?
May
be
dist
ance
alo
ng b
each
, or
pref
eren
ce d
ista
nce
(e.g
. car
col
ors)
An
Eco
logi
cal T
heor
y of
Dif
fere
ntia
tion
•H
anna
n &
Fre
eman
(19
89)
borr
ow f
rom
eco
logy
•Fi
rst,
unde
rsta
nd s
impl
e lo
gist
ic g
row
th m
odel
:
– –N
num
ber
of f
irm
s, t
time,
K c
arry
ing
capa
city
,r
grow
th r
ate
–N
umbe
r of
fir
ms
grow
s un
til r
each
ing
carr
ying
cap
acity
–T
hink
of
N a
s po
pula
tion,
r(K
–N)/
K a
s bi
rth-
deat
h ra
te
–So
lutio
n is
:
dN dtrN
KN
K=
−% &
( )
NK
NN
KN
et
rt=
+−
−0
00
()
Eco
logi
cal T
heor
y C
ont.
•N
ext,
allo
w f
or d
iffe
rent
mar
ket “
nich
es”:
– –M
uch
like
befo
re, b
ut e
ach
i (or
j) r
efer
s to
a d
iffe
rent
mar
ket n
iche
(a
diff
eren
t “sp
ecie
s”)
–0
ij is
com
petit
ion
coef
fici
ent b
etw
een
spec
ies
i and
j
–E
ach
spec
ies
grow
s to
its
carr
ying
cap
acity
or
isel
imin
ated
by
com
petit
ion
–So
lve
by c
ompu
ter
(you
can
use
Ste
lla I
I in
Com
pute
r C
entr
e)
dN dtrN
KN
N
Ki
ii
ii
ijj
ji
i
=
−−
% &' '
( )* *≠∑0
Firm
Str
ateg
ies
and
Dif
fere
ntia
tion
•Pr
oduc
e 1
or m
ultip
le v
arie
ties?
•Pr
oduc
ts d
iffe
rent
fro
m c
ompe
titor
s, o
r sa
me?
•“L
eapf
rog”
to r
etai
n be
st p
ositi
ons?
•Pr
oduc
e be
st v
arie
ties
righ
t aw
ay?
•R
elat
ion
to e
ntry
and
exi
t?
Stra
tegi
es in
UK
Fer
tiliz
erM
anuf
actu
re•
Tes
t ide
as a
bout
pro
duct
dif
fere
ntia
tion
•U
sing
UK
fer
tiliz
ers
as a
n ex
ampl
e (S
haw
, 198
2)
•Sq
uare
pro
duct
dif
fere
ntia
tion
spa
ce–
Tw
o se
ts o
f ch
emic
al e
lem
ents
to h
elp
plan
ts g
row
–N
itrog
en
–Ph
osph
orou
s &
pot
assi
um (
usu.
in e
qual
pro
port
ions
)
Tec
hnol
ogy
and
Dif
fere
ntia
tion
•Sh
akeo
ut th
eory
has
ver
tica
l dif
fere
ntia
tion
–Fi
rms
inno
vate
to im
prov
e th
eir
prod
uct q
ualit
y
•M
arke
t lea
ders
hip
turn
over
has
hor
izon
tal a
nd/o
rve
rtic
al d
iffe
rent
iati
on–
Dis
rupt
ive
tech
nolo
gy y
ield
s be
tter
qual
ity (
calc
ulat
ors)
–O
r cr
eate
s a
com
petin
g m
arke
t nic
he (
hard
dis
ks)
•W
hat i
f te
chno
logy
cre
ates
new
, lar
gely
inde
pend
ent s
ubm
arke
ts?
–Pr
otec
ts f
irm
s fr
om c
ompe
titio
n (p
enic
illin
)
Differentiation and Success inUS Penicillin Manufacture
• Penicillin: drug, attacks bacterial disease
• Produced naturally by some mold
• Natural penicillin: types G, O, or V– Patents disallowed (World War II projects)
– Produced by many firms
• Circa 1958, new techniques:– Extract chemical produced by mold
– Modify it by chemical methods
– Give to mold, get new types of penicillin
• Semi-synthetic penicillins: phenethicillin,ampicillin, …– Treat different diseases from natural penicillin
– Developed by specific firms, patented
– Legal battles limited patent rights for first 2
– Later types patented, licensed to few firms
Table: US PenicillinManufacturers by Type
• Also in course notes, pp. 109-110
• Drawn from Klepper and Simons (1997)
• Main source Synthetic Organic Chemicals
• 1948-1993
• Organized by penicillin type– Earliest first
– 1988+ if made only by Beecham not all listed
• Key innovations identified by †– According to Achilladelis (1993)
• Innovating firm identified by *
• Table has 4 pages, numbered at bottom
Penicillin G† [Innovation dated as 1942 by Achilladelis]Abbott 1948-1964Baker 1948-1952Bristol 1948-1959, 1961, 1974, 1977, 1981-1983Commercial Solvents 1948-1959Cutter 1948-1954Heyden 1948-1953Hoffman LaRoche 1948-1949Lederle 1948-1949, 1954-1955Lilly 1948-1969, 1972, 1976-1981Merck 1948-1986Pfizer 1948-1992Schenley 1948-1953Squibb 1948-1982X 1948-1953, Upjohn 1954-1957Wyeth 1948-1993U.S. Rubber 1949Monsanto 1954Penick, S.B. 1954-1955
Penicillin OX 1951-1953, Upjohn 1954-1964, 1966Pfizer 1968-1975
Penicillin V† *Lilly 1955-1990Abbott 1956-1974Wyeth 1956-1976, 1983-1985Bristol 1958, 1970-1985, 1987, 1989-1993Squibb 1968-1976Pfizer 1976-1988[*Also developed by Glaxo (UK).] 1
Phenethicillin†*Bristol 1959-1975Pfizer 1960-1965, 1967-1971Squibb 1960-1961, 1963-1964Wyeth 1962-1966[*Also developed by Beecham(UK).]
Ampicillin†Bristol 1963-1993Wyeth 1966-1993*Beecham 1968-1990Squibb 1968-1976Trade Enterprises 1971-1981Biocraft 1972-1993Kanasco 1986-1992
NEP penicillinMerck 1963
MethicillinBristol 1961-1985Beecham 1972-1982Wyeth 1991
OxacillinBristol 1961-1985Beecham 1969-1992Biocraft 1979-1992 2
Cloxacillin Bristol 1964-1985Beecham 1968-1993Biocraft 1980-1988, 1990-1993Kanasco 1991-1992
Nafcillin Wyeth 1964-1990Beecham 1975-1976, 1984-1988, 1990-1991Bristol 1975-1977, 1979-1985, 1987
Dicloxacillin Bristol 1966-1979, 1981-1985Beecham 1968-1992Wyeth 1968, 1970-1975, 1977-1989Biocraft 1983-1993Kanasco 1990, 1992
Hetacillin Bristol 1966-1979, 1981-1991
Carbenicillin† [Innovation dated as 1969 by Achilladelis]*Beecham 1970-1985, 1987, 1989+?*Pfizer 1972-1986, 1988Biocraft 1986
3
Amoxicillin Beecham 1973-1993Biocraft 1976-1993Bristol 1977-1993Trade Enterprises 1978-1979Wyeth 1980-1985Kanasco 1986-1992
Ticarcillin Beecham 1976-1993
Azlocillin† [*Developed 1978 by Bayer (Germany)]
Cyclacillin Wyeth 1978-1985Bristol 1984-1985Biocraft 1986, 1988Epicillin Trade Enterprises 1978-1982Kanasco 1986
Piperacillin Bristol 1982-1993
Amdinocillin† [*Developed 1984 by Roche (Switzerland)]
Sulbactam† [*Developed 1986 by Pfizer]
Floxacillin Beecham 1989+?
4
EC
2212
Ind
ustr
ial G
row
than
d C
ompe
titio
n
Lec
ture
9
Firm
siz
e an
d ag
e af
fect
gro
wth
,an
d th
us th
e fi
rm s
ize
dist
ribu
tion.
Mea
sure
s of
Fir
m S
ize
•D
iffe
rent
mea
sure
s fo
r di
ffer
ent p
urpo
ses
–Fi
nanc
ial o
r st
ock
mar
ket v
alue
–E
mpl
oyee
s
–Pr
oduc
tive
cap
acit
y
–V
alue
of
prod
uctio
n
–V
alue
add
ed o
f pr
oduc
tion
(ou
tput
– in
puts
)
•H
ighl
y co
rrel
ated
•E
mpl
oyee
s, p
rodu
ctio
n m
ain
focu
s to
day
No
Equ
ilibr
ium
Fir
m S
ize
•Si
ze c
hang
es g
radu
ally
–N
o in
stan
t exp
ansi
on to
a d
esir
ed s
ize
–Fi
nanc
e, h
irin
g, tr
aini
ng, p
urch
ase
& s
et-u
p of
equi
pmen
t and
ope
ratin
g m
etho
ds ta
ke ti
me
•Fi
rm s
ize
appa
rent
ly u
nlim
ited
–U
-sha
ped
long
run
cos
t cur
ves
of f
irm
s ar
e m
ythi
cal
–M
ay b
e U
-sha
ped
shor
t run
cur
ves,
long
run
cur
ves
for
a pa
rtic
ular
pla
nt
–M
ulti-
prod
uct f
irm
s ca
n ke
ep e
xpan
ding
, reo
rgan
izin
g
–W
itnes
s IB
M, G
ener
al M
otor
s, M
icro
soft
, etc
.
Lim
its to
Fir
m G
row
th
•Pe
nros
e (1
959)
, The
The
ory
of th
e G
row
thof
the
Fir
m
•N
o eq
uilib
rium
am
ount
of
outp
ut
•O
ptim
al g
row
th r
ate
inst
ead
•G
row
th is
lim
ited
–M
anag
eria
l lim
its
to e
xpan
sion
act
ivit
ies
–T
rain
ing
of n
ew e
mpl
oyee
s by
old
–E
.g.,
conv
ex c
osts
of
grow
th, g
'>0,
g''>
0
Sim
ple
Rep
rese
ntat
ions
of
Gro
wth
•G
ibra
t’s
“law
”–
All
fir
ms
have
sam
e pr
obab
ility
dis
trib
utio
n fo
r%
gro
wth
(at
a g
iven
tim
e)
–E
.g.,
firm
s w
ith 1
0 vs
. 10,
000
empl
oyee
s ha
vesa
me
chan
ce to
gro
w 5
0% o
r m
ore
in a
yea
r
•M
odif
icat
ions
:–
Seri
al c
orre
latio
n
–M
erge
r
–E
ffec
ts o
f fi
rm s
ize,
age
, ski
ll, te
chno
logy
, ...
Obs
erve
d G
row
th o
f Fi
rms
& P
lant
s
•E
xam
ine
patte
rns
for
US
plan
ts
•Pa
ttern
s ar
e si
mila
r fo
r fi
rms
•R
elat
ion
betw
een
% g
row
th a
nd a
ge, s
ize
•(I
gnor
ing
man
y ot
her
corr
elat
es o
f gr
owth
.)
•D
unne
, Rob
erts
, and
Sam
uels
on (
1989
)
Ave
rage
Fiv
e-Y
ear
Gro
wth
Rat
esam
ong
Surv
ivin
g Pl
ants
Age
Plan
t S
ize
(# o
f em
plo
yees
)(y
ears
)5
-19
20
-49
50
-99
10
0-2
49
25
0+
1-5
61
%3
0%
19
%1
3%
7%
6-1
03
4%
14
%7
%1
%-1
%1
1-1
53
1%
6%
-1%
-2%
-2%
Ave
rage
Fiv
e-Y
ear
Exi
t Rat
es
Age
Plan
t S
ize
(# o
f em
plo
yees
)(y
ears
)5
-19
20
-49
50
-99
10
0-2
49
25
0+
1-5
41
%4
0%
39
%3
3%
23
%6
-10
35
%2
7%
28
%2
5%
16
%1
1-1
53
0%
21
%2
3%
21
%1
3%
Ave
rage
Fiv
e-Y
ear
Gro
wth
Rat
es, w
ith E
xit =
-10
0% G
row
th
Age
Plan
t S
ize
(# o
f em
plo
yees
)(y
ears
)5
-19
20
-49
50
-99
10
0-2
49
25
0+
1-5
-6%
-22
%-2
8%
-24
%-1
8%
6-1
0-1
3%
-17
%-2
3%
-24
%-1
7%
11
-15
-9%
-16
%-2
4%
-22
%-1
5%
Obs
erve
d G
row
th o
f Fi
rms
in a
nIn
dust
ry
•M
easu
re s
ize
with
in th
e pr
oduc
t ind
ustr
y
•G
row
th li
mite
d by
mar
ket s
ize
•Si
ze a
t tim
e of
ent
ry–
Ent
rant
s at
dif
fere
nt ti
mes
hav
e si
mila
r si
zes?
–O
r do
es in
itial
siz
e gr
ow/f
all o
ver
time?
Why
?
•W
hich
fir
ms
grow
s ho
w m
uch,
whi
ch e
xit?
•H
ence
how
doe
s si
ze d
istr
ibut
ion
evol
ve?
Size
1 Y
ear
Aft
er E
ntry
, US
Tir
es
J=$2
,500
-9,9
99 K
=$1
0,00
0-49
,999
L=
50,0
00-1
99,9
99 M
=20
0,00
0-99
9,99
9 N
=$1
M+
Not
adj
uste
d fo
r in
flatio
n. S
ome
firm
s ar
e m
ulti-
prod
uct f
irm
s. M
edia
n sh
aded
.
Cop
yrig
ht ©
2001
Ken
neth
L. S
imon
s.
En
try
No.
fir
ms
by I
nitia
l ca
pita
lizat
ion
(US
$)
To
tal
Ye
ar
?J
KL
MN
Fir
ms
19
05
-09
17
26
22
33
21
91
0-1
42
92
01
92
01
84
11
01
91
5-1
92
31
13
33
25
81
03
19
20
-24
69
71
54
83
31
01
82
19
25
-29
19
03
12
70
41
19
30
-80
21
04
11
14
23
73
To
tal
17
83
06
01
26
99
48
54
1
Size
Dis
trib
utio
n, U
S T
ires
J=$2
,500
-9,9
99 K
=$1
0,00
0-49
,999
L=
50,0
00-1
99,9
99 M
=20
0,00
0-99
9,99
9 N
=$1
M+
Not
adj
uste
d fo
r in
flatio
n. S
ome
firm
s ar
e m
ulti-
prod
uct f
irm
s. M
edia
n sh
aded
.
Cop
yrig
ht ©
2001
Ken
neth
L. S
imon
s.
No.
fir
ms
by C
apita
lizat
ion
(US
$)
To
tal
Ye
ar
?J
KL
MN
Fir
ms
19
10
14
11
01
46
44
91
92
05
99
23
53
51
35
23
01
93
01
61
61
92
72
89
71
94
06
03
10
10
23
52
19
50
70
23
82
14
11
96
03
02
42
28
39
19
70
30
11
11
82
41
98
07
00
13
18
29
Firm
Siz
e D
istr
ibut
ions
•H
ow m
any
firm
s ar
e of
eac
h si
ze
•U
sed
to c
alcu
late
mea
sure
s of
con
cent
ratio
n–
N-f
irm
con
cent
ratio
n ra
tio: /
s i, N
larg
est f
irm
s
–H
erfi
ndah
l ind
ex: /
s i2 ,
all
firm
s
•A
ffec
ted
by–
Ent
ry a
nd e
xit (
and
size
s of
ent
rant
s &
exi
tors
)
–G
row
th
Skew
Siz
e D
istr
ibut
ions
Res
ult
•G
ibra
t’s
law
+ e
ntry
yie
lds
skew
siz
edi
stri
buti
on–
Ijir
i and
Sim
on (1
977)
, Ske
w D
istr
ibut
ions
and
the
Size
s of
Bus
ines
s F
irm
s(w
ith B
onin
i)
–W
ho p
rodu
ces
each
new
uni
t of
outp
ut?
•Pr
obab
ility0
of
prod
uctio
n by
a n
ew f
irm
•O
ther
wis
e, p
roba
bilit
y pr
opor
tiona
te to
fir
m s
ize
•Sk
ew d
istr
ibut
ion
(man
y sm
all f
irm
s, f
ewla
rge)
res
ults
as
# of
dra
ws .
∞
Skew
Dis
trib
utio
ns: A
Gen
eral
Phen
omen
on•
Ijir
i & S
imon
poi
nt o
ut g
ener
ality
:–
Size
s of
bus
ines
s fi
rms
(For
tune
500
)
–P
opul
atio
ns o
f ci
ties
–H
ow o
ften
wor
ds a
ppea
r in
a b
ook
•Sa
me
prin
cipl
e w
orks
in e
ach
case
–N
ew e
ntit
ies
appe
ar w
ith
som
e pr
obab
ilit
y
–L
ikel
ihoo
d of
nex
t app
eara
nce
prop
orti
onat
e to
num
ber
of p
ast a
ppea
ranc
es
You
Hav
e L
earn
ed
•Pe
nros
e’s
theo
ry o
f lim
ited
firm
gro
wth
•G
ibra
t’s
“law
”: g
row
th in
depe
nden
t of
size
•G
row
th f
aste
r fo
r yo
unge
r, s
mal
ler
plan
ts
•W
ithin
indu
stry
: ent
ry, e
xit,
grow
th–
Ent
ry s
ize
grew
som
ewha
t in
tires
–Su
rviv
ing
firm
s’ m
edia
n si
ze g
rew
in ti
res
•E
volu
tion
of s
kew
siz
e di
stri
butio
ns
EC
2212
Ind
ustr
ial G
row
than
d C
ompe
titio
n
Lec
ture
10
Firm
s’ te
chno
logi
cal s
ucce
ss f
uels
natio
ns’
econ
omic
gro
wth
.
Com
para
tive
Adv
anta
ge in
Inte
rnat
iona
l Tra
de•
Cla
ssic
al in
tern
atio
nal t
rade
mod
els
–D
iffe
rent
cou
ntri
es’
com
para
tive
adva
ntag
e in
dif
fere
ntin
dust
ries
–Pr
oduc
e w
here
the
adva
ntag
e ex
ists
, and
trad
e
•In
tern
atio
nal t
rade
in n
ew g
row
th m
odel
s–
Con
cent
ratio
ns o
f su
cces
s bu
ild u
p
–Su
cces
s br
eeds
suc
cess
, and
may
per
tain
to a
llin
dust
ries
–Po
ssib
ly y
ield
ing
a po
vert
y tr
ap
Com
para
tive
Adv
anta
ge E
xam
ple
•U
tili
ty =
F*E
, fro
m F
ood
and
Ent
erta
inm
ent
•10
0 pe
ople
eac
h in
cou
ntri
es A
and
B
Cou
ntry
AC
ount
ry B
Pvit
y O
utpu
t C
ons.
Pvit
y O
utpu
t C
ons.
Food
2 2
00 1
00
1
0
100
Ent
ert.
1
0
100
2
200
100
Pove
rty
Tra
p E
xam
ple
•U
tili
ty =
F*E
, fro
m F
ood
and
Ent
erta
inm
ent
•10
0 pe
ople
eac
h in
cou
ntri
es A
and
B
Cou
ntry
AC
ount
ry B
Pvit
y O
utpu
t C
ons.
Pvit
y O
utpu
t C
ons.
Food
2 1
00 1
00
1 5
0
50
Ent
ert.
2 1
00 1
00
1 5
0
50
•N
o tr
ade
need
s to
occ
ur,
A b
ette
r of
f th
an B
•C
an B
impr
ove
its p
rodu
ctiv
ity?
Lin
ear
and
Lea
pfro
g M
odel
sof
Tec
hnol
ogy
Dev
elop
men
t•
Dep
icti
ons
of te
chno
logy
dev
elop
men
t for
deve
lopi
ng c
ount
ries
–L
inea
r: S
tart
with
sim
ples
t tec
hnol
ogy,
gra
dual
ly b
uild
up
–L
eapf
rog:
Jum
p im
med
iate
ly to
a c
ompl
ex te
chno
logy
,us
e fo
r in
tern
atio
nal c
ompe
titio
n
•T
hese
dev
elop
men
t pat
tern
s ha
ppen
not
just
for
aco
untr
y in
agg
rega
te, b
ut a
lso
for
its
com
pani
es–
Com
pani
es a
re th
e bu
lk o
f th
e te
chno
logy
dev
elop
men
tpr
oces
s
Ana
m I
ndus
tria
l: L
inea
rD
evel
opm
ent
•K
orea
’s A
nam
Ind
ustr
ial,
wor
ld’s
larg
est
sem
icon
duct
or p
acka
ging
com
pany
•H
obda
y (1
995)
cha
ract
eriz
es li
near
gro
wth
pha
ses:
–19
68-8
0 Pa
ckag
ing
chip
s in
to p
last
ic/c
eram
ic c
ases
•m
achi
nery
, eng
inee
ring
, pro
duct
des
ign,
mat
eria
ls f
rom
US
–19
80-8
5 G
reat
er in
-hou
se p
roce
ss e
ngin
eeri
ng•
aide
d by
US
com
pani
es s
uch
as T
exas
Ins
trum
ents
–19
86-1
990s
In-
hous
e pr
oces
s en
gine
erin
g, p
rodu
cts
•te
chno
logy
dif
fusi
on &
in-h
ouse
eng
inee
ring
gro
wth
–B
y 19
92, U
S $1
.8 b
illio
n of
exp
ort s
ales
Foun
der:
Lea
pfro
g D
evel
opm
ent
•C
hina
’s F
ound
er, f
or C
hine
se la
ngua
ge p
rint
ing
syst
ems
•L
u (2
000)
des
crib
es it
s gr
owth
:–
Dre
w o
n kn
owle
dge
and
skill
s fr
om B
eijin
g U
nive
rsity
–C
opie
d &
cre
ated
sta
te-o
f-th
e-ar
t las
er p
rint
ing
tech
nolo
gies
–In
trod
uced
a p
rodu
ct c
ompe
titiv
e in
tern
atio
nally
–A
ble
to o
utco
mpe
te B
ritis
h, J
apan
ese,
US
firm
s
Lea
pfro
ggin
g A
rgum
ent
In h
igh-
tech
are
as, t
here
are
big
lags
bet
wee
n ou
rco
untr
y an
d ad
vanc
ed c
ount
ries
. M
any
new
idea
san
d m
etho
ds o
rigi
nate
abr
oad…
. H
owev
er, w
esh
ould
not
be
satis
fied
with
mer
ely
catc
hing
up
beca
use
this
wou
ld n
ot c
ome
up w
ith
com
petit
ive
prod
ucts
. It
was
inev
itab
le th
at w
e w
ould
cat
ch u
pfo
r qu
ite a
long
tim
e. H
owev
er, i
t was
pos
sibl
e to
leap
for
war
d ba
sed
on o
ur in
dige
nous
inno
vati
veca
pabi
litie
s.-
Chi
ef D
esig
ner
of th
e Fo
unde
r Sy
stem
Whe
n C
an L
eapf
rogg
ing
Wor
k?
•E
ithe
r co
mpe
titor
s do
n’t e
xist
yet
–E
nter
ear
ly in
an
indu
stry
with
a s
hake
out
•O
r ne
w te
chno
logy
hel
ps s
urpa
ss th
e co
mpe
titio
n–
Faci
litat
es tu
rnov
er o
f co
rpor
ate
lead
ersh
ip
•O
r te
chno
logy
(etc
.) n
ot c
ompe
titiv
ely
cruc
ial
–N
o sh
akeo
ut, n
o di
sadv
anta
ge to
late
ent
rant
s
•In
dige
nous
adv
anta
ges,
suc
h as
init
iall
y lo
w la
bor
cost
s, a
lso
help
Cat
chin
g U
p to
the
Com
petit
ion
•A
. Pro
duce
rs w
ithin
a c
ount
ry’s
pro
tect
ed m
arke
t–
Var
y in
qua
lity,
eff
icie
ncy
–Q
ualit
y an
d ef
fici
ency
enh
ance
pro
fit
–M
inim
um c
ompe
titiv
e qu
ality
, eff
icie
ncy
yiel
ds 0
pro
fit
•B
. Int
erna
tion
al p
rodu
cers
: hig
her
qual
ity,
effi
cien
cy–
How
can
you
mea
sure
wha
t it t
akes
to c
atch
up,
i.e.
, be
com
petit
ive
inte
rnat
iona
lly?
–L
ook
at q
ualit
y, e
ffic
ienc
y di
ffer
ence
s
–B
ench
mar
king
A. W
ithin
-Cou
ntry
Com
petit
ion
Qua
lity
Effi
cien
cy
Equ
ipro
fit l
ine #
= #
2
Equ
ipro
fit
line #
= #
1
B. B
etw
een-
Cou
ntry
Com
petit
ion
Qua
lity
Effi
cien
cy
Inte
rnat
iona
l pr
odu
cers
Hom
e-co
unt
ry
prod
uce
rs
Prot
ect
ive
gap#
p
Min
imu
mca
tch
-up #
c
#1
#2
#3
#4
#5
C. C
atch
ing
Up
•B
ench
mar
k to
det
erm
ine
diff
eren
ces
now
•E
stim
ate
tech
nolo
gy g
row
th ra
tes
for
com
petit
ors
•H
ow m
uch
tech
nolo
gy is
nee
ded
in f
utur
e to
be
com
peti
tive
?
•H
ow c
an y
ou c
atch
up?
–W
hat i
mpr
ovem
ents
to m
ake,
str
ateg
ies,
how
to g
ette
chno
logy
?
–C
ost?
Wor
th th
e in
vest
men
t?
–W
hat a
re th
e od
ds?
Nat
iona
l Tec
hnol
ogy
Polic
y
•W
hat c
an a
nat
iona
l gov
ernm
ent d
o?
•N
atio
nal t
echn
olog
y pr
ogra
ms:
–M
ITI
(Min
istr
y of
Int
erna
tiona
l Tra
de a
nd I
ndus
try,
now
ME
TI:
Min
istr
y of
Eco
nom
y,T
rade
, and
Ind
ustr
y)
–A
dvan
ced
Tec
hnol
ogy
Prog
ram
–B
ig g
over
nmen
t-fu
nded
pro
ject
s of
ten
spec
tacu
lar
failu
res,
are
out
of
favo
r
–B
ut n
ot a
lway
s fa
ilure
s, o
pera
te a
t var
ying
sca
les
–B
est n
ot to
pre
sum
e in
ter-
firm
coo
pera
tion
Nat
iona
l Tec
hnol
ogy
Polic
y co
nt.
•O
ther
pol
icie
s:–
Loc
atio
n in
cent
ives
(in
clud
ing
for
mul
tinat
iona
ls)
–T
ax &
fun
ding
ince
ntiv
es ta
rget
ed b
y te
chno
logy
–U
nive
rsity
R&
D f
undi
ng
–E
duca
tion
–C
hang
e cu
lture
s th
at h
inde
r in
nova
tion
–Sh
ift e
mph
asis
off
mili
tary
R&
D, o
r en
cour
age
tran
sfer
of m
ilita
ry te
chno
logy
top related