EC2212 Industrial Growth
and Competition
Lecture 6
Sustaining technology lets skilled
early entrants destroy competitors.
Shakeouts
• New products: often rise then fall in number of
producers
• Fall in number of producers often called a
“shakeout”
• Most products have shakeouts, within 3+ decades
of when the market forms
• Can be very dramatic: US automobiles went from
273 producers to 5
• Concentrated market shares tend to result
US Automobile Producers, 1896-1966
275
firms
220
165
110
entry
55
0
1894
exit
1912
1930
1948
1966
Why Do Shakeouts Happen?
• Focus on the main cause
• … in products with severe shakeouts
• You will see
– Evidence on entry and exit
– Theory that best fits the facts
• approximately, Klepper (1996, 2001)
– Evidence on early-movers, technology
• Then discuss ramifications
Firms, Entry, Exit in Four Products (US)
Automobiles
Tires
32%
25%
%exit
5-year
moving
average
%exit
5-year
moving
average
0%
0%
275
275
firms
220
220
165
165
110
firms
110
entry
55
entry
55
exit
exit
0
1894
1912
1930
1948
1966
0
1905
1920
1935
Televisions
29%
5-year
moving
average
0%
0%
30
firms
88
24
66
18
44
12
firms
entry
22
0
1945
1980
%exit
5-year
moving
average
110
1965
Penicillin
%exit
24%
1950
exit
6
exit
1956
1967
1978
1989
entry
0
1942
1952
1962
1972
1982
1992
Price and Output in the Four Products
Aut omobiles
$4,500
millions
sold
price
4.5
Tires
300
80
price
index
millions
sold
3
$3,000
60
200
40
1.5
$1,500
100
20
$0
0
1900
$1,200
1910
1920
1930
1940
Televisions
millions
produced
price
$800
4
$0
0
millions
produced
Color
price
8
$400
4
1952
1959
1966
1973
1928
1937
Penicillin
price (1 0 3 $/ lb.)
12
prod’n ( 10 6 lbs.)
4
8
2
4
0
0
12
$800
$0
1945
1919
6
8
$400
$1,200
0
1910
12
B&W
0
0
1980
1945
1955
1965
1975
Explanation of Shakeouts
Part 1 of 3
Entrants and Their Skills
More Skilled Firms Can Earn More Profit
Pote nt ial Ent rants in Year X
% of
potenti al
entrant s
Shaded region: firm s
have enough skill to earn
profit > 0 afte r ent ry
low
high
compet ence at R&D
Entry & Growth Drive Down Price
• Limited number of firms have skills needed
to enter, at any point in time
• Each year some number of firms can enter
• Firms enter fairly small, but then grow
• Entry and growth increase total output
• More output, lower price (demand curve)
Skill Needed to Enter Rises over Time
Pote nt ial Ent rants in Year X
% of
potenti al
entrant s
Shaded region: firm s
have enough skill to earn
profit > 0 afte r ent ry
With lower price, need
more skill to earn a profit
low
high
compet ence at R&D
Entry Eventually Stops
Time 1
2
3
Entrants need increasing skill to earn profit > 0, since price falls
%
low
high
skill
low
high
low
skill
high
skill
May be more potential entrants, but eventually no entrants
EXAMPLE:
60 potential firms
40 enter
300 potential firms
70 enter
800 potential firms
0 enter
Explanation of Shakeouts
Part 2 of 3
R&D, Size, and Profit
R&D with Imitation
•
•
•
•
R&D improves quality, lowers cost
Decreasing returns to R&D
Cost-per-unit-of quality c = c(R), c'<0,c''>0
Firms benefit from R&D during 1 time
period
• Firms imitate all past innovations in the
next period
Firm i’s Profit at Time t
Π it = ( p t − [c t − s i c( R it )])Q it − R it − g(Q it − Q it −1 )
• pt price per unit of quality, pt = ƒ(∑Qit)
• [ct – sic(Rit)] cost per unit produced
– ct highest possible cost given imitation of past R&D
– si firm i’s skill at R&D
– c(Rit) cost decreases with current R&D, c'<0, c''>0
• Qit output produced
• Rit spending on R&D
• g(Qit – Qit-1) cost of growth, g'>0, g''>0
Implications of the Profit Function
• Firms choose Rit, Qit to maximize profit
• Larger firms spend more on R&D
– Spread cost of R&D over more output
– Remember lecture 3
• Growth is limited
– Firms grow each period
– Increasing marginal cost limits growth
• Size (Qit) and skill (si) enhance profit
Explanation of Shakeouts
Part 3 of 3
Exit (given Size and Skill)
Who Exits When?
• Firms exit if Πit < 0
• Growth causes exit at every t
– Growth → ΣQit → pt → profit
• Exiting firms are smallest, least-skilled
– Since size and skill enhance profit
• Earlier entrants are larger, ceteris paribus
– Have had more time and incentive to grow
• Skilled early entrants are long-run survivors
Summary in Course Notes, p. 89
( Potent ial) Ent rant s 1895-1904
(Pot enti al) Entr ant s 1905-1909
(Pot enti al) Entrants 1910-1916
% of
pot ent ial
entrants
% of
pot ent ial
entrants
incompetent :
higher cost s
% of
pot enti al
ent rant s
competent:
lower cost s
low
high
competence of managers
How big are firms th at entered in 1895-1904?
(By mid-1920s,
entr y becomes
impossible.)
low
high
competence of managers
How big... ent ered in 1905-1909?
circa 1904
small
circa 1909
medium (b ut 80% have exit ed)
small
circa 1916
large ( but 90% have exit ed)
medium ( but 80% have exit ed)
Firms always ent er at small sizes.
As t ime goes on, surviving firms grow.
At any point in t ime, earlier ent rant s are larger t han lat er ent rants.
Size and competence reduce a firm's cost s. Because of
survival of th e fitt est, firms in each group are forced
out unt il only compet ent early ent rant s remain.
low
high
competence of managers
How big... ent ered in 1910-1916?
small
Implications of the Theory
• Shakeout
– Entry eventually stops
– Exit continues forever, causing shakeout
• Earlier entrants have lower chance of exit
– Maybe not at first (depends on skill distribution)
– But eventually even high-skilled late entrants exit
• Earlier entrants do more R&D
• Firms successful at R&D survive better
% Survival by Entry Date of Automobile Producers
Firms Surviving
100%
S2
S1
31.6%
10.0%
Entrants in 1895-1904
3.2%
1.0%
1905-09
0.3%
1910-66
0.1%
0
14
28
42
56
70
Years of Production
% Survival by Entry Date in the Four Products
Automobiles
Firms Surviving
Tires
Firms Surviving
100%
100%
S2
S1
S2
31.6%
S3
S1
31.6%
10.0%
Entrants in 1895-1904
Entrants in 1901-1906
3.2%
10.0%
1907-16
1.0%
1905-09
3.2%
1917-22
0.3%
1910-66
0.1%
1923+
1.0%
0
14
28
42
56
70
0
20
40
60
Televisions
Firms Surviving
100%
80
Years of Production
Years of Production
Penicillin
Firms Surviving
100%
S1
S1
S2
S2
Entrants in 1943
31.6%
31.6%
1945-52
1953+
10.0%
10.0%
Entrants in 1946-1948
3.2%
3.2%
1949-1951
1952+
1.0%
0
9
18
1.0%
27
36
Years of Production
45
0
10
20
30
40
Years of Production
50
% Survival by Entry Date in a Non-Shakeout Product
Pens,
Ballpoint
100%
1968-79
1955-67
31.6%
1951-54
Entrants 1946-50
10.0%
3.2%
1.0%
0
7
1 4
2 1
2 8
3 5
Years of production
Innovation, % Adoption, by Entry Time in the Four Products
Use same entry-time cohorts as previously, but divide tires cohort 1
Relative innovation rates by product & innovation type — compare cohorts
Product
Automobiles
Automobiles
Tires
Tires
Tires
Tires
Televisions
Televisions
Penicillin
Innovation type
Product
Process
Product
Cord 1917
Cord 1920
Balloon 1923
Product
Process
Process
Cohort 1
9
3
1
36%
8%
100%
73%
63%
16%
2
63
5
Cohort 2
2
0
0
Cohort 3
1
0.1
0
62%
7%
1
7
0
0
0
0
Innovation and Exit
A ut o m o b i l e s
Smoot hed Hazard
.28
.21
Non-Innovat ors
.14
.07
Innovato rs
0
1895
1913
1931
1949
1967
Innovation and Exit in the Four Products
A u t o m o b i le s
Smoot hed Hazard
Tires
Smooth ed Hazard
.2 4
.28
Non-Innovat ors
.18
.21
Non-Innovat ors
.14
.12
.07
.06
Cord Tires
Innovat ors
Cord &
Balloon Tires
0
0
1895
1913
1931
1949
1935
1950
1965
1980
Pe n i c i l l i n
Telev isions
Smoot hed Hazard
1920
1967
Smoot hed Hazard
.36
.24
.27
Non-Innovat ors
.18
Non-Innovat ors
Innovato rs
.09
Semisynt hetic Producers
0
0
1952
1959
1966
1973
1980
1959
1967.25
1975.5
1983.75
1992
Ramifications of Shakeouts
• In industries with strong sustaining R&D
• High-skilled early entrants dominate
• Other firms may profit for a while
– But eventually forced to exit
• Enter early, keep up with R&D, to survive
• Concentration is a natural result
– Anti-trust authorities often investifate
– But expect concentration with legal behavior