Learning from Learning by Doing : Lessons for Economic Growth M *1 * 2 1997 2004 10 1-2013 12 29 Version1.0 2014 3 2 *1 c 2010 MIT *2 c 2004-2013
i 1993 40 MIT 1991
ii R.M.S. * 1 *1
iii i 1 1 2 15 3 27 4 43 57
1 1 (1962) *1 1962 *1 :middlebrow theory.
2 1 30 1962 1/3 RAND
3 1962 (Solow, Tobin, von Weizsäcker, and Yaari, 1965 ) ( ) product
4 1 30 g(t) t G(t) t 1 a G bg n n 1/3 ) ) G G L = b g n dg (1.1) G G x = a(g G ) (1.2) (1.1) G (1.2) x = ag [ 1 ( ) 1/(1 n) ] 1 L cg 1 n (1.3) n 1 c = b/(1 n) n = 1 (1.3) G G
5 n < 1 n > 1 G L n > 1 n 1 n = 1 n > 1 k k n n = 1/3 k 1/1 + 1/2 n + 1/2 n + 1/2 n... + 1/k n n > 1 k=1 k n n > 1 n 1 n > 1 n > 1 L n = 1/3 (1.3) n 1 ( n = 0 ) 0 < n < 1 L cg 1 n cg 1 n L n > 1 c = b/(1 n) 1 G ( ) n/(1 n) x G = 1 1 L cg 1 n (1.4) n 1 ) ( ) 2n 1 2 x G 2 = nl c Gn 2 1 L 1 n cg 1 n (1.5) n < 1 n > 1 G G
6 1 n > 1 () L L G 1 n γ γ/(1 n) n < 1 (s ) 1962 n < 1 (1.3) G L 1/1 n 1 n < 1 n > 1 1 n > 1 n > 1 1 1
7 n = 1 (1.3) n = 1 c = b/(1 n) (1.1) (1.2) (1.3) 1962 x = ag ( 1 e L/b) (1.6) G 1 s (1.6) G sa(1 e L/b ) G s sa 50 n = 1 AK 1962
8 1 (1.6) n = 1 n 1 n n < 1 n < 1 n = 1 n > 1 AK 1 1 1
9 1 n 1 1 1 2.1 1.9 AK
10 1 AK
11 dh dt dh dt = khf(l H) = kh m f(l H ) m < 1 m > 1 L H m = 1 H kf(l H ) L H H AK AK MIT
12 1 1962
13
15 2
16 2 G bg n B + bg n a a/b B = 0 B 1.3
17 n = 1/2 1 0 n n G L = b (B + bg 1/2 dg G (2.1) x = a(g G ) (2.2) n = 1/2 2.1 G 2.2 x = a (2.3) d = 2b/B 2.3 L 2.1 B(G G ) < L 2.2 x < al/b 2.3 B-17
18 2 B q < 1 k B q k B 0 B 0 q q q q 1 m h mh 1 mh m a k
19 k + 1 (q k q k+1 )B 0 k + 1 1 G q k+1 B 0 + bg 1/2 k k 1 2.1 2.3 G G B q 1 i G a/(q i B 0 + bg 1/2 ) G
20 2 m q G n 1962 1
21 factor share drift bg n
22 2 q k B 0 + = r nr B + bg n nr B q m 1/m 1/5 5 1 q 0.9 10 2 m(1 q) nr m(1 q) q k B 0 bg n )
23 (m) (1 q) T k mt T t 0 t 0 + T x B t 0 t 0 + T B x ln q xt 1 ln q m ln q ln q = ln[1 (1 q)] (1 q) x mt mt 1 (ln q) 2 m(1 q) 2 /T
24 2 (n) (1 q) t k t t k+1 > t > t k [G(t) G(t k )] n > G(t) n **************************** 30
25
27 3 1 2.1 2.2 B = B 0 q k B 0 q k m a[bg n + B 0 q k ] 1 B 0 a 1 [bg n + q k ] 1 b t G(t) G(t) G t = G t 1 + x t (1 + r) t x t 1 r 0.03
28 3 G 1a 1b 100 50 b = 1, n = 0.33, m = 0.2, q = 0.95 1/3 5 1a 1b 100 10-15 50 5 50 1.25 0.1
29 q b m G m q r 2a 2b 1a 1b b 1 2 3a 3b b = 1 n = 0.33 m = 0.5 q = 0.98 m(1 q)
30 3 1 (1 q) 1a 1b 2a 2b q = 0.95 q q 1.00.980.950.90 0.1, 0.2, 0.5, 0.2 q 0.9511 q = 0 0.1 q 4a, 4b q
31 b = 1 m = 0.2 5a, 5b m = 0.5q = 0.98 2 (n) 0.33 0.5 G(t) 5c
32 3
33 m 0 = 0.5 t t 1 m t = (1 + u) m t 1 t 1 m t = (1 + u) 1 m t 1 u 0.05, 0.1, 0.15 b = 1, n = 0.33, q = 0.98 100 50 6a, 6b, 6c u = 0.05, 0.1, 0.15 6a 6b 6c t = 20 t = 35 t = 50 15 (1 + u) (1 + u) 1 (1 + u) (1 + v) 1 m t n b
34 3
35-1 n 1 1 n < 1 50 100 100 200 200 50 3 51 100 b = 1, n = 0.33, m = 0.2, q 0.95 100 200 100 t = 10 7a 100 3 2.5 3 7b 51 1 100 3 2.5 2 7a 7c 5 4 3 4 7a 7c 50 20-25 0.5 0.5
36 3
37 3 5 b 2 n 0.5 m 0 = 0.5 v = u = 0.05 5 5 m t 0.8 0.2 () () b = 1, n = 0.33, q = 0.98 m(1 q) 0.01 0.5 0.525 0.475 0.1 0.105 0.095 50 3 50 1,2 4, 5 3 100 5 200 5 8a,b,c,d,e 51-100 g
38 3 t = 100 t = 50 50 3
39 t = 50 t = 100 8c 3 1.35 2.1 0.6 1.5 8a 1.25 1.82 3 1 50 30
40 3 25 (35 25)/35 (1.10 0.82)/1.10 8b 2 8a 8c 8e 1.45 2.35 2.4 3 5 50 30 25 4 8d 8c 8e 8c 55 20-25 50 5 50 20 60
41 2
43 4
44 4 4.1 Energizer Bunny. Still going and going and going... *1 1988 GNP USD37,608 1985 GNP Summers and Heston 6.5 8.5 *1
45 1 2 30
46 4 30 1950 1960 1970 OPEC 1 1
47 GNP 20 25
48 4 1
49 GDP x y 1950 1960 1 GNP 2
50 4 3 1950 1960 1980 2
51
52 4 30 40 2 30 138 A D
53 GDP GDP GDP )
54 4 2 30 30 ky
55 GDP 40 30
57 Kenneth Arrow The Economic Implications of Learning by Doing, Review of Economic Studies 28 (1962): 155-73 David Levhari Further Implications of Learning by Doing, Review of Economic Studies 33 (1967): 31-38 Extensions of Arrow s Learning by Doing, Review of Economic Studies 33 (1967): 117-31 Eytan Sheshinski Tests of the Learning-by-Doing Hypothesis, Review of Economics and Statistics 49 (1967): 568-78 1930 Werner Hirsch, Manufacturing Progress Functions, Review of Economics and Statistics 34 (1952): 143-55 Armen Alchian, Reliability of Progress Curves in Airframe Production, Econometrica 31 (1963): 679-93 RAND Linda Argote and Dennis Epple, Learning Curves in Manufacturing, Science 247 (1990): 920-24 Sherwin Rosen, Learning by Experience as Joint Production, textitquarterly Journal of Economics 86 (1972):366-82 Drew Fudenberg and Jean Tirole, Learning by Doing and Market Performance, Bell Journal 14 (1983): 522-30 Alwyn Young, Learning by Doing and the Dynamic Effects of International Trade, Quarterly Journal of Economics 106 (1991): 369-406 Young, Invention and Bounded Learning by Doing, Journal of Political Economy 101 (1993): 443-72 Philippe Aghion and Peter Howitt, A Model of Growth Throught Creative Destruction, Econometrica 60 (1992): 323-51; Robert Solow, James Tobin, Christian von Weizäcker, and Menachem Yaari, Neoclassical Growth with Fixed Factor Proportions, Review of Economic Studies 33 (1965): 79-115; Robert Summers and Alan Heston, The Penn World Table (Mark 5): An Expanded Set of International Comparisons, 1950-1988, Quarterly Journal of Economics 106 (1991): 327-68.