Supplemental Materials for Natarajan Balasubramanian and Marvin B Lieberman, “Learning-byDoing and Market Structure”, The Journal of Industrial Economics
*DISCLAIMER: The research in this supplement was conducted while the authors were Census
Bureau research associates at the California Census Research Data Center (CCRDC). Research
results and conclusions expressed are those of the authors and do not necessarily indicate concurrence
by the Bureau of the Census. The results presented in this supplement have been screened to ensure
that no confidential data are revealed.
1
LIST OF APPENDICES & TABLES
Table A1: Learning Coefficients (Olley-Pakes with selection correction)
Appendix A2: Potential issues with measuring learning
Table A2.1: Rank correlations with alternative learning rates
Table A3: Learning by doing and lower bounds of concentration (Other
quantiles)
Table A4: Learning by doing and lower bounds of concentration (SIC-4
Level)
Table A5: Learning by doing and lower bounds of concentration (Industries
with missing R&D and advertising omitted)
Table A6:Learning by doing and lower bounds of concentration (Alternative
cutoffs for imprecise estimates)
Table A7: Learning by doing and lower bounds of concentration (Outliers
dropped)
Table A8: Learning by doing and lower bounds of concentration (Other
estimation approaches)
Table A9: Learning by doing and lower bounds of concentration
(4 quartiles of R&D and advertising)
Table A10: Learning by doing and lower bounds of concentration (National
Markets)
Table A11: Learning by doing and lower bounds of concentration
(Non-advertising industries)
Appendix A12: Learning coefficients using other estimation approaches
2
Ind
2011
2013
2015
2022
2023
2024
2026
2032
2033
2035
2037
2038
2041
2045
2047
2048
2051
2052
2053
2064
2066
2068
2077
2084
2086
2087
2091
2092
2096
2097
2098
Coeff
0.18
0.37
0.25
0.24
0.53
0.40
0.26
0.62
0.42
0.33
0.31
0.42
0.23
0.29
0.48
0.39
0.38
0.36
0.06
0.55
0.40
0.49
0.30
0.42
0.31
0.46
0.41
0.30
0.30
0.31
0.41
S.E.
(0.05)
(0.08)
(0.08)
(0.09)
(0.24)
(0.08)
(0.07)
(0.31)
(0.1)
(0.12)
(0.09)
(0.09)
(0.08)
(0.11)
(0.23)
(0.03)
(0.06)
(0.1)
(0.22)
(0.19)
(0.41)
(0.25)
(0.12)
(0.07)
(0.06)
(0.17)
(0.37)
(0.05)
(0.1)
(0.06)
(0.15)
Table A1: Learning Coefficients (Olley-Pakes with Selection Correction)
Rank
Ind
Coeff
S.E.
Rank
Ind
Coeff
S.E.
306
2099
0.36
(0.04)
105
2337 0.40
(0.05)
88
2211
0.27
(0.16)
193
2339 0.36
(0.06)
219
2221
0.55
(0.13)
14
2341 0.30
(0.06)
240
2231
-0.31
(0.27)
368
2353 0.23
(0.08)
19
2241
0.18
(0.1)
308
2361 0.41
(0.11)
72
2251
0.36
(0.21)
94
2369 0.20
(0.09)
199
2252
0.11
(0.07)
349
2371 0.18
(0.41)
8
2253
0.34
(0.06)
112
2386 0.11
(0.26)
54
2254
-0.07
(0.11)
364
2387 0.17
(0.12)
119
2257
0.31
(0.07)
141
2389 0.86
(0.24)
139
2258
0.21
(0.07)
271
2391 0.28
(0.06)
58
2259
0.52
(0.29)
23
2392 0.28
(0.05)
253
2261
0.52
(0.19)
24
2393 0.11
(0.18)
175
2262
0.24
(0.12)
241
2394 0.29
(0.05)
31
2269
0.12
(0.13)
343
2395 0.24
(0.08)
74
2273
0.36
(0.07)
100
2396 0.32
(0.05)
79
2281
0.12
(0.05)
341
2397 0.34
(0.28)
98
2282
0.15
(0.08)
328
2399 0.30
(0.07)
358
2295
0.21
(0.16)
275
2411 0.24
(0.03)
12
2297
0.43
(0.13)
51
2421 0.25
(0.03)
70
2298
0.43
(0.18)
50
2426 0.18
(0.06)
30
2299
0.32
(0.07)
136
2429 0.47
(0.17)
158
2311
0.32
(0.12)
132
2431 0.21
(0.03)
57
2321
0.24
(0.06)
244
2434 0.27
(0.03)
146
2322
0.28
(0.21)
187
2435 0.17
(0.08)
38
2323
0.13
(0.37)
339
2436 0.19
(0.08)
61
2325
0.20
(0.09)
294
2439 0.21
(0.04)
166
2326
0.33
(0.08)
124
2441 0.14
(0.07)
167
2329
0.21
(0.05)
277
2448 0.15
(0.03)
147
2331
0.39
(0.04)
75
2449 0.28
(0.13)
62
2335
0.23
(0.07)
250
2451 0.12
(0.03)
Rank
68
99
155
248
66
281
310
345
321
1
186
185
350
168
231
138
116
163
243
222
307
36
280
195
313
302
270
335
330
189
342
3
Ind
2452
2491
2493
2499
2511
2512
2514
2515
2519
2521
2522
2531
2541
2542
2591
2599
2621
2652
2653
2655
2657
2672
2673
2674
2675
2676
2677
2678
2679
2813
2819
2821
Coeff
0.32
0.17
0.21
0.22
0.26
0.24
0.25
0.36
0.15
0.20
0.33
0.37
0.23
0.20
0.25
0.27
0.36
0.38
0.26
0.25
0.06
0.44
0.38
0.28
0.30
0.66
0.21
-0.79
0.26
0.45
0.37
0.42
S.E.
(0.05)
(0.08)
(0.08)
(0.04)
(0.03)
(0.05)
(0.06)
(0.1)
(0.14)
(0.06)
(0.06)
(0.07)
(0.05)
(0.04)
(0.11)
(0.06)
(0.13)
(0.13)
(0.02)
(0.07)
(0.06)
(0.11)
(0.10)
(0.25)
(0.08)
(0.23)
(0.13)
(0.57)
(0.08)
(0.08)
(0.08)
(0.08)
Rank
134
316
278
258
203
242
221
103
329
284
118
90
247
289
215
191
102
80
213
217
359
42
83
180
160
4
279
369
198
40
92
55
Ind
2822
2833
2834
2835
2836
2841
2842
2843
2844
2851
2865
2869
2873
2874
2875
2879
2891
2892
2893
2899
2911
2951
2952
2992
3011
3052
3053
3061
3069
3081
3082
3083
Coeff
0.04
0.64
0.77
0.75
0.19
0.50
0.46
0.19
0.49
0.52
0.56
0.52
0.08
0.41
0.35
0.42
0.35
0.19
0.38
0.53
0.48
0.33
0.20
0.39
0.36
0.17
0.32
0.24
0.29
0.36
0.25
0.22
S.E.
(0.25)
(0.20)
(0.11)
(0.20)
(0.10)
(0.12)
(0.10)
(0.22)
(0.08)
(0.06)
(0.16)
(0.07)
(0.23)
(0.46)
(0.12)
(0.19)
(0.05)
(0.14)
(0.04)
(0.05)
(0.43)
(0.03)
(0.11)
(0.13)
(0.19)
(0.10)
(0.08)
(0.06)
(0.06)
(0.05)
(0.07)
(0.13)
Rank
361
7
2
3
299
27
37
304
29
22
11
21
354
63
111
56
108
296
84
18
35
122
282
78
106
319
131
232
170
101
225
259
Ind
Coeff
3084 0.35
3085 0.19
3086 0.19
3087 0.37
3088 0.39
3089 0.26
3111 0.10
3131 -0.17
3143 0.17
3144 0.41
3149 0.24
3161 0.20
3171 0.03
3172 0.17
3199 0.22
3229 0.32
3231 0.29
3241 0.54
3251 0.39
3253 0.13
3255 0.51
3261 0.36
3264 0.24
3269 0.19
3271 0.20
3272 0.29
3273 0.25
3274 -0.10
3275 0.23
3281 0.22
3291 0.43
3295 0.33
S.E.
(0.06)
(0.06)
(0.04)
(0.08)
(0.10)
(0.02)
(0.12)
(0.35)
(0.1)
(0.21)
(0.16)
(0.11)
(0.08)
(0.19)
(0.11)
(0.11)
(0.04)
(0.2)
(0.15)
(0.19)
(0.15)
(0.20)
(0.12)
(0.06)
(0.05)
(0.02)
(0.02)
(0.51)
(0.12)
(0.06)
(0.09)
(0.12)
Rank
110
303
301
93
77
208
352
366
318
60
235
286
363
317
265
129
176
15
76
340
26
96
246
300
283
169
216
365
251
264
47
117
4
Ind
3296
3297
3299
3312
3315
3316
3317
3321
3324
3325
3341
3351
3354
3356
3357
3363
3364
3365
3366
3369
3398
3399
3411
3412
3421
3423
3425
3429
3431
3432
3433
3441
Coeff
0.44
0.22
0.30
0.27
0.25
0.28
0.28
0.20
0.24
0.07
0.31
0.27
0.22
0.14
0.31
0.34
0.06
0.19
0.24
0.26
0.22
0.21
0.55
0.11
-0.24
0.17
0.31
0.22
0.29
0.54
0.33
0.24
S.E.
(0.11)
(0.17)
(0.11)
(0.06)
(0.06)
(0.14)
(0.1)
(0.04)
(0.1)
(0.11)
(0.09)
(0.14)
(0.09)
(0.11)
(0.04)
(0.09)
(0.11)
(0.08)
(0.11)
(0.28)
(0.04)
(0.07)
(0.09)
(0.11)
(0.32)
(0.04)
(0.14)
(0.04)
(0.33)
(0.29)
(0.08)
(0.03)
Rank
43
257
153
190
218
188
184
285
239
357
150
192
266
333
149
114
360
297
245
197
260
274
13
347
367
314
140
262
174
16
120
227
Ind
3442
3443
3444
3446
3448
3449
3451
3452
3462
3465
3469
3471
3479
3484
3491
3492
3493
3494
3495
3496
3497
3498
3499
3511
3519
3523
3524
3531
3534
3535
3536
3537
Coeff
0.24
0.20
0.26
0.25
0.33
0.26
0.20
0.26
0.26
0.32
0.23
0.21
0.20
0.15
0.18
0.21
0.36
0.18
0.61
0.28
0.30
0.20
0.24
0.31
0.23
0.27
0.04
0.36
0.29
0.25
0.33
0.24
S.E.
(0.04)
(0.04)
(0.02)
(0.04)
(0.08)
(0.05)
(0.04)
(0.06)
(0.08)
(0.06)
(0.03)
(0.04)
(0.03)
(0.28)
(0.07)
(0.16)
(0.23)
(0.09)
(0.09)
(0.05)
(0.21)
(0.05)
(0.02)
(0.28)
(0.07)
(0.05)
(0.22)
(0.09)
(0.13)
(0.04)
(0.12)
(0.09)
Rank
230
293
204
223
123
207
288
206
214
128
254
276
287
327
309
272
97
312
9
179
162
291
236
142
256
194
362
95
177
226
127
228
Ind
3541
3542
3543
3544
3545
3546
3548
3549
3552
3553
3554
3555
3556
3559
3561
3562
3563
3564
3565
3566
3567
3568
3569
3571
3572
3575
3577
3578
3579
3581
3585
3589
Coeff
0.34
0.38
0.17
0.18
0.24
0.43
0.19
0.32
0.29
0.15
0.40
0.28
0.26
0.30
0.41
0.09
0.22
0.08
0.31
0.24
0.16
0.22
0.28
0.66
0.37
0.39
0.42
0.56
0.38
0.31
0.23
0.31
S.E.
(0.08)
(0.12)
(0.04)
(0.02)
(0.04)
(0.17)
(0.19)
(0.07)
(0.09)
(0.08)
(0.11)
(0.07)
(0.11)
(0.03)
(0.06)
(0.08)
(0.08)
(0.09)
(0.07)
(0.08)
(0.08)
(0.12)
(0.04)
(0.07)
(0.2)
(0.16)
(0.08)
(0.26)
(0.12)
(0.33)
(0.05)
(0.05)
Rank
115
87
323
311
237
48
295
133
173
331
69
183
205
164
67
353
261
356
143
233
325
263
182
5
91
73
59
10
85
152
252
148
5
Ind
3592
3593
3594
3596
3599
3612
3613
3621
3624
3625
3629
3634
3643
3644
3645
3646
3648
3651
3652
3661
3663
3669
3671
3672
3674
3675
3676
3677
3678
3679
3691
3694
Coeff
0.32
0.54
0.16
0.10
0.20
0.37
0.38
0.24
0.31
0.33
0.45
0.48
0.21
0.48
0.24
0.38
0.11
0.38
0.25
0.51
0.32
0.30
0.11
0.25
0.44
0.14
0.08
0.14
0.26
0.33
0.50
0.26
S.E.
(0.16)
(0.18)
(0.16)
(0.13)
(0.01)
(0.13)
(0.06)
(0.06)
(0.27)
(0.08)
(0.13)
(0.15)
(0.09)
(0.22)
(0.09)
(0.11)
(0.2)
(0.12)
(0.09)
(0.11)
(0.06)
(0.17)
(0.29)
(0.07)
(0.04)
(0.08)
(0.11)
(0.05)
(0.09)
(0.03)
(0.16)
(0.06)
Rank
130
17
324
351
290
89
82
238
144
126
39
34
268
32
234
86
348
81
224
25
135
154
346
220
44
337
355
338
209
125
28
210
Ind
3695
3699
3711
3713
3714
3715
3716
3721
3724
3728
3731
3732
3743
3751
3792
3799
3812
3821
3822
3823
3824
3825
3826
3827
3829
3841
3842
3942
3944
3949
3951
3952
Coeff
0.28
0.30
0.17
0.29
0.30
0.17
0.33
0.21
0.31
0.42
0.16
0.23
0.26
0.43
0.14
0.24
0.21
0.43
0.41
0.29
0.42
0.35
0.40
0.20
0.27
0.36
0.35
0.32
0.34
0.26
0.17
0.45
S.E.
(0.16)
(0.06)
(0.07)
(0.05)
(0.03)
(0.05)
(0.09)
(0.12)
(0.09)
(0.06)
(0.05)
(0.04)
(0.09)
(0.12)
(0.05)
(0.1)
(0.07)
(0.15)
(0.1)
(0.06)
(0.1)
(0.04)
(0.09)
(0.08)
(0.05)
(0.04)
(0.04)
(0.12)
(0.08)
(0.04)
(0.37)
(0.21)
Rank
178
161
315
172
157
320
121
267
151
52
326
249
200
45
332
229
273
49
65
171
53
107
71
292
196
104
109
137
113
211
322
41
Ind
3953
3955
3961
3965
3991
3993
3995
3999
Coeff
0.14
0.30
0.41
0.26
0.43
0.21
0.30
0.19
S.E.
(0.07)
(0.18)
(0.10)
(0.17)
(0.13)
(0.03)
(0.17)
(0.02)
Rank
334
159
64
202
46
269
156
305
6
Appendix A2: Potential Issues with Measuring Learning
This appendix along with Table A2.1 identifies and addresses some important issues with
measuring learning.
Alternative estimation approaches (Rows 1-9)
We used a number of different alternative estimation approaches to ensure that the ranking of
industries were not driven by the choice of estimation approach. Specifically, we (a) dropped the
selection correction in the OP approach (OPNS) (b) used an extension of Ackerberg, Caves and
Frazer [2005] described in Appendix A10, (ACF) and (c) applied OLS. For the baseline ACF
approach, we adopted the most precise learning estimates (current period instruments along with
the materials decision to back out the unobserved heterogeneity).
Sample selection (Row 10)
Using a sample that excluded plants with any gaps produced learning estimates that were highly
correlated with baseline estimates.
Other measures of experience (Rows 11-12)
Using age in additional to cumulative output, and lagged cumulative output produced learning
estimates that were highly correlated with baseline estimates.
Measurement errors in capital (Row 13)
It is well known that there are errors with measuring capital. If it were true that such
measurement errors were more prevalent in some industries, then we may observe a high
measured rate of learning in such industries. While there is no easy way to rule this out, learning
coefficients were re-estimated using the year-end book value of assets. The resulting learning
coefficients were highly correlated with the baseline estimates.
Alternative production functions (Rows 14-15)
The following alternative production function forms (instead of the Cobb-Douglas with value
added) adopted here were tested – (a) a version of the Translog production function
(yijt=ajt+αj.kijt+ α’j.(kijt)2+βj.lijt+ β’j.(lijt)2+ γj (kijt)(lijt) + λj.xijt+εijt) and (b) a non-linear least
squares with allowance for depreciation in the experience term. These learning coefficients were
highly correlated with the baseline estimates.
R&D investments (Row 16-17)
Refer discussion in text.
Inter-temporal and inter-cohort variations (Rows 18-21)
The learning rates were re-estimated by industry for two different time periods (roughly equal
sub-samples), and for two different entry cohorts (roughly equal sub-samples). The resulting
learning estimates were highly correlated with the baseline estimates.
Geography (Row 22).
In order to rule out geography as a potential factor, learning rates were re-estimated with SMSA
fixed effects as controls for the plant’s location. The resulting learning estimates were highly
correlated with the baseline estimates.
7
TABLE A2.1: RANK CORRELATIONS WITH ALTERNATIVE LEARNING RATES
Rank Correlation Coefficient with
SIC4 Baseline Rates
Alternative estimation approaches
1
OLS
0.628***
2
OP with no selection correction
ACF with materials to invert, and
current-period variables as instruments
0.760***
0.548***
3
Rank Correlation Coefficient with
SIC3 Level Rates
Alternative estimation approaches
4
OLS
0.737***
5
OP with no selection correction
ACF with materials to invert, and
current-period variables as instruments
ACF with materials to invert, and lagged
experience as instrument
ACF with investments to invert, and
current-period variables as instruments
ACF with investments to invert, and
lagged experience as instrument
OLS learning rates excluding plants with
any gaps in observations
0.813***
6
7
8
9
10
11
12
13
Alternative measures of experience
OLS with age and cumulative output
OLS with lagged cumulative output
Alternative measure of capital
OLS with year end assets instead of
perpetual inventory
0.567***
0.540***
0.696***
0.650***
0.553***
0.666***
0.804***
0.690***
8
Rank Correlation Coefficient with
SIC3 Level Rates
14
15
Alternative production function
specifications
OLS with translog production function 0.745***
NLS with depreciated cumulative output 0.646***
16
17
Controlling for R&D
OLS with R&D expenditure
OLS with R&D stock
18
19
20
21
22
Time and Cohort Effects
OLS limited to observations between
1973 and 1984
OLS limited to observations between
1985 and 2000
OLS with entry cohorts between 1973
and 1984
OLS with entry cohorts between 1985
and 2000
Geography
OLS with SMSA fixed effects
0.558***
0.693***
0.559***
0.750***
0.619***
0.601***
0.704***
This table is extracted from Balasubramanian [2007], which studies inter-industry heterogeneity
in learning in greater detail. The primary level of aggregation in that study was SIC-3 due to
other data limitations. Given the administrative difficulties in obtaining Census disclosure for
multiple results of the same nature, all but the first three robustness checks above were limited to
the SIC-3 level, and using a sample of plants that with no gaps greater than three years between
consecutive observations. The SIC3 learning rates were significantly correlated with the
weighted-average SIC3 rates computed using the baseline SIC-4 rates (spearman’s rank
correlation coefficient: 0.67). Also, given the computational costs, we limited the checks in
many instances to OLS. *p < 0.10 **p < 0.05 ***p < 0.01.
9
TABLE A3: Learning by doing and lower bounds of concentration (Other quantiles)
C8
1st
10th
25th
50th
Constant
-6.37
-5.09
-4.01
-3.08
(1.45)*** (1.1)*** (0.79)*** (0.66)***
Learning Quartile 2 (D2)
0.45
0.23
-0.41
-0.74
(2.25)
(1.27)
(0.91)
(0.76)
Learning Quartile 3 (D3)
2.71
1.97
0.80
0.00
(1.65)
(1.39)
(1.06)
(0.75)
Learning Quartile 4 (D4)
4.48
2.80
1.36
0.60
(1.6)***
(1.23)** (0.91)
(0.94)
1/ln(S/σ)
30.68
26.66
22.92
20.17
(10.11)*** (7.1)*** (5.63)*** (4.86)***
D2*(1/ln(S/σ))
-1.98
(17.24)
-14.37
(11.69)
-26.56
(11.37)**
1.62
(1.2)
-9.08
(8.68)
-1.52
(0.96)
12.39
(7.01)*
-1.10
(8.86)
-11.96
(9.55)
-17.68
(8.29)**
1.63
(0.66)**
-8.19
(5.11)
-0.20
(0.57)
2.84
(4.5)
2.62
(6.42)
-4.39
(7.64)
-8.70
(6.64)
1.40
(0.56)**
-6.73
(4.19)
0.32
(0.52)
-1.18
(3.96)
3.77
(5.74)
0.49
(5.52)
-3.53
(7.42)
1.83
(0.53)***
-9.58
(4.12)**
0.27
(0.42)
-1.10
(3.35)
1,103
1,103
1,103
1,103
D3*(1/ln(S/σ))
D4*(1/ln(S/σ))
RND (DR)
DR*(1/ln(S/σ))
ADV (DA)
DA *(1/ln(S/σ))
N
Note: This table replicates results in Table III (C8) of the main text at the 1st, 10th, 25th and 50th percentiles.
10
TABLE A4: Learning by doing and lower bounds of concentration (SIC- 4 Level)
Constant
Learning Quartile 2 (D2)
Learning Quartile 3 (D3)
Learning Quartile 4 (D4)
C4
-7.23
(1.41)***
3.53
(1.82)*
4.13
(1.9)**
3.44
(1.53)**
C8
C20
-4.55
-5.17
(1.8)** (1.97)***
-0.32
0.63
(2.00)
(2.26)
1.69
2.54
(1.91)
(2.05)
3.43
3.90
(1.80)* (1.93)**
1/ln(S/σ)
35.21
19.87
(10.25)*** (13.65)
30.52
(15.07)**
D2*(1/ln(S/σ))
-25.82
(14.41)*
-27.71
(13.92)**
-22.14
(11.05)**
2.91
(1.1)***
-16.01
(8.31)*
-0.85
(1.09)
8.53
(8.42)
2.32
(15.57)
-9.42
(14.06)
-22.97
(13.19)*
0.68
(1.41)
0.51
(11.19)
-0.85
(1.48)
8.05
(11.78)
-4.87
(17.61)
-16.87
(15.09)
-27.38
(14.32)*
0.96
(1.51)
-0.56
(11.88)
-0.75
(1.58)
7.23
(12.46)
255
255
255
D3*(1/ln(S/σ))
D4*(1/ln(S/σ))
RND (DR)
DR*(1/ln(S/σ))
ADV (DA)
DA *(1/ln(S/σ))
N
Note: This table replicates results in Table III of the main text at the SIC4 industry level.
11
TABLE A5: Learning by doing and lower bounds of concentration
(Industries with missing R&D and advertising omitted)
C4
C8
C20
Constant
-5.98
-5.99
-5.88
(1.04)*** (1.17)*** (1.51)***
Learning Quartile 2 (D2) 1.42
1.70
1.55
(1.37)
(1.53)
(1.83)
Learning Quartile 3 (D3) 2.44
2.51
2.41
(1.25)*
(1.38)*
(1.71)
Learning Quartile 4 (D4) 3.64
3.96
3.99
(1.2)*** (1.29)*** (1.61)**
1/ln(S/σ)
26.34
(6.8)***
30.66
36.00
(7.75)*** (10.18)***
D2*(1/ln(S/σ))
-9.90
(9.83)
-14.00
(8.49)
-22.51
(8.16)***
2.03
(0.71)***
-11.56
(5.27)**
-1.46
(0.67)**
11.82
(5.07)**
-11.68
(10.8)
-14.46
(9.39)
-24.97
(8.72)***
2.04
(0.76)***
-11.13
(5.7)*
-1.10
(0.69)
9.18
(5.25)*
-11.47
(12.95)
-13.94
(11.86)
-25.35
(10.97)**
1.97
(0.92)**
-10.36
(7.08)
-0.57
(0.75)
5.33
(5.79)
1,025
1,027
1,027
D3*(1/ln(S/σ))
D4*(1/ln(S/σ))
RND (DR)
DR*(1/ln(S/σ))
ADV (DA)
DA *(1/ln(S/σ))
N
Note: This table replicates results in Table III (in the main text)
omitting industries with no R&D and advertising data.
12
TABLE A6: Learning by doing and lower bounds of concentration
(Alternative cutoffs for imprecise estimates)
C8
C8
C8
Constant
-5.35
(0.93)***
-6.06
(1.04)***
-6.12
(1.1)***
Learning Quartile 2 (D2)
1.09
(1.3)
2.18
(1.24)*
3.20
(1.06)***
1.62
(1.36)
2.72
(1.34)**
3.87
(1.15)***
1.74
(1.35)
2.71
(1.34)**
4.08
(1.19)***
1/ln(S/σ)
25.60
(5.69)***
31.15
(6.79)***
31.47
(7.29)***
D2*(1/ln(S/σ))
-6.91
(9.28)
-11.82
(8.38)
-18.56
(6.98)***
-10.64
(9.58)
-15.83
(9.09)*
-23.74
(7.71)***
-11.63
(9.61)
-15.71
(9.27)*
-25.78
(8.08)***
1.84
(0.73)**
-9.50
(5.3)*
-1.38
(0.68)**
11.73
(5.17)**
2.00
(0.75)***
-11.03
(5.52)**
-1.08
(0.66)
9.06
(5)*
2.07
(0.73)***
-11.48
(5.43)**
-1.10
(0.71)
9.18
(5.36)*
1,171
1,125
1,078
Learning Quartile 3 (D3)
Learning Quartile 4 (D4)
D3*(1/ln(S/σ))
D4*(1/ln(S/σ))
RND (DR)
DR*(1/ln(S/σ))
ADV (DA)
DA *(1/ln(S/σ))
N
Note: This table replicates Table III for C8 using alternative cutoffs for dropping imprecise
estimates. The decision rule was as follows: for a given cut-off n, if adding (subtracting) n
times the standard error to the estimated learning rate moved an industry from the first (fourth)
to the fourth (first) quartiles, the industry was excluded.
13
TABLE A7: Learning by doing and lower bounds of concentration (Outliers dropped)
C4
C8
C20
Constant
-6.02
-6.07
-5.99
(0.96)*** (1.08)*** (1.35)***
Learning Quartile 2 (D2) 1.46
1.65
1.30
(1.23)
(1.34)
(1.59)
Learning Quartile 3 (D3) 2.43
2.67
2.39
(1.21)** (1.33)** (1.6)
Learning Quartile 4 (D4) 3.62
3.88
4.22
(1.09)*** (1.17)*** (1.42)***
1/ln(S/σ)
26.73
31.20
36.57
(6.32)*** (7.16)*** (9.22)***
D2*(1/ln(S/σ))
-10.17
(8.99)
-13.92
(8.28)
-22.22
(7.35)***
2.08
(0.72)***
-12.01
(5.27)**
-1.41
(0.7)**
11.36
(5.31)**
-10.92
(9.54)
-15.36
(9.08)*
-23.80
(7.85)***
2.02
(0.75)***
-11.20
(5.55)**
-1.10
(0.72)
9.25
(5.43)*
-8.82
(11.42)
-13.35
(11.08)
-26.57
(9.64)***
1.95
(0.9)**
-10.64
(6.79)
-0.67
(0.74)
6.26
(5.71)
1,120
1,122
1,122
D3*(1/ln(S/σ))
D4*(1/ln(S/σ))
RND (DR)
DR*(1/ln(S/σ))
ADV (DA)
DA *(1/ln(S/σ))
N
Note: This table replicates results in Table III (in the main text)
omitting outlier industries in the top and bottom 2% of learning
rates. No imprecise estimates are dropped.
14
TABLE A8: Learning by doing and lower bounds of concentration
(Other estimation approaches)
OPNS
ACF
OLS
Constant
-6.54
-6.83
-6.78
(1.57)***
(1.06)***
(0.86)***
Learning Quartile 2 (D2)
1.60
1.66
2.49
(1.82)
(1.58)
(0.93)***
Learning Quartile 3 (D3)
2.15
3.37
2.41
(1.64)
(1.12)***
(1.08)**
Learning Quartile 4 (D4)
3.48
2.18
3.24
(1.65)**
(1.35)
(0.95)***
1/ln(S/σ)
34.92
(11.49)***
37.88
(7.6)***
36.14
(6.19)***
D2*(1/ln(S/σ))
-8.83
(13.32)
-13.25
(11.94)
-20.70
(11.83)*
1.36
(0.93)
-6.26
(7.04)
-0.15
(0.62)
1.55
(4.83)
-11.57
(10.83)
-22.40
(8.17)***
-13.69
(9.81)
0.50
(0.92)
0.40
(6.66)
0.53
(0.69)
-2.88
(5.13)
-12.43
(6.81)*
-14.36
(7.66)*
-18.00
(6.83)***
1.55
(0.67)**
-8.70
(4.92)*
-0.03
(0.62)
1.58
(4.62)
1,100
1,141
1,171
D3*(1/ln(S/σ))
D4*(1/ln(S/σ))
RND (DR)
DR*(1/ln(S/σ))
ADV (DA)
DA *(1/ln(S/σ))
N
Note: This table replicates results in Table III (in the main text) for C8 using
learning estimates from other estimation approaches. See Appendix A10 for
additional details on these approaches.
15
TABLE A9: Learning by doing and lower bounds of concentration
(4 quartiles of R&D and advertising)
C4
C8
C20
Constant
-5.84
-5.70
-5.47
(1.21)*** (1.33)*** (1.58)***
Learning Quartile 2 (D2)
1.42
(1.38)
1.85
(1.25)
3.84
(1.30)***
Learning Quartile 3 (D3)
Learning Quartile 4 (D4)
1.76
(1.51)
2.20
(1.4)
3.90
(1.39)***
1.99
(1.8)
2.40
(1.71)
3.99
(1.63)**
1/ln(S/σ)
24.94
27.40
31.77
(8.18)*** (9.13)*** (10.93)***
D2*(1/ln(S/σ))
-10.34
(10.07)
-9.94
(8.67)
-24.79
(9.14)***
-12.50
(10.95)
-12.03
(9.98)
-24.30
(9.89)**
-14.89
(12.99)
-13.65
(12.24)
-25.31
(11.52)**
-0.26
(0.69)
1.45
(1.2)
2.33
(0.96)**
2.65
(4.99)
-8.21
(8.83)
-13.77
(7.43)*
-0.72
(0.75)
1.44
(1.26)
1.87
(1.05)*
6.02
(5.55)
-7.97
(9.28)
-9.88
(8.34)
-0.87
(0.75)
1.06
(1.37)
1.78
(1.26)
7.07
(5.69)
-4.53
(10.16)
-9.18
(10.15)
-0.27
(0.82)
-1.47
(0.86)*
-1.27
(1.09)
3.57
(6.55)
13.63
(6.62)**
10.88
(8.07)
0.09
(0.89)
-0.98
(0.86)
-1.56
(1.13)
1.58
(7.2)
9.90
(6.76)
13.49
(8.45)
0.45
(1.09)
-0.79
(0.94)
-1.51
(1.23)
-1.49
(8.87)
8.38
(7.51)
13.58
(9.26)
1,101
1,103
1,103
D3*(1/ln(S/σ))
D4*(1/ln(S/σ))
RND2 (DR2)
RND3 (D R3)
RND4 (D R4)
RND2*(1/ln(S/σ))
RND3*(1/ln(S/σ))
RND4*(1/ln(S/σ))
ADV2 (DA2)
ADV3 (DA3)
ADV4 (DA4)
ADV2*(1/ln(S/σ))
ADV3*(1/ln(S/σ))
ADV4*(1/ln(S/σ))
N
Note: This table replicates results in Table III (in the main text) using 4 quartiles of R&D and advertising
instead of two classes.
16
TABLE A10: Learning by doing and lower bounds of concentration (National Markets)
Constant
Learning Quartile 2 (D2)
Learning Quartile 3 (D3)
Learning Quartile 4 (D4)
C4
-6.02
(1.61)***
1.54
(1.60)
2.92
(1.73)
3.47
(1.55)**
C8
C20
-5.76
-7.42
(1.75)*** (2.31)***
2.13
2.58
(1.84)
(2.33)
2.70
3.82
(1.94)
(2.42)
4.23
5.11
(1.75)**
(2.35)**
1/ln(S/σ)
31.07
32.93
52.41
(12.24)*** (13.26)*** (17.50)***
D2*(1/ln(S/σ))
-8.21
(12.48)
-19.29
(13.07)
-23.57
(11.76)**
1.85
(0.95)*
-13.14
(7.65)
-0.62
(0.89)
-4.17
(7.25)
-12.83
(14.51)
-16.70
(14.66)
-28.65
(13.22)**
1.47
(0.94)
-9.50
(7.38)
0.11
(0.92)
-0.06
(7.45)
-16.85
(18.16)
-25.78
(18.44)
-34.56
(17.84)**
1.89
(1.16)
-12.90
(8.98)
1.24
(0.97)
-9.53
(7.69)
519
519
519
D3*(1/ln(S/σ))
D4*(1/ln(S/σ))
RND (DR)
DR*(1/ln(S/σ))
ADV (DA)
DA *(1/ln(S/σ))
N
Note: This table replicates results in Table III (in the main text) for industries with ‘national
markets’. We used the method adopted in Syverson [2004]. (We thank Chad Syverson for
providing these data). Using data from the 1977 Commodity Transportation Survey, this method
computes, by industry, the tonnage-weighted average distance the output is shipped. The CTS
provides seven distance classes (<100, 100-199, 200-299,300-499, 500-999, 1000-1499, >1500
miles) and the tonnage in each class. It is assumed that distance is distributed uniformly, and
used the average distance in each class (e.g. 50 miles in the <100 miles class). For the >1500
miles class, 1750 miles were used. Industries with tonnage-weighted average distance greater
than the median tonnage-weighted average distance were classified as ‘national markets’.
17
TABLE A11: Learning by doing and lower bounds of concentration
(Non-advertising industries)
C4
C8
C20
Constant
-5.95
-5.65
-5.60
(1.03)*** (1.15)*** (1.26)***
Learning Quartile 2 (D2) 0.67
1.46
1.99
(1.54)
(1.71)
(1.83)
Learning Quartile 3 (D3) 2.25
2.11
2.29
(1.32)
(1.47)
(1.71)
Learning Quartile 4 (D4) 3.04
2.86
2.77
(1.13)** (1.20)** (1.36)**
1/ln(S/σ)
26.81
28.28
33.99
(7.02)*** (7.75)*** (8.87)***
D2*(1/ln(S/σ))
-4.37
(11.75)
-14.09
(9.40)
-20.21
(8.37)**
2.45
(0.79)***
-14.09
(6.08)**
600
D3*(1/ln(S/σ))
D4*(1/ln(S/σ))
RND (DR)
DR*(1/ln(S/σ))
N
-10.85
(12.99)
-12.65
(10.43)
-17.62
(8.78)**
2.59
(0.89)***
-14.15
(6.72)**
601
-16.37
(14.00)
-14.48
(12.48)
-17.24
(10.19*
2.50
(1.12)**
-12.91
(8.68)
601
Note: This table replicates results in Table III (in the main text) for SIC-3 industries that report
no advertising or have advertising intensities below the median (0.37% of industry sales). It also
excludes advertising terms in the empirical specification.
18
SIC
2011
2013
2015
2022
2023
2024
2026
2032
2033
2035
2037
2038
2041
2045
2047
2048
2051
2052
2053
2064
2066
2068
2077
2084
2086
2087
2091
2092
2096
2097
Appendix A12: Learning coefficients using other estimation approaches
OP (NS)
ACF
OLS
Coeff
S.E.
Rank
Coeff
S.E. Rank
Coeff
S.E. Rank
0.14 (0.05) 337
0.02 (0.06) 353
0.07 (0.02) 331
0.18 (0.1) 318
0.48 (0.06) 33
0.24 (0.02) 82
0.24 (0.07) 260
0.25 (0.06) 173
0.15 (0.03) 207
0.24 (0.11) 256
0.25 (0.1) 176
0.19 (0.03) 145
0.18 (0.26) 316
0.26 (0.2) 158
0.19 (0.08) 146
0.44 (0.08) 51
0.31 (0.1) 125
0.27 (0.03) 48
0.27 (0.08) 197
0.28 (0.08) 140
0.24 (0.03) 85
0.63 (0.47) 9
0.90 (0.41) 3
0.46 (0.08) 6
0.42 (0.09) 59
0.38 (0.11) 71
0.25 (0.04) 71
0.38 (0.12) 90
0.24 (0.14) 187
0.23 (0.05) 90
0.32 (0.1) 147
0.24 (0.16) 184
0.28 (0.04) 46
0.45 (0.08) 43
0.38 (0.1) 74
0.23 (0.03) 89
0.27 (0.09) 210
0.34 (0.08) 100
0.18 (0.04) 157
0.23 (0.13) 269
0.35 (0.16) 92
0.25 (0.04) 70
0.56 (0.21) 16
0.41 (0.15) 59
0.24 (0.05) 84
0.41 (0.03) 65
0.50 (0.04) 30
0.30 (0.02) 32
0.26 (0.09) 225
0.31 (0.06) 123
0.26 (0.01) 62
0.34 (0.1) 117
0.35 (0.13) 91
0.20 (0.03) 127
0.06 (0.22) 358
0.11 (0.17) 318
(0.07) (0.07) 366
0.01 (0.25) 363
0.47 (0.18) 34
0.15 (0.04) 210
0.25 (0.25) 245
0.21 (0.16) 214
0.10 (0.06) 306
0.48 (0.35) 33
0.46 (0.14) 39
0.30 (0.07) 28
0.28 (0.09) 194
0.42 (0.09) 55
0.21 (0.04) 122
0.39 (0.08) 83
0.44 (0.09) 49
0.29 (0.04) 34
0.29 (0.07) 181
0.32 (0.07) 114
0.23 (0.02) 97
0.40 (0.2) 69
0.80 (0.18) 4
0.55 (0.05) 2
0.52 (0.4) 24
0.22 (0.29) 194
0.27 (0.08) 50
0.29 (0.05) 179
0.07 (0.07) 335
0.20 (0.03) 136
0.33 (0.11) 130
0.12 (0.08) 310
0.07 (0.05) 332
0.31 (0.06) 158
0.16 (0.06) 264
0.22 (0.03) 109
19
SIC
2098
2099
2211
2221
2231
2241
2251
2252
2253
2254
2257
2258
2259
2261
2262
2269
2273
2281
2282
2295
2297
2298
2299
2311
2321
2322
2323
2325
2326
2329
2331
OP (NS)
Coeff
S.E.
Rank
0.39 (0.17) 80
0.36 (0.04) 94
0.23 (0.13) 279
0.52 (0.13) 23
0.15 (0.31) 332
0.16 (0.1) 325
0.36 (0.32) 98
0.12 (0.08) 351
0.34 (0.05) 120
(0.04) (0.12) 364
0.33 (0.09) 133
0.20 (0.07) 299
0.74 (0.35) 4
0.45 (0.23) 39
0.28 (0.14) 192
0.23 (0.13) 272
0.37 (0.06) 93
0.15 (0.04) 331
0.16 (0.11) 328
0.31 (0.13) 152
0.36 (0.09) 104
0.39 (0.15) 84
0.33 (0.07) 136
0.34 (0.11) 128
0.26 (0.06) 230
0.51 (0.21) 26
0.04 (0.4) 361
0.19 (0.1) 307
0.31 (0.07) 150
0.23 (0.05) 278
0.40 (0.03) 70
ACF
Coeff
S.E.
0.29 (0.12)
0.20 (0.05)
0.18 (0.11)
0.56 (0.21)
(0.14) (0.16)
0.13 (0.11)
0.10 (0.16)
0.03 (0.08)
0.40 (0.09)
(0.05) (0.19)
0.38 (0.06)
0.27 (0.07)
0.27 (0.18)
0.22 (0.13)
0.13 (0.12)
0.24 (0.11)
0.31 (0.07)
0.24 (0.12)
0.01 (0.16)
0.34 (0.15)
0.07 (0.09)
0.25 (0.12)
0.34 (0.06)
0.40 (0.12)
0.28 (0.07)
0.45 (0.16)
0.02 (0.19)
0.37 (0.1)
0.45 (0.13)
0.26 (0.07)
0.29 (0.05)
Rank
135
223
247
26
367
299
319
349
62
364
73
152
156
196
301
179
120
177
355
94
338
175
97
61
147
41
354
75
43
162
137
Coeff
0.20
0.18
0.10
0.22
0.01
0.12
0.24
0.06
0.23
0.04
0.14
0.19
0.24
0.17
0.14
0.17
0.20
0.07
0.10
0.19
0.21
0.20
0.27
0.24
0.18
0.28
0.20
0.26
0.18
0.18
0.25
OLS
S.E.
(0.05)
(0.02)
(0.04)
(0.03)
(0.07)
(0.04)
(0.06)
(0.03)
(0.02)
(0.06)
(0.03)
(0.03)
(0.07)
(0.05)
(0.04)
(0.04)
(0.03)
(0.03)
(0.04)
(0.04)
(0.04)
(0.05)
(0.03)
(0.03)
(0.02)
(0.03)
(0.11)
(0.03)
(0.04)
(0.02)
(0.01)
Rank
128
160
312
104
361
270
79
345
96
353
235
144
80
169
236
168
131
339
303
149
115
133
60
81
166
44
132
63
167
153
74
20
SIC
2335
2337
2339
2341
2353
2361
2369
2371
2386
2387
2389
2391
2392
2393
2394
2395
2396
2397
2399
2411
2421
2426
2429
2431
2434
2435
2436
2439
2441
2448
2449
OP (NS)
Coeff
S.E.
Rank
0.39 (0.06) 79
0.41 (0.06) 63
0.35 (0.04) 113
0.31 (0.06) 160
0.24 (0.08) 251
0.36 (0.14) 101
0.22 (0.09) 286
(0.24) (0.75) 367
(0.41) (0.37) 368
0.15 (0.15) 330
0.80 (0.15) 3
0.25 (0.06) 236
0.28 (0.05) 193
0.36 (0.19) 102
0.31 (0.05) 159
0.25 (0.08) 243
0.33 (0.05) 140
0.19 (0.32) 310
0.32 (0.07) 146
0.24 (0.03) 248
0.27 (0.03) 198
0.19 (0.06) 312
0.45 (0.18) 40
0.21 (0.03) 290
0.29 (0.06) 178
0.20 (0.08) 306
0.22 (0.11) 285
0.23 (0.04) 275
0.04 (0.07) 360
0.16 (0.04) 329
0.30 (0.15) 166
ACF
Coeff
S.E.
0.43 (0.02)
0.37 (0.06)
0.34 (0.05)
0.22 (0.07)
0.23 (0.11)
0.09 (0.11)
0.14 (0.1)
0.07 (0.16)
0.56 (0.19)
(0.07) (0.16)
0.34 (0.08)
0.16 (0.06)
0.28 (0.05)
0.25 (0.12)
0.11 (0.04)
0.12 (0.07)
0.22 (0.07)
0.01 (0.09)
0.26 (0.08)
0.17 (0.02)
0.30 (0.02)
0.09 (0.05)
0.20 (0.11)
0.13 (0.06)
0.15 (0.04)
0.28 (0.09)
0.27 (0.12)
0.19 (0.06)
0.08 (0.06)
0.11 (0.03)
0.23 (0.12)
Rank
51
77
102
198
193
326
296
339
25
365
103
258
143
171
311
309
195
358
166
257
131
329
218
302
278
142
150
238
334
317
191
Coeff
0.26
0.28
0.20
0.26
0.10
0.18
0.22
0.23
0.18
0.11
0.27
0.16
0.13
0.15
0.15
0.16
0.17
0.12
0.17
0.12
0.17
0.07
0.29
0.14
0.10
0.13
0.21
0.15
0.06
0.03
0.11
OLS
S.E.
(0.01)
(0.02)
(0.01)
(0.02)
(0.03)
(0.03)
(0.03)
(0.06)
(0.07)
(0.04)
(0.05)
(0.02)
(0.02)
(0.04)
(0.02)
(0.03)
(0.02)
(0.04)
(0.02)
(0.01)
(0.01)
(0.02)
(0.06)
(0.01)
(0.01)
(0.04)
(0.03)
(0.02)
(0.03)
(0.01)
(0.05)
Rank
64
45
135
65
297
163
106
93
159
283
56
204
252
209
216
205
180
272
179
274
176
336
37
241
311
253
120
218
341
355
282
21
SIC
2451
2452
2491
2493
2499
2511
2512
2514
2515
2519
2521
2522
2531
2541
2542
2591
2599
2621
2652
2653
2655
2657
2672
2673
2674
2675
2676
2677
2678
2679
2813
OP (NS)
Coeff
S.E.
Rank
0.13 (0.03) 347
0.31 (0.04) 161
0.14 (0.08) 339
0.23 (0.09) 276
0.19 (0.04) 314
0.26 (0.03) 231
0.27 (0.05) 208
0.26 (0.07) 226
0.35 (0.08) 107
0.08 (0.13) 356
0.20 (0.06) 300
0.14 (0.09) 338
0.41 (0.07) 64
0.23 (0.05) 270
0.20 (0.04) 303
0.25 (0.11) 237
0.30 (0.05) 170
0.34 (0.17) 127
0.35 (0.12) 106
0.24 (0.03) 261
0.26 (0.06) 227
0.07 (0.06) 357
0.45 (0.1) 45
0.39 (0.09) 87
0.35 (0.28) 109
0.27 (0.09) 205
0.59 (0.22) 12
0.24 (0.14) 250
(0.92) (1.18) 369
0.27 (0.08) 214
0.45 (0.11) 42
ACF
Coeff
S.E.
0.18 (0.06)
0.27 (0.07)
(0.00) (0.1)
0.28 (0.08)
0.15 (0.03)
0.08 (0.02)
0.15 (0.05)
0.19 (0.09)
0.33 (0.05)
0.24 (0.11)
0.01 (0.06)
0.24 (0.06)
0.27 (0.07)
0.12 (0.03)
0.12 (0.06)
0.20 (0.08)
0.15 (0.05)
0.38 (0.2)
0.42 (0.22)
0.31 (0.03)
0.21 (0.09)
0.14 (0.04)
0.64 (0.15)
0.33 (0.08)
0.34 (0.16)
0.20 (0.07)
0.05 (0.16)
0.09 (0.16)
(0.47) (0.16)
0.35 (0.09)
0.57 (0.09)
Rank
245
153
363
141
269
331
268
229
104
181
356
178
155
307
304
216
276
69
56
118
211
294
15
105
96
221
345
328
369
85
24
OLS
Coeff
S.E.
0.08 (0.02)
0.16 (0.02)
0.15 (0.03)
0.16 (0.03)
0.13 (0.01)
0.13 (0.01)
0.13 (0.01)
0.10 (0.03)
0.16 (0.02)
0.10 (0.04)
0.07 (0.02)
0.17 (0.03)
0.13 (0.03)
0.10 (0.01)
0.03 (0.02)
0.16 (0.03)
0.11 (0.02)
0.09 (0.04)
0.18 (0.05)
0.16 (0.01)
0.18 (0.02)
0.02 (0.02)
0.34 (0.04)
0.17 (0.03)
0.09 (0.04)
0.13 (0.03)
0.11 (0.05)
0.05 (0.05)
(0.25) (0.1)
0.14 (0.02)
0.28 (0.04)
Rank
327
188
213
201
266
261
250
310
189
304
337
185
255
307
354
202
290
313
164
206
161
356
21
175
319
264
287
350
369
226
42
22
SIC
2819
2821
2822
2833
2834
2835
2836
2841
2842
2843
2844
2851
2865
2869
2873
2874
2875
2879
2891
2892
2893
2899
2911
2951
2952
2992
3011
3052
3053
3061
3069
Coeff
0.36
0.41
0.17
0.58
0.82
0.72
0.24
0.52
0.43
0.12
0.48
0.55
0.63
0.50
0.14
0.34
0.39
0.49
0.35
0.20
0.40
0.54
0.53
0.35
0.19
0.40
0.36
0.22
0.34
0.28
0.29
OP (NS)
S.E.
Rank
(0.08) 96
(0.07) 62
(0.67) 322
(0.2) 14
(0.21) 2
(0.18) 5
(0.14) 252
(0.12) 22
(0.11) 56
(0.19) 350
(0.1) 32
(0.06) 18
(0.2) 10
(0.07) 27
(0.3) 336
(0.55) 126
(0.18) 78
(0.18) 30
(0.05) 116
(0.15) 301
(0.04) 73
(0.05) 19
(0.41) 21
(0.03) 114
(0.13) 308
(0.12) 72
(0.22) 105
(0.08) 287
(0.08) 119
(0.06) 185
(0.05) 174
Coeff
0.38
0.46
0.23
0.33
0.70
0.71
0.08
0.43
0.44
0.21
0.35
0.44
0.72
0.47
0.35
0.43
0.45
0.60
0.34
0.16
0.37
0.60
0.77
0.42
0.41
0.43
0.33
0.21
0.18
0.24
0.14
ACF
S.E.
(0.13)
(0.08)
(0.16)
(0.28)
(0.14)
(0.2)
(0.16)
(0.1)
(0.12)
(0.29)
(0.11)
(0.09)
(0.21)
(0.1)
(0.22)
(0.49)
(0.08)
(0.21)
(0.08)
(0.15)
(0.11)
(0.06)
(0.31)
(0.04)
(0.1)
(0.19)
(0.15)
(0.11)
(0.1)
(0.06)
(0.06)
Rank
72
38
188
107
9
8
332
53
46
205
90
45
7
36
88
50
44
19
99
262
81
20
5
54
60
52
108
210
239
185
288
Coeff
0.25
0.38
0.19
0.51
0.49
0.37
0.10
0.33
0.28
0.12
0.27
0.27
0.27
0.36
0.45
0.56
0.27
0.32
0.25
0.14
0.26
0.43
0.55
0.24
0.10
0.29
0.29
0.12
0.19
0.08
0.16
OLS
S.E.
(0.03)
(0.03)
(0.08)
(0.04)
(0.03)
(0.04)
(0.03)
(0.03)
(0.04)
(0.06)
(0.03)
(0.02)
(0.05)
(0.03)
(0.07)
(0.12)
(0.04)
(0.05)
(0.02)
(0.05)
(0.02)
(0.02)
(0.1)
(0.01)
(0.03)
(0.04)
(0.05)
(0.05)
(0.03)
(0.02)
(0.02)
Rank
76
14
152
4
5
15
309
23
43
267
55
47
53
17
7
1
54
24
77
227
61
9
3
86
296
38
35
271
141
322
195
23
SIC
3081
3082
3083
3084
3085
3086
3087
3088
3089
3111
3131
3143
3144
3149
3161
3171
3172
3199
3229
3231
3241
3251
3253
3255
3261
3264
3269
3271
3272
3273
3274
OP (NS)
Coeff
S.E.
Rank
0.35 (0.05) 110
0.27 (0.07) 204
0.28 (0.13) 191
0.33 (0.04) 134
0.22 (0.07) 280
0.21 (0.05) 291
0.39 (0.07) 86
0.43 (0.15) 58
0.27 (0.02) 201
0.09 (0.1) 355
(0.08) (0.46) 365
0.13 (0.12) 345
0.47 (0.23) 35
0.27 (0.16) 199
0.25 (0.12) 233
0.03 (0.11) 362
0.22 (0.21) 284
0.25 (0.15) 240
0.26 (0.1) 229
0.27 (0.04) 203
0.56 (0.17) 17
0.40 (0.13) 68
0.43 (0.24) 57
0.36 (0.14) 97
0.32 (0.29) 141
0.29 (0.13) 177
0.19 (0.06) 311
0.23 (0.06) 264
0.33 (0.02) 135
0.28 (0.02) 186
1.16 (0.72) 1
ACF
Coeff
S.E.
0.40 (0.05)
0.26 (0.09)
0.15 (0.09)
0.53 (0.11)
0.20 (0.08)
0.37 (0.05)
0.49 (0.09)
0.45 (0.12)
0.31 (0.02)
0.28 (0.08)
0.62 (0.33)
0.12 (0.13)
0.22 (0.12)
0.16 (0.21)
0.15 (0.08)
0.12 (0.12)
0.06 (0.1)
0.26 (0.13)
0.14 (0.1)
0.23 (0.05)
0.20 (0.16)
0.33 (0.07)
0.02 (0.15)
0.41 (0.14)
0.30 (0.1)
(0.00) (0.12)
0.14 (0.06)
0.19 (0.05)
0.25 (0.02)
0.35 (0.02)
0.10 (0.28)
Rank
63
165
275
29
225
79
31
40
119
146
18
305
200
267
280
306
342
160
292
189
224
110
352
57
126
362
284
230
167
89
324
OLS
Coeff
S.E.
0.28 (0.02)
0.08 (0.03)
0.12 (0.04)
0.22 (0.03)
0.15 (0.03)
0.13 (0.02)
0.23 (0.03)
0.17 (0.03)
0.11 (0.01)
0.06 (0.05)
0.20 (0.08)
0.08 (0.04)
0.23 (0.05)
0.35 (0.06)
0.13 (0.04)
0.06 (0.04)
(0.00) (0.06)
0.09 (0.03)
0.10 (0.03)
0.14 (0.01)
0.34 (0.06)
0.30 (0.04)
0.08 (0.05)
0.27 (0.04)
0.29 (0.03)
0.06 (0.05)
0.06 (0.02)
0.12 (0.02)
0.17 (0.01)
0.16 (0.01)
0.19 (0.09)
Rank
41
324
268
107
212
251
98
170
286
342
126
328
91
18
248
346
363
315
301
232
20
29
320
57
36
344
347
279
184
191
151
24
SIC
3275
3281
3291
3295
3296
3297
3299
3312
3315
3316
3317
3321
3324
3325
3341
3351
3354
3356
3357
3363
3364
3365
3366
3369
3398
3399
3411
3412
3421
3423
3425
OP (NS)
Coeff
S.E.
Rank
0.25 (0.13) 246
0.21 (0.05) 297
0.44 (0.08) 49
0.32 (0.11) 148
0.44 (0.13) 52
0.27 (0.2) 207
0.32 (0.11) 149
0.28 (0.08) 190
0.26 (0.06) 220
0.26 (0.14) 215
0.28 (0.11) 182
0.19 (0.04) 313
0.22 (0.11) 281
0.11 (0.11) 353
0.34 (0.08) 122
0.26 (0.14) 223
0.24 (0.08) 253
0.19 (0.12) 315
0.28 (0.04) 187
0.32 (0.09) 144
0.28 (0.1) 189
0.17 (0.08) 319
0.25 (0.1) 234
0.27 (0.31) 196
0.24 (0.04) 255
0.22 (0.07) 282
0.48 (0.08) 34
0.09 (0.13) 354
(0.17) (0.37) 366
0.20 (0.04) 302
0.25 (0.19) 239
Coeff
0.08
0.17
0.33
0.29
0.39
0.17
0.36
0.41
0.37
0.14
0.25
0.25
0.09
0.19
0.57
0.28
0.14
0.21
0.25
0.22
0.29
0.19
0.16
0.33
0.19
0.33
0.59
0.06
0.10
0.07
0.19
ACF
S.E.
(0.1)
(0.06)
(0.07)
(0.12)
(0.09)
(0.17)
(0.08)
(0.07)
(0.05)
(0.16)
(0.15)
(0.04)
(0.1)
(0.09)
(0.13)
(0.23)
(0.09)
(0.07)
(0.05)
(0.11)
(0.08)
(0.07)
(0.06)
(0.14)
(0.09)
(0.1)
(0.07)
(0.16)
(0.15)
(0.03)
(0.08)
Rank
333
255
112
136
67
252
84
58
78
286
170
168
327
231
23
149
283
208
172
199
134
235
266
106
233
109
22
343
321
336
232
Coeff
0.10
0.14
0.26
0.16
0.20
0.30
0.24
0.22
0.23
0.16
0.19
0.14
0.19
0.09
0.27
0.17
0.17
0.12
0.18
0.20
0.22
0.07
0.01
0.21
0.08
0.15
0.25
0.08
0.21
0.13
0.14
OLS
S.E.
(0.05)
(0.02)
(0.03)
(0.03)
(0.03)
(0.04)
(0.03)
(0.03)
(0.03)
(0.05)
(0.04)
(0.02)
(0.04)
(0.04)
(0.04)
(0.06)
(0.04)
(0.04)
(0.02)
(0.03)
(0.05)
(0.03)
(0.05)
(0.05)
(0.02)
(0.03)
(0.02)
(0.05)
(0.06)
(0.02)
(0.05)
Rank
298
240
69
190
130
31
87
112
88
193
147
238
150
316
51
172
171
273
155
134
105
333
362
116
323
215
75
321
121
265
230
25
SIC
3429
3431
3432
3433
3441
3442
3443
3444
3446
3448
3449
3451
3452
3462
3465
3469
3471
3479
3484
3491
3492
3493
3494
3495
3496
3497
3498
3499
3511
3519
3523
Coeff
0.24
0.21
0.44
0.36
0.27
0.28
0.23
0.29
0.25
0.39
0.26
0.21
0.28
0.26
0.33
0.26
0.23
0.25
0.30
0.28
0.21
0.67
0.14
0.59
0.31
0.33
0.19
0.26
0.31
0.24
0.26
OP (NS)
S.E.
Rank
(0.04) 258
(0.6) 296
(0.32) 47
(0.08) 99
(0.03) 209
(0.04) 184
(0.04) 268
(0.02) 175
(0.04) 241
(0.06) 85
(0.05) 221
(0.04) 295
(0.06) 188
(0.1) 228
(0.07) 131
(0.03) 232
(0.04) 267
(0.05) 247
(0.3) 168
(0.08) 183
(0.16) 293
(0.24) 6
(0.08) 334
(0.07) 13
(0.05) 156
(0.21) 139
(0.06) 309
(0.02) 217
(0.3) 153
(0.06) 257
(0.05) 218
ACF
Coeff
S.E.
0.28 (0.06)
0.37 (0.25)
0.49 (0.33)
0.22 (0.07)
0.30 (0.04)
0.10 (0.06)
0.24 (0.04)
0.29 (0.03)
0.25 (0.03)
0.17 (0.1)
0.32 (0.07)
0.21 (0.06)
0.36 (0.08)
0.18 (0.09)
0.34 (0.07)
0.18 (0.04)
0.18 (0.03)
0.28 (0.05)
0.11 (0.23)
0.17 (0.12)
0.06 (0.21)
(0.08) (0.14)
0.01 (0.07)
0.54 (0.07)
0.37 (0.06)
0.64 (0.18)
0.05 (0.07)
0.19 (0.03)
0.21 (0.19)
0.28 (0.08)
0.14 (0.04)
Rank
148
76
32
202
130
322
186
138
169
254
116
215
82
243
101
242
241
144
312
249
341
366
357
27
80
14
346
237
212
145
282
Coeff
0.14
0.13
0.29
0.21
0.15
0.14
0.11
0.16
0.12
0.15
0.14
0.09
0.15
0.22
0.18
0.12
0.09
0.11
0.21
0.11
0.17
0.13
0.10
0.40
0.17
0.20
0.10
0.14
0.14
0.21
0.14
OLS
S.E.
(0.02)
(0.08)
(0.07)
(0.03)
(0.01)
(0.02)
(0.02)
(0.01)
(0.02)
(0.02)
(0.02)
(0.01)
(0.02)
(0.03)
(0.03)
(0.01)
(0.01)
(0.01)
(0.06)
(0.03)
(0.04)
(0.05)
(0.03)
(0.03)
(0.02)
(0.05)
(0.02)
(0.01)
(0.06)
(0.03)
(0.02)
Rank
237
257
39
119
217
224
294
196
269
222
242
317
220
110
162
275
318
291
118
280
174
247
302
11
177
124
295
244
228
113
243
26
SIC
3524
3531
3532
3533
3534
3535
3536
3537
3541
3542
3543
3544
3545
3546
3548
3549
3552
3553
3554
3555
3556
3559
3561
3562
3563
3564
3565
3566
3567
3568
3569
Coeff
0.45
0.33
0.25
0.24
0.27
0.25
0.34
0.32
0.37
0.37
0.18
0.20
0.30
0.46
0.23
0.34
0.34
0.13
0.38
0.30
0.29
0.31
0.40
0.12
0.21
0.31
0.33
0.24
0.17
0.21
0.30
OP (NS)
S.E.
Rank
(0.16) 38
(0.09) 132
(0.11) 235
(0.04) 263
(0.14) 200
(0.04) 238
(0.16) 124
(0.11) 145
(0.08) 92
(0.1) 91
(0.05) 317
(0.02) 305
(0.04) 164
(0.2) 36
(0.21) 265
(0.07) 118
(0.15) 129
(0.11) 343
(0.13) 89
(0.07) 169
(0.1) 176
(0.03) 162
(0.06) 75
(0.09) 352
(0.08) 292
(0.04) 155
(0.06) 137
(0.07) 249
(0.08) 320
(0.11) 289
(0.04) 171
Coeff
0.23
0.30
0.32
0.17
0.16
0.10
0.14
0.23
0.29
0.31
0.12
0.14
0.25
0.35
0.04
0.15
0.29
0.18
0.53
0.20
0.14
0.19
0.44
0.08
0.34
0.15
0.27
0.30
0.06
0.15
0.16
ACF
S.E.
(0.15)
(0.08)
(0.1)
(0.05)
(0.11)
(0.06)
(0.1)
(0.09)
(0.06)
(0.1)
(0.04)
(0.01)
(0.05)
(0.11)
(0.15)
(0.05)
(0.08)
(0.07)
(0.2)
(0.08)
(0.07)
(0.04)
(0.08)
(0.08)
(0.11)
(0.06)
(0.1)
(0.08)
(0.09)
(0.12)
(0.04)
Rank
192
127
117
250
263
320
290
190
139
124
303
285
174
93
348
277
133
246
28
226
281
234
48
330
98
279
157
129
344
271
261
Coeff
0.08
0.18
0.13
0.12
0.14
0.13
0.21
0.20
0.23
0.23
0.10
0.07
0.16
0.24
0.16
0.16
0.19
0.04
0.26
0.17
0.13
0.15
0.30
0.07
0.16
0.17
0.18
0.22
0.11
0.16
0.17
OLS
S.E.
(0.03)
(0.02)
(0.03)
(0.03)
(0.06)
(0.02)
(0.04)
(0.03)
(0.03)
(0.04)
(0.02)
(0.01)
(0.01)
(0.04)
(0.06)
(0.03)
(0.03)
(0.04)
(0.04)
(0.03)
(0.03)
(0.01)
(0.03)
(0.04)
(0.03)
(0.02)
(0.02)
(0.03)
(0.03)
(0.03)
(0.02)
Rank
329
154
245
278
223
258
123
137
92
99
300
330
203
78
192
199
142
352
68
181
260
211
30
334
194
183
156
108
292
187
182
27
SIC
3571
3572
3575
3577
3578
3579
3581
3585
3589
3592
3593
3594
3596
3599
3612
3613
3621
3624
3625
3629
3634
3643
3644
3645
3646
3648
3651
3652
3661
3663
3669
Coeff
0.65
0.53
0.44
0.44
0.58
0.39
0.41
0.25
0.32
0.39
0.61
0.14
0.14
0.23
0.40
0.39
0.25
0.27
0.31
0.46
0.36
0.24
0.43
0.26
0.38
0.13
0.29
0.26
0.49
0.33
0.35
OP (NS)
S.E.
Rank
(0.12) 8
(0.2) 20
(0.18) 48
(0.07) 46
(0.32) 15
(0.12) 81
(0.41) 61
(0.05) 244
(0.04) 142
(0.2) 82
(0.18) 11
(0.2) 340
(0.12) 335
(0.01) 273
(0.11) 76
(0.06) 77
(0.06) 242
(0.23) 206
(0.1) 154
(0.12) 37
(0.16) 95
(0.08) 254
(0.18) 54
(0.08) 219
(0.08) 88
(0.25) 342
(0.1) 180
(0.08) 216
(0.11) 31
(0.06) 138
(0.17) 111
Coeff
0.97
0.73
0.21
0.62
0.63
0.35
0.31
0.22
0.16
0.00
0.39
0.00
0.13
0.14
0.24
0.32
0.21
0.32
0.14
0.45
0.66
0.20
0.26
0.22
0.36
0.18
0.59
0.11
0.44
0.20
0.24
ACF
S.E.
(0.05)
(0.22)
(0.28)
(0.1)
(0.19)
(0.11)
(0.22)
(0.05)
(0.04)
(0.16)
(0.12)
(0.19)
(0.09)
(0.01)
(0.08)
(0.08)
(0.08)
(0.24)
(0.07)
(0.1)
(0.12)
(0.09)
(0.16)
(0.07)
(0.13)
(0.22)
(0.06)
(0.1)
(0.12)
(0.1)
(0.21)
Rank
1
6
207
17
16
86
122
201
265
360
65
361
298
287
183
115
204
113
289
42
12
222
161
197
83
244
21
316
47
220
180
OLS
Coeff
S.E.
0.44 (0.03)
0.34 (0.08)
0.23 (0.05)
0.27 (0.03)
0.41 (0.08)
0.33 (0.05)
0.27 (0.09)
0.15 (0.02)
0.15 (0.02)
0.02 (0.04)
0.26 (0.04)
(0.09) (0.07)
0.14 (0.05)
0.08 (0)
0.25 (0.03)
0.27 (0.02)
0.16 (0.02)
0.23 (0.06)
0.13 (0.02)
0.23 (0.03)
0.37 (0.05)
0.14 (0.03)
0.25 (0.04)
0.10 (0.03)
0.20 (0.05)
0.18 (0.04)
0.23 (0.04)
(0.01) (0.03)
0.21 (0.03)
0.13 (0.02)
0.11 (0.04)
Rank
8
19
102
58
10
22
52
219
214
360
67
368
233
325
72
49
197
101
262
95
16
234
73
299
125
158
94
365
114
254
288
28
SIC
3671
3672
3674
3675
3676
3677
3678
3679
3691
3694
3695
3699
3711
3713
3714
3715
3716
3721
3724
3728
3731
3732
3743
3751
3792
3799
3812
3821
3822
3823
3824
Coeff
0.13
0.27
0.44
0.13
0.05
0.12
0.29
0.35
0.45
0.26
0.30
0.32
0.20
0.27
0.30
0.16
0.34
0.14
0.26
0.40
0.17
0.23
0.30
0.43
0.16
0.22
0.23
0.41
0.35
0.27
0.43
OP (NS)
S.E.
Rank
(0.38) 346
(0.07) 202
(0.04) 50
(0.08) 344
(0.11) 359
(0.06) 349
(0.09) 173
(0.03) 115
(0.13) 44
(0.05) 222
(0.14) 165
(0.06) 143
(0.06) 304
(0.05) 212
(0.03) 172
(0.05) 323
(0.09) 121
(0.12) 341
(0.08) 224
(0.07) 74
(0.05) 321
(0.04) 266
(0.07) 167
(0.15) 55
(0.05) 324
(0.1) 288
(0.08) 274
(0.14) 60
(0.09) 108
(0.06) 211
(0.12) 53
Coeff
0.16
0.21
0.95
0.14
0.20
0.10
0.12
0.30
0.46
0.27
0.47
0.26
0.22
0.15
0.35
0.15
0.40
0.10
0.17
0.27
0.02
0.17
0.20
0.33
0.11
0.04
0.11
0.30
0.21
0.01
0.20
ACF
S.E.
(0.13)
(0.06)
(0.07)
(0.1)
(0.16)
(0.08)
(0.15)
(0.04)
(0.11)
(0.07)
(0.2)
(0.07)
(0.07)
(0.07)
(0.03)
(0.06)
(0.16)
(0.08)
(0.11)
(0.08)
(0.04)
(0.04)
(0.08)
(0.07)
(0.09)
(0.07)
(0.05)
(0.12)
(0.09)
(0.05)
(0.09)
Rank
259
213
2
291
219
323
308
132
37
151
35
164
203
270
87
274
64
325
253
154
351
256
217
111
314
347
313
128
206
359
227
Coeff
0.16
0.10
0.27
0.05
0.08
0.06
0.23
0.21
0.38
0.14
0.20
0.19
0.13
0.19
0.18
0.10
0.17
0.06
0.16
0.19
0.05
0.13
0.14
0.11
0.07
0.02
0.14
0.19
0.29
0.11
0.24
OLS
S.E.
(0.05)
(0.02)
(0.02)
(0.05)
(0.05)
(0.03)
(0.04)
(0.01)
(0.04)
(0.03)
(0.06)
(0.02)
(0.03)
(0.02)
(0.01)
(0.03)
(0.04)
(0.05)
(0.03)
(0.02)
(0.02)
(0.02)
(0.04)
(0.04)
(0.02)
(0.02)
(0.02)
(0.04)
(0.04)
(0.02)
(0.04)
Rank
198
308
59
351
326
343
100
117
12
229
129
139
263
140
165
305
186
348
200
148
349
246
239
285
338
358
231
143
40
289
83
29
SIC
OP (NS)
ACF
OLS
Coeff
S.E.
Rank
Coeff
S.E. Rank
Coeff
S.E. Rank
3825
0.36 (0.05) 103
0.15 (0.06) 273
0.15 (0.02) 221
3826
0.40 (0.09) 71
0.20 (0.08) 228
0.13 (0.03) 259
3827
0.20 (0.1) 298
0.16 (0.1) 260
0.11 (0.03) 293
3829
0.23 (0.05) 271
0.07 (0.04) 337
0.11 (0.02) 284
3841
0.36 (0.04) 100
0.26 (0.04) 163
0.15 (0.02) 208
3842
0.34 (0.04) 125
0.24 (0.04) 182
0.19 (0.01) 138
3843
0.34 (0.07) 123
0.14 (0.03) 297
0.09 (0.02) 314
3844
0.52 (0.3) 25
0.03 (0.21) 350
(0.01) (0.08) 364
3845
0.66 (0.17) 7
0.68 (0.25) 11
0.38 (0.04) 13
3851
0.15 (0.08) 333
0.19 (0.07) 236
0.12 (0.02) 276
3861
0.50 (0.08) 28
0.31 (0.12) 121
0.30 (0.02) 25
3873
0.45 (0.12) 41
0.17 (0.11) 251
0.22 (0.04) 111
3911
0.27 (0.04) 213
0.26 (0.04) 159
0.14 (0.01) 225
3914
0.35 (0.23) 112
0.39 (0.22) 66
0.02 (0.07) 359
3915
0.21 (0.11) 294
0.11 (0.1) 315
0.07 (0.04) 335
3931
0.16 (0.09) 327
0.15 (0.08) 272
0.12 (0.02) 277
3942
0.31 (0.17) 157
0.39 (0.17) 68
0.30 (0.05) 26
3944
0.40 (0.09) 67
0.34 (0.08) 95
0.17 (0.02) 173
3949
0.28 (0.05) 195
0.21 (0.04) 209
0.17 (0.02) 178
3951
0.31 (0.32) 151
(0.34) (0.36) 368
(0.09) (0.08) 367
3952
0.12 (0.25) 348
0.18 (0.14) 240
0.30 (0.07) 27
3953
0.16 (0.07) 326
0.07 (0.07) 340
0.02 (0.02) 357
3955
0.24 (0.16) 259
0.70 (0.15) 10
0.30 (0.06) 33
3961
0.40 (0.08) 66
0.14 (0.06) 293
0.13 (0.02) 249
3965
0.23 (0.19) 277
0.14 (0.12) 295
0.06 (0.05) 340
3991
0.49 (0.16) 29
0.38 (0.11) 70
0.26 (0.04) 66
3993
0.24 (0.03) 262
0.13 (0.02) 300
0.11 (0.01) 281
3995
0.30 (0.18) 163
0.66 (0.26) 13
0.23 (0.06) 103
3999
0.22 (0.03) 283
0.17 (0.04) 248
0.13 (0.01) 256
OPNS refers to Olley-Pakes with no selection correction; ACF to Ackerberg, Caves, Frazer [2005] using materials to invert, and
current-period variables as instruments
30
Algorithm for ACF Method
Step 1: Regress using OLS, yt on a polynomial of lt, mt, kt and xt. Obtain predicted values of yt, which are unbiased estimates
of αkt+βlt,+λxt +ωt (So, no coefficients are estimated here)
Step 2: Choose some initial set of parameter values for α, β and λ. Using the predicted values from Step 1 and the chosen
parameter values, obtain the estimated ωt.
Step 3: Regress using OLS, the ωt from Step 2 on a polynomial in ωt-1. The residuals from the regression give the estimated ξt.
Step 4: Using the estimated ξt, say ξ*t, construct the empirical equivalent of the moment conditions i.e. (1/N)(Σ k2t(ξ*t)2
2
+x t(ξ*t)2) where N is the sample size.
Step 5: Check if the sum of squares in Step 4 is close to zero. If yes, the parameter values in Step 2 are the estimated
coefficients. If not, change parameter values in Step 2 and repeat till the condition is achieved.
31