* 1)., Heckman Selection. 50.,. 1990 40, -. I.,., (the young old) (active aging). 1/3. 55 60 70.,. 2001 55 64 55%, 60%,,. 65 75%. 55 64 25%, 32%
, 65 55%, 53% (, 2001)... 1998, 8% 41.5% ( 1998). 2002 7.8% 4.0%, 36.9%.,?,.....,...?,,? 20 100, 5 [ 1]. 1980. 30 40 50. 1985 50. 1990. 1990 30 40. 2000 50 20 1.7 30. 50 [ 2]. 30-34
(100). 2) 1990. 1990., 50 1980. 55 90, 60. 55 1997.,,., 1990.,. [ 1] (, 20 24 =100) : ( Labor SIS) [ 2] (, 35 39 =100) 2) ( ) 20.
: ( Labor SIS)??..,? 40.,,??,?,??,? II. 1.
?,..,.?,. (proxy variable).. (Human Capital Theory)., (seniority).,.,,., - (Heterogeneity) (Altonji & Shakatko 1987, Abranham & Farber, 1987).,., (return). (firm- specific human capital) (Topel 1991). (on-the-job-training) (Brown 1989)., (Lazear 1981).,. ( )., (, 1998; 2002).,,.
, (, 1998; 2002). 2. -. (L) (X) (utility). (w) (market wage) (reservation wage). Quinn, Burthause, and Myers(1990).,,....,,. 3.,., 50 60 1990, ( ). (return).,, 50,. 1.. -.
..., (randomly).,., -. Heckman Selection Model. 2.,, -. 3. ( ). (, 2002;, 1997), - 50 ( 1 ). (, 1998; 2002), 1990.,., ( ),,., (bridge job).?,.
. 4.. 5.. III. 1. (Korean Labor and Income Panel Survey) 1 4 (1998 2001 ). (5,000 ) 1 1,,, (longitudinal survey). 4 45. 2001 4 15 (Life Course) ( ), (Schooling History), (Work History) (Labor Market Transitions). < 1>. < 1>
15-29 30-49 50-79 1,380 (46.8) 2,236 (50.5) 1,433 (46.5) 5,049 (48.3) 1,571 (53.2) 2,191 (49.5) 1,647 (53.5) 5,409 (51.7) 13 (.4) 375 ( 8.5) 1,630 (52.9) 2,018 (19.3) 1,435 (48.6) 2,857 (64.5) 1,181 (38.3) 5,473 (52.3) 1,503 (50.9) 1,195 (27.0) 269 ( 8.7) 2,967 (28.4) 2,435 (82.5) 370 (8.4) 13 (.4) 2,818 (26.9) 507 (17.2) 3,870 (87.4) 2,364 (76.8) 6,741 (64.5) 9 (.3) 187 (4.2) 703 (22.8) 899 (8.6). 235 ( 8.0) 2,113 (47.7) 1,764 (57.3) 4,112 (39.3). 2,716 (92.0) 2,314 (52.3) 1,316 (42.7) 6,346 (60.7) 4.14 4.04 3.55 3.92 2.12 2.47 3.23 2.61 198.54 216.54 165.26 195.40 (%) 706 (23.9) 931 (21.0) 969 (31.5) 2,606 (24.9) (%) 153 (5.2) 212 (4.8) 213 (7.0) 578 (5.6) 2,951 (28.2) 4,427 (42.3) 3,080 (29.5) 10,458 (100.0) < 1> ( )
** ** 15-29 30-49 50-79 1,125 (38.1) 3,168 (71.6) 1,404 (45.6) 5,697 (54.5) 9 (.8) 116 ( 3.7) 344 (24.7) 469 ( 8.3) 256 (23.1) 798 (25.4) 176 (12.6) 1,230 (21.8) 47 ( 4.2) 278 ( 8.8) 135 ( 9.7) 460 ( 8.2) 63 ( 5.7) 246 ( 7.8) 82 ( 5.9) 391 ( 6.9) 284 (25.7) 882 (28.1) 322 (23.1) 1,488 (26.4) 172 (15.5) 295 ( 9.4) 128 ( 9.2) 595 (10.5) 276 (24.9) 529 (16.8) 205 (14.7) 1,010 (17.9) 1,107 (19.6) 3,144 (55.7) 1,392 (24.7) 5,643 (100.0) 330 (29.7) 638 (20.4) 144 (10.4) 1,112 (19.8) 298 (26.8) 304 ( 9.7) 41 ( 2.9) 643 (11.4) 220 (19.8) 770 (24.6) 305 (21.9) 1,295 (23.0) 7 (.6) 104 ( 3.3) 343 (24.7) 454 ( 8.1) 257 (23.1) 1,312 (41.9) 557 (40.1) 2,126 (37.8) 1,112 (19.8) 3,128 (55.6) 1,390 (24.7) 5,630 (100.0) 860 (77.7) 1,634 (52.7) 389 (28.6) 2,883 (51.8) / 180 (16.3) 404 (13.0) 245 (18.0) 829 (14.9) 67 ( 6.1) 1,060 (34.2) 724 (53.3) 1,851 (33.3) 1,107 (19.9) 3,098 (55.7) 1,358 (24.4) 5,563 (100.0) 93.74 136.19 107.22 119.30 : **. 2.. 4 OLS Heckman Selection Model. OLS -,.,,. ( ),.,. Heckman(1979)
(sample selection bias),,,. (self- selection) : (parameters).. (1a) Y 1i = X 1i 1 + U 1i, (1b) Y 2 i = X 2i 2 + U 2 i ( i = 1,..., I ) where E ( U j i ) = 0, E ( U j i U j ' i' ' ) = jj ', i = i ' ', = 0, i i' '..., (1a) Y??. E ( Y 1i X 1i ) = X 1i 1 ( i = 1,..., I ).. E ( Y 1i X 1i, sample selection rule ) = X 1i 1 + E ( U 1i sample selection rule ), U 0,.,. Y2 0 Y1 Y1, E ( U 1i X 1i, sample selection rule ) = E ( U 1i X 1i, Y 2i 0)
= E ( U 1i X 1i, Y 2 i - X 2i 2 ). U1 U2, Y1. E ( U 1i X 1i, Y 2 i 0) = X 1i 1 + E ( U 1i U 2 i - X 2i 2 ). Heckman.. Heckman Selection Model. 5 (binomial logit model) (multinomial logit model). /.,, ( ), ( ) 4... logitij = log [ ij / ij ] i, j.,,,,, ( =4) ( =2) logiti2 = log [ i2 / i4 ]. logitik = ik + xi k. ik (parameter) k (regression parameter).,,,,,.
IV. 1. < 2>. 40. 30. < 2> ( :, ) 20 108.9 (418) 87.1 (543) 96.6 (961) 30 158.2 (771) 92.9 (325) 138.8 (1,096) 40 162.4 (570) 87.3 (365) 133.1 (935) 50 136.3 (321) 73.2 (141) 117.0 (462) 60 104.6 (93) 48.0 (60) 82.4 (153) 70 72.8 (8) 40.0 (7) 57.5 (15) 144.0 (2,181) 85.2 (1,441) 120.6 (3,622), (ln(wage)). [ 3] (ln( )) OLS ( ). 40.,, 40,., Heckman Selection OLS.,.,. 50,
.. [ 1] [ 2].. -., -. [ 3] ln( ) 2. 50 < 3>., 50 (linear relationship), -, 50, 45. - 0.28..,.., 50 (coef. - 0.02-0.03). 50,,.
< 3>, 50,,.,, 45 50. < 3> 50 : OLS/Heckman selection model OLS Heckman selection model + 5.94 (0.30) *** 5.66 (0.39) *** - 0.03 (0.00) *** - 0.02 (0.01) * ( ) - 0.02 (0.06) - 0.01 (0.06) 0.19 (0.07) * 0.19 (0.08) * ( ) 0.39 (0.18) * 0.18 (0.28) 0.22 (0.06) ** 0.40 (0.10) *** - 0.28 (0.20) - 0.92 (0.20) *** 0.11 (0.07) 0.27 (0.11) * ( ) - 0.21 (0.47) - 0.41 (0.10) *** 0.07 (0.08) - 0.01 (0.08) - 0.08 (0.09) - 0.12 (0.07) * / - 0.28 (0.11) * - 0.36 (0.10) *** / - 0.19 (0.08) * - 0.24 (0.07) *** 0.03 (0.08) - 0.06 (0.08) ( ) 0.64 (0.08) *** 0.66 (0.08) *** 0.32 (0.09) ** 0.34 (0.11) ** 0.28 (0.14) * 0.35 (0.11) ** 0.37 (0.51) 0.48 (0.16) ** 0.27 (0.06) *** 0.25 (0.07) *** rho - 0.56 (0.15) n 421 sigma 0.51 (0.04) R Square 0.556 lambda - 0.28 (0.09) : + Wald test of indep. eqns. (rho = 0) : chi2(1)= 8.46 Prob> chi2= 0.0036 p <.1 * p <.01 ** p <.001 *** : (KLIPS) 4
V. OLS Heckman Selection Model. 5.?.,, 45 50 ( 4-1, 4-2). 45 ( ). 45 50 50.. 45 50, 50,., < 4> 50. < 5>,, 45. 3), 45 (linear relationship) I, 45 50., 45., 45 50 2 II 3) 45.
., 45 45. 50. 45 50.. 50.,... 50, 50.,,.
< 4-1> ( ) 0.49 (0.12) *** 0.02 (0.02) 2-0.03 (0.01) * ) - 0.76 (0.28) ** - 0.69 (0.24) ** - 0.75 (0.28) ** - 0.69 (0.25) ** 45-0.64 (0.19) ** - 0.41 (0.18) * 45 2 0.06 (0.03) * 0.05 (0.03) * ( ) - 0.48 (0.53) - 0.46 (0.53) 0.12 (0.21) 0.10 (0.21) - 0.48 (0.28) * - 0.50 (0.28) * - 1.24 (0.26) *** - 1.25 (0.26) *** 0.44 (0.37) 0.45 (0.38) - 0.71 (0.24) ** - 0.66 (0.24) ** ( ) 0.71 (0.25) ** 0.74 (0.25) ** 0.38 (0.28) 0.40 (0.29) 0.99 (0.26) *** 1.05 (0.26) *** 1.95 (0.23) *** 1.95 (0.24) ***
< 4-2> ( ) - 0.08 (0.19) 0.05 (0.04) 2 0.02 (0.03) 0.00 (0.00) ) - 0.97 (0.49) * - 0.94 (0.28) ** - 1.02 (0.55) * - 0.71 (0.29) * 45 3.41 (9.46) 3.51 (9.53) 0.63 (0.15) *** 0.61 (0.15) *** 45 0.55 (0.25) * 0.60 (0.24) * 45 2-0.04 (0.02) - 0.04 (0.02) * ( ) - 0.89 (0.39) * - 0.92 (0.40) * - 1.01 (0.60) * - 0.89 (0.60) - 2.04 (1.23) * - 1.85 (1.24) - 1.06 (0.28) *** - 1.06 (0.28) *** - 0.91 (0.65) - 0.81 (0.64) - 0.27 (0.39) - 0.30 (0.39) ( ) 1.39 (0.59) * 1.47 (0.59) * 0.56 (0.82) 0.78 (0.83) 1.30 (0.25) *** 1.33 (0.25) *** 1.91 (0.19) *** 1.90 (0.19) ** :,,,,, 45.. : ( ) * p<.1 ** p<.01 *** p<.001 :, 4, 2001
< 5> logit ( / ) logit ( / ) logit ( / ) 45-0.14 (0.08) * - 0.07 (0.11) 0.07 (0.06) 45-0.08 (0.27) - 0.77 (0.24) ** - 0.47 (0.20) * 45 2-0.03 (0.05) 0.07 (0.03) * 0.07 (0.03) * ( ) 1.78 (0.80) * 0.82 (0.75) - 1.07 (0.60) * 0.08 (0.29) 1.40 (0.28) *** - 0.95 (0.31) ** 0.55 (0.33) * - 0.77 (0.65) - 1.52 (0.44) ** - 0.08 (0.34) - 0.79 (0.54) - 2.12 (0.32) *** 1.33 (0.43) ** 0.46 (0.70) - 0.26 (0.48) 0.49 (0.29) * 0.02 (0.46) - 2.56 (0.39) *** ( ) 0.46 (0.29) - 2.16 (1.06) * 1.24 (0.30) *** 0.17 (0.34) 0.47 (0.45) 0.24 (0.42) - 0.34 (0.36) - 0.29 (0.54) 2.12 (0.31) *** - 0.09 (0.48) 1.00 (0.40) * 2.29 (0.25) *** :,,,,,.. : ( ) * p<.1 ** p<.01 *** p<.001 :, 4, 2001 VI.., 1980., 1990. 1990 1998. 1990 ( 1, 2)... 50...,
. 40..., Heckman Selection. 50. (,2002),,. 1990 1997. 1990, 1990,.,. 1990 40, -..,..
(2001),,. (2000),,, 24,. (2002), :,. (1999), :, 1955-1995, 38 4 1999 12,. (2001), 2001,. (1998),. (2001), 1-4.. (2002),,. (1998a),, 21 1,. (1998b), -,. Abraham, Katharine G. and Farber Henry S.(1987), "Job Duration, Seniority, and Earnings", The American Economic Review, Vol.77 No.3. Altonji, J. and R. Shakotko (1987), "Do Wages Rises with Job Seniority?", Review of Economic Studies, Vol.54. Brown, J. (1989), "Why do Wages Increase with T enure? On- the- Job Training and Life- Cycle Wage Growth observed within Firms", American Economic Review, Vol.79 No.5. Clark, R.C. and N. Ogawa (1992), "The Effect of Mandatory Retirement on Earnings Profiles in Japan", Industrial and Labor Relations Review, Vol.45 No.2. (1997), Transitions from Career and Jobs to Retirement in Japan", Industrial Relations, Vol.36 No.2 Heckman, J. (1979), "Sample Selection Bias as a Specification Error", Econometirca, Vol.47. T opel, R. (1991), " Specific Capital, Mobility, and Wages: Wages Rise with Job Seniority", Journal of Political Economy, Vol.99 No.1. http://laborstat.molab.go.kr ( Labor SIS).