Migration, fixed costs, and location-specific amenities: a hazard analysis for a panel of males.

spell W([ft.sub.is]) The average real hourly predicted wage for new

destinations over the resident spell [DELTA]W([t.sub.is]) W([ot.sub.is]) - W([dt.sub.is]) AGE([t.sub.is]) Age at the beginning of resident spell EDU([t.sub.is]) Level of education at the beginning of resident

spell DSLFEMP([t.sub.is]) Dummy variable, equal to 1 if self-employed or a

farmer, 0 otherwise UNION([t.sub.is]) Dummy variable, equal to 1 if an individual is a

union member at the beginning of the spell, 0

otherwise UNEM([t.sub.is]) Average annual unemployment hours in the

resident spell (hour/year) MARR([t.sub.is]) Share of time married in a resident spell CHILD([t.sub.is]) Number of children who are school age at the

beginning of resident spell CRIME([t.sub.is]) Average annual crime rate in the state

corresponding to resident spell, relative to

the U.S. average PARKS([t.sub.is]) Share of state area in state and national parks

corresponding to resident spell, relative to

U.S. average JAN([t.sub.is]) Average (thirty years) January temperature

corresponding to state of the resident spell

relative to the U.S. average JULY([t.sub.is]) Average (thirty years) July temperature

corresponding to residence the state of the

resident spell relative to the U.S. average RACE([t.sub.is]) Dummy variable, equal to 1 if white, 0 otherwise Number of resident

Spells

Completed Censored Symbol Spells Spells W([ht.sub.is]) W([ft.sub.is]) [DELTA]W([t.sub.is]) -0.199 -0.161 AGE([t.sub.is]) 34.473 34.341 EDU([t.sub.is]) 14.043 12.544 DSLFEMP([t.sub.is]) 0.150 0.315 UNION([t.sub.is]) 0.135 0.258 UNEM([t.sub.is]) 70.474 42.855 MARR([t.sub.is]) 0.713 0.796 CHILD([t.sub.is]) 0.766 1.305 CRIME([t.sub.is]) 0.195 -0.472 PARKS([t.sub.is]) -0.116 -0.050 JAN([t.sub.is]) 4.781 1.563 JULY([t.sub.is]) 1.382 0.138 RACE([t.sub.is]) 0.986 0.921 Number of resident 207 1,061

Spells Table 3. Male Wage Equation: Variable Names, Sample Means, and Least-Squares Estimates: Twenty Years of Panel Data Starting in 1968 (Asymptotic t-Values in Parentheses) Symbol Variable Description Dependent variable:

ln([W.sub.iy]/[P.sub.y]) Log of real hourly wage of male i in

year y (ln$/hour) (a) Personal characteristics:

[EDU.sub.iy] Education of individual i in year y

(in years)

[EXP.sub.iy] Experience of individual i in year y

(AGE-EDU-6 in years)

[EXP.sup.2.sub.iy] Experience squared/100

[RACE.sub.i] Dummy variable, equal to 1 if

individual i is white and 0

otherwise Cost of living and amenities:

ln([PLANDA.sub.ky]/[P.sub.y]) Log of real price of land in state k

where male lives in year y ($/acre)

(a)

[URBAN.sub.ky] Proportion of population urban in

state k in year y (percent)

[CRIME.sub.ky] Crime rate in state k in year y

(percent)

[JAN.sub.k] Thirty-year average of January

temperature in state k (degrees F.)

[JULY.sub.k] Thirty-year average of July

temperature in state k (degrees F.) State labor market indicators:

[PJOBGR.sub.ky] Predicted job growth in state k

between years y and y-1

[PURATE.sub.ky] Predicted unemployment rate in

state k in year y

[RSHOCK.sub.ky] Relative employment shock in state

k in year y

[RURATE.sub.ky] Residual unemployment rate in state

k in year y Regional dummies and trend:

[DS.sub.iy] Dummy variable equals 1 if male i

lives in the South and 0 otherwise

[DW.sub.iy] Dummy variable equals 1 if male i

lives in the West and 0 otherwise

[DNC.sub.iy] Dummy variable equals 1 if male i

lives in the North Central region

and 0 otherwise

[TIME.sub.y] Time indicator, 1968 = 1, ..., 1987 =

20

[TIME.sup.2.sub.y] Time squared/100 Number of observations = 15,367 (b)

Sample Estimated Asympt. Symbol Mean Coefficient t-Values Dependent variable:

ln([W.sub.iy]/[P.sub.y]) 2.431 Personal characteristics:

[EDU.sub.iy] 12.709 0.075 (34.07)

[EXP.sub.iy] 23.597 0.051 (21.83)

[EXP.sup.2.sub.iy] 6.581 -0.098 (-19.62)

[RACE.sub.i] 0.929 0.110 (4.73) Cost of living and amenities:

ln([PLANDA.sub.ky]/[P.sub.y]) 1,118.55 0.048 (3.07)

[URBAN.sub.ky] 71.771 0.295 (4.10)

[CRIME.sub.ky] 8.090 0.015 (5.94)

[JAN.sub.k] 33.167 -0.004 (-3.39)

[JULY.sub.k] 75.648 -0.017 (-10.42) State labor market indicators:

[PJOBGR.sub.ky] 0.213 0.060 (7.46)

[PURATE.sub.ky] 6.357 0.043 (7.15)

[RSHOCK.sub.ky] 0.000 0.009 (2.92)

[RURATE.sub.ky] 0.040 -0.005 (-1.06) Regional dummies and trend:

[DS.sub.iy] 0.290 -0.089 (-3.03)

[DW.sub.iy] 0.226 -0.137 (-4.47)

[DNC.sub.iy] 0.272 -0.018 (-0.92)

[TIME.sub.y] 10.123 -0.012 (-1.88)

[TIME.sup.2.sub.y] 1.357 0.008 (-0.29) Number of observations = 15,367 (b) (a) [P.sub.y] = Implicit price deflator for personal consumption expenditure (1987 = 1.00). See Council of Economic Advisors (1993, p. 352). (b) From the 915 individuals observed over twenty years, we obtain 15,367 observations. Table 4. Maximum Likelihood Estimates of Hazard Function for Interstate Migration of U.S. Working-Age Males, Nineteen to Forty-Five Years of Age in 1968: Twenty Years of Migration Experience (Asymptotic t-Values Are in Parentheses)

Constant Hazard Rate Variables (1) (2) INTERCEPT -3.575 -3.990

(-3.47) (-2.09) [DELTA]W -0.215 -0.286

(-2.40) (-1.71) AGE -0.180 -0.199

(-4.16) (-2.13) [AGE.sup.2]/100 0.269 0.308

(4.97) (2.49) EDU 0.135 0.212

(5.93) (4.90) DSLFEMP -1.344 -1.744

(-7.03) (-5.99) UNION -0.507 -0.76

(-2.31) (-2.50) UNEM 0.001 0.002

(4.91) (2.16) MARR -1.139 -1.24

(-2.32) (-2.12) CHILD -0.139 -0.167

(-2.32) (-2.12) CRIME 0.002 0.052

(0.11) (1.59) PARKS -0.060 -0.039

(-1.57) (-0.74) JAN 0.024 0.020

(3.41) (1.82) JULY 0.015 0.022

(1.28) (1.12) RACE 1.693 2.073

(2.98) (3.17) [sigma] [delta] 3.814

(4.47) Log-likelihood -793.5 -755.9 Completed spells 207 207 Total spells 1,268 1,268

Time-Dependent Hazard Rate

Treatment 1 Treatment 2 Variables (3) (4) INTERCEPT -6.614 -6.634

(-7.85) (-11.47) [DELTA]W -0.603 -0.549

(-5.05) (-6.76) AGE 0.271 0.274