The Long-Term Wage Effects of Job Displacement: Frequency or Cumulative Duration of Unemployment

logW_i= ^X_i+ γ₁SUM₋DUR_i+ γ₂FREQ_i+ _i

W_i i

X_i SUM₋DUR

FREQ

RECNT₋DUR REMAIN₋DUR

logW_i= η^X_i+ ₁R EM A IN₋D U R_i+ ₂R E C N T₋D UR_i+ ₃F R EQ_i+ w_i

1

1 2

2 1

REMAIN₋DUR

RECNT₋DUR_i= α^Z_i+ µ₁REMAIN₋DUR_i+ µ₂FREQ_i+ µ₃logW_i+ v_i

Z v

FREQ REMAIN₋DUR

RECNT₋DUR_i

y_1i=D_i y_2i y_3i= y_2i− D_i

y_i=^



y_1i= D_i if y_3i 0 y_2i if y_3i 0

y_i D_i y_2i

y_2i D_i D_i

Prob(y_2i D_i) = 1−F(D_i) f(y_2i) F^(y_2i) = f (y_2i) = αy_2i^α−1e ^Xi−y2iαe

Xi

E (y_2i│X_i) = ^X_i

lnL =_y

Σ

2i Di

ln^1− F (D_i)^€+_y

Σ

2 i Di

lnf (y_2i)

RECNT₋DUR) REMAIN₋DUR)

R²

σ

abstract

On the basis of data from the National Longitudinal Survey of Youth (NLSY), this paper investigates wage consequences of cumulative experience of job displacement. Unlike previous studies, we consider two measures of cumulative unemployment experience simultaneously: the total frequency and the cumulative duration of unemployment induced by job displacement. When frequency and cumulative duration compete in a wage equation, only cumulative duration remains significant for men, while only frequency matters for women. For men, a one-month increase in the cumulative duration of displacement-initiated unemployment leads to a fall in wages by 0.4 percent. This finding is quite robust with respect to various sample restrictions and/or estimation methods. For women, approximately a 2.5 percent wage reduction is associated with an additional event of displacement

Keywords: displacement frequency, cumulative duration, wages, human capital, gender differences