Table C5 Observed and Selectivity Corrected Wage Differentials CPS 19812000

Wage differential Observed corrected for wage selectivity into Endowment Coefficient Selection

Year differential3 the labor marketb effect0 effectd effecte

1981

0.1587

0.1315

0.0386

0.0929

0.0272

1982

0.1294

0.1318

0.0307

0.1011

- 0.0024

1983

0.1717

0.1991

0.0056

0.1935

- 0.0275

1984

0.1885

0.1902

0.0312

0.1590

-0.0017

1985

0.1706

0.2037

0.0476

0.1561

- 0.0332

1986

0.1948

0.2144

0.0634

0.1510

-0.0197

1987

0.2008

0.2342

0.0591

0.1751

- 0.0334

1988

0.2089

0.1954

0.0436

0.1518

0.0135

1989

0.2512

0.2009

0.0766

0.1243

0.0503

1990

0.2469

0.2605

0.0807

0.1798

-0.0136

1991

0.2649

0.2494

0.0769

0.1725

0.0154

1992

0.2644

0.2901

0.0890

0.2011

- 0.0258

1993

0.2723

0.3114

0.0852

0.2262

- 0.0391

1994

0.2733

0.3337

0.0911

0.2427

- 0.0605

1995

0.2400

0.2768

0.0792

0.1976

- 0.0368

1996

0.2638

0.2642

0.0929

0.1713

- 0.0004

1997

0.2117

0.2090

0.0791

0.1299

0.0027

1998

0.2843

0.2776

0.0884

0.1892

0.0066

1999

0.2941

0.3028

0.0881

0.2147

- 0.0087

2000

0.2809

0.2566

0.0674

0.1892

0.0243

NOTE: The first-stage probit estimation included the following regressors: age; age squared; nonwhite, female, disabled, single household, education, and worked-last-year dummy variables; and nonlabor income. Regressions for 1981 and 1982 do not include a union dummy. Second-stage wage estimations included the following regressors: hour of work; age; age squared; and union, female, single household, nonwhite, education, region, industry, occupation, and government dummy variables. Regressions for 1983 do not include dummies for the service; farming, fishing, and forestry; or the craft occupations due to the absence of representation of disabled workers in these occupations in the sample. (3nd (coefficients from the nondisabled estimation) was used to represent the ''nondiscriminatory'' world since the disabled make up such a small proportion of the whole.

Table C.6 Marginal Effect of Being Disabled on the Probability of

Employer-Provided Health Insurance and of Being Included in the Employer's Pension Plan, CPS

Table C.6 Marginal Effect of Being Disabled on the Probability of

Employer-Provided Health Insurance and of Being Included in the Employer's Pension Plan, CPS

Measure of

earnings not

Measure of earnings included in

included in probit estimation

probit estimation

Year

¿$(HealthIns)

(Pension)

¿$(HealthIns)

(Pension)

¿Disabled

¿Disabled

¿Disabled

¿Disabled

1980

-0.0848

-0.0890

(0.0094)

(0.0098)

1981

-0.0859

-0.0924

(0.0096)

(0.0100)

1982

-0.0888

- 0.0751

(0.0101)

(0.0104)

1983

- 0.0822

- 0.0835

(0.0099)

(0.0102)

1984

- 0.0740

- 0.0677

(0.0098)

(0.0100)

1985

- 0.0861

- 0.0849

(0.0098)

(0.0102)

1986

- 0.0956

- 0.0824

(0.0099)

(0.0103)

1987

- 0.0850

- 0.0901

-0.0243

-0.0356

(0.0100)

(0.0105)

(0.0096)

(0.0101)

1988

- 0.0990

-0.0911

-0.0428

- 0.0402

(0.0099)

(0.0104)

(0.0095)

(0.0100)

1989

- 0.0891

- 0.0820

-0.0315

- 0.0285

(0.0095)

(0.0097)

(0.0091)

(0.0093)

1990

-0.1050

- 0.0900

-0.0415

- 0.0347

(0.0099)

(0.0101)

(0.0095)

(0.0097)

1991

-0.1150

-0.1059

- 0.0626

- 0.0595

(0.0099)

(0.0101)

(0.0095)

(0.0097)

1992

-0.1188

- 0.0967

- 0.0638

- 0.0462

(0.0101)

(0.0101)

(0.0098)

(0.0097)

1993

- 0.0741

- 0.0856

- 0.0329

- 0.0424

(0.0100)

(0.0102)

(0.0097)

(0.0098)

1994

- 0.0831

- 0.0872

- 0.0380

- 0.0450

(0.0107)

(0.0108)

(0.0104)

(0.0105)

1995

-0.1070

-0.1210

- 0.0869

-0.1045

(0.0105)

(0.0108)

(0.0105)

(0.0108)

1996

-0.1159

-0.1281

- 0.0970

-0.1094

(0.0104)

(0.0108)

(0.0104)

(0.0107)

Measure of earnings not Measure of earnings included in included in probit estimation probit estimation

Year

dO(Healthlns)

(Pension)

dO(Healthlns)

dO (Pension)

dDisabled

dDisabled

dDisabled

dDisabled

1997

-0.1101

-0.1110

-0.0913

-0.0944

(0.0109)

(0.0113)

(0.0109)

(0.0112)

1998

-0.1018

-0.1022

- 0.0822

-0.0828

(0.0111)

(0.0115)

(0.0111)

(0.0114)

1999

-0.0728

-0.1036

- 0.0526

- 0.0832

(0.0108)

(0.0111)

(0.0107)

(0.0110)

NOTE: A reliable earnings variable was not available in the years 1981-1987. Re-gressors in the probit estimation included nonwhite, female, education, single household, government, industry, occupation, and disabled dummy variables; age; age squared; earnings from the job; and usual hours worked. Standard errors for the partial derivatives are in parentheses. $ is the standard normal cumulative distribution function.

Table C.7 Impact of Being Disabled on the Probability of Part-Time and Voluntary Part-Time Employment, CPS, 1981-2000

Predicted marginal effect of a disability on probability

Year

Being employed part-timea (1)

Being voluntarily employed part-timeb (2)

1981

0.0557

-0.0030

1982

0.0581

-0.0010

1983

0.0836

-0.0643

1984

0.0665

-0.0098

1985

0.0664

-0.0744

1986

0.0850

-0.0145

1987

0.0594

- 0.0085

1988

0.0820

- 0.0399

1989

0.0924

0.0067

1990

0.0824

-0.0214

1991

0.0844

- 0.0367

1992

0.0886

0.0053

1993

0.1134

0.0173

1994

0.1425

0.0365

1995

0.1227

0.0187

1996

0.1380

0.0266

1997

0.1492

0.0676

1998

0.1407

0.0385

1999

0.1569

0.0594

2000

0.1250

0.0182

NOTE: The marginal effects are calculated at the sample means, except disability status, which is varied between 0 and 1 to generate the marginal effect. See Long (1997).

a Predicted marginal effects in column 1 result from estimation of a bivariate model of employment and part-time employment, accounting for selection at the employment stage. Regressors in both the employment and part-time employment equations included age, education, region, race, gender, marital status, and a central city residence indicator. Regressors unique to the employment equation included the state unemployment rate and the number of weeks worked last year. Regressors unique to the part-time employment equation included occupation and industry dummies, nonlabor income, and a government employer indicator.

b Predicted marginal effects in column 2 result from estimation of a univariate probit model of the probability of being voluntarily employed part-time. The sample is restricted to all part-time workers, so results are not generalizable beyond this sample. Voluntary part-time is defined as did not want to work full time (1994-1999), and the reason for working under 35 hours per week (1981-1993) was coded as 07-15 (see Chapter 4, Note 8, and Stratton 1994). Regressors included age, education, region, race, gender, marital status, nonlabor income, and occupation dummies, industry dummies, a government employer dummy, and an indicator of whether the person worked last year or not.

Was this article helpful?

0 0
Stocks and Shares Retirement Rescue

Stocks and Shares Retirement Rescue

Get All The Support And Guidance You Need To Be A Success At Investing In Stocks And Shares. This Book Is One Of The Most Valuable Resources In The World When It Comes To

Get My Free Ebook


Post a comment