Temporary help service firms' use of employer tax credits: implications for disadvantaged workers' labor market outcomes.

Matched 0.746 0.746 Education Unmatched 1.706 1.616

Matched 1.694 1.701 Black Unmatched 0.435 0.469

Matched 0.435 0.432 Hispanic Unmatched 0.060 0.052

Matched 0.059 0.057 Other race Unmatched 0.196 0.168

Matched 0.196 0.196 Total kids under Unmatched 1.042 1.255

6 years of age Matched 1.049 1.062 Total kids ages Unmatched 0.896 0.901

6-17 Matched 0.910 0.902 Milwaukee Unmatched 0.607 0.609

Matched 0.606 0.600 Firm "size" in Unmatched 15.826 23.430

1000s Matched 15.949 16.777 WOTC target Unmatched 5.943 6.214

group Matched 6.002 5.963 Occupation Unmatched 3.119 2.552

Matched 3.106 3.038 Conditioning Test Results: Initial Bias and % p-Value Variable Reduction in Bias after Matching (Matched) Age % Initial bias 15.6 0.972

% Reduction in bias 98.4 Female % Initial bias -31.3 0.99

% Reduction in bias 99.7 Education % Initial bias 13.8 0.87

% Reduction in bias 92 Black % Initial bias -6.9 0.937

% Reduction in bias 92.3 Hispanic % Initial bias 3.5 0.927

% Reduction in bias 81.9 Other race % Initial bias 7.3 0.992

% Reduction in bias 99.1 Total kids under % Initial bias -19.4 0.86

6 years of age % Reduction in bias 94 Total kids ages % Initial bias -0.4 0.93

6-17 % Reduction in bias -57.4 Milwaukee % Initial bias -0.4 0.852

% Reduction in bias -255.7 Firm "size" in % Initial bias -32 0.538

1000s % Reduction in bias 89.1 WOTC target % Initial bias -9.5 0.839

group % Reduction in bias 85.7 Occupation % Initial bias 65.7 0.221

% Reduction in bias 87.9 This balancing test follows the propensity score matching estimation of THS employment effects. The interpretation of the p-values is that low values (such as those below 0.10 or 0.05) suggest that there is remaining selection bias for a given variable even after matching. Table A2. Balancing Test Results for Selection into WOTC Certification

Mean of Conditioning

Variable Conditioning Variable Sample Treatment Comparison Age Unmatched 25.812 25.044

Matched 25.735 25.668 Female Unmatched 0.84688 0.84937

Matched 0.84395 0.84699 Education Unmatched 1.6719 1.46

Matched 1.6656 1.6698 Black Unmatched 0.39687 0.60083

Matched 0.40446 0.39325 Hispanic Unmatched 0.075 0.09688

Matched 0.07643 0.07524 Other race Unmatched 0.20313 0.17741

Matched 0.20701 0.20059 Total kids under Unmatched 1.1844 1.514

6 years of age Matched 1.1911 1.2213 Total kids ages Unmatched 1.0313 1.2035

6-17 Matched 1.0287 1.0178 Milwaukee Unmatched 0.58437 0.79525

Matched 0.59554 0.57733 Firm headquarters Unmatched 0.5625 0.66162

in Wisconsin Matched 0.56688 0.57038 Firm "size" in Unmatched 15.048 18.806

1000S Matched 14.87 14.923 Average earnings/ Unmatched 1324 1027.1

quarter at firm Matched 1316 1273.9

last ear

Test Results: Initial Conditioning Bias and % Reduction p-Value Variable in Bias after Matching (Matched) Age % Initial bias 11.1 0.908

% Reduction in 91.4

bias Female % Initial bias -0.7 0.916

% Reduction in -22.1

bias Education % Initial bias 34.3 0.935

% Reduction in 98

bias Black % Initial bias -41.6 0.775

% Reduction in 94.5

bias Hispanic % Initial bias -7.8 0.955

% Reduction in 94.5

bias Other race % Initial bias 6.5 0.842

% Reduction in 75.1

bias Total kids under % Initial bias -28.5 0.726

6 years of age % Reduction in 90.8

bias Total kids ages % Initial bias -11.5 0.924

6-17 % Reduction in 93.7

bias Milwaukee % Initial bias -46.8 0.644

% Reduction in 91.4

bias Firm headquarters % Initial bias -20.4 0.93

in Wisconsin % Reduction in 96.5

bias Firm "size" in % Initial bias -22.3 0.965

1000S % Reduction in 98.6

bias Average earnings/ % Initial bias 75.7 0.265

quarter at firm % Reduction in 85.8

last ear bias This balancing test follows the propensity score matching estimation of WOTC certification effects. The interpretation of the p-values is that low values (such as those below 0.10 or 0.05) suggest that there is remaining selection bias for a given variable even after matching.

We thank the W. E. Upjohn Institute for Employment Research and the University of Wisconsin Graduate Research Fund for financial support for this research. We are grateful to the Wisconsin Department of Workforce Development and the Institute for Research on Poverty for facilitating our access to administrative data and for computer programming support. We also thank Karl Scholz, Mark Killingsworth, and attendees of Economics Department seminars at the University of Florida and the University of South Florida for helpful comments. Additional thanks go to participants in sessions at the Econometric Society 2006 North American Winter Meetings (Welfare and Active Labor Market Policies session) and the 2005 Institute for Research on Poverty Low Income Summer Workshop.

Received October 2006; accepted May 2007.

References

Altonji, Joseph G., Todd E. Elder, and Christopher R. Taber. 2005. Selection on observed and unobserved variables: Assessing the effectiveness of catholic schools. Journal of Political Economy 113:151-84.

Angrist, Joshua D., and Alan B. Krueger. 1999. Empirical strategies in labor economics. In Handbook of labor economics 3, edited by Orley C. Ashenfelter and David Card. Oxford: Elsevier, pp. 1277-366.

Autor, David. 2003. Outsourcing at will: The contribution of unjust dismissal doctrine to the growth of employment outsourcing. Journal of Labor Economics 21:1-42.

Autor, David, and Susan Houseman. 2005. Do temporary help jobs improve labor market outcomes for low-skilled workers? Evidence from random assignments. NBER Working Paper No. 11743.

Barker, Kathleen, and Kathleen Christensen. 1998. Contingent work: American employment relations in transition. Ithaca, NY: Cornell University Press.

Blank, Rebecca M. 1998. Contingent work in a changing labor market. In Generating jobs: How to increase demand for less-skilled workers, edited by Richard B. Freeman and Peter Gottschalk. New York: Russell Sage Foundation, pp. 258-94.

Carre, Francoise J. 1992. Temporary employment in the eighties. In New policies for the part-time and contingent workforce, edited by Virginia L. duRivage. New York: ME Sharpe, pp. 45-88.

Cohany, Sharon R. 1998. Workers in alternative employment arrangements: A second look. Monthly Labor Review 121:3-21.

Hamersma, Sarah. 2003. The work opportunity and welfare-to-work tax credits: Participation rates among eligible workers. National Tax Journal 56:725-38.

Hamersma, Sarah. 2007. Why don't eligible firms claim hiring subsidies? The role of job duration. Journal of Policy Analysis and Management. In press.

Hamersma, Sarah. 2008. The effects of an employer subsidy on employment outcomes: A study of the work opportunity and welfare to work tax credits. Unpublished paper, University of Florida.

Heckman, James J., Hidehiko Ichimura, Jeffrey Smith, and Petra Todd. 1998. Characterizing selection bias using experimental data. Econometrica 66:1017-98.