Evaluating job training in two Chinese cities

Abstract

Recent years have seen a surge in work on the impacts of active labor market programs for numerous countries. However, little evidence has been presented on the effectiveness of such programs in China. Recent economic reforms, associated with massive lay-offs, and the accompanying public retraining programs make China fertile ground for rigorous impact evaluations. This study uses survey data from the two large industrial cities Shenyang and Wuhan, covering the period 1998 to 2000, to evaluate retraining programs for over 2000 workers two years after they had been observed as displaced and unemployed. Using a comparison group design, this study is, to our knowledge, the first evaluation of its kind in China. The evidence suggests that retraining helped workers find jobs in Wuhan, but had little effect in Shenyang. The study raises questions about the overall effectiveness of retraining expenditures, and it offers some directions for policy-makers about future interventions to help laid-off workers.

Keywords:

• active labor market programs
• job training
• impact evaluation
• propensity score matching
• China

JEL Classifications:

• J23
• J31

Acknowledgements

The authors thank participants at the Eighteenth Chinese Economic Association (UK) Annual Meetings in Nottingham, and an anonymous referee for helpful comments and suggestions. Any remaining errors and omissions are the authors’ own. The views expressed here are those of the authors and should not be attributed to the World Bank or any of its member countries or affiliated institutions.

Notes

Notes

1. See Heckman et al. (1999) for a comprehensive review of impact evaluations in OECD countries; Dar and Gill (1998) for a review of 11 studies covering the United States, Sweden, Australia, Canada and France; Galasso, Ravallion, and Salvia (2001) for a study on the Argentinean Proemplio experiment; Jimenez and Kugler (1987) for a study on Columbia's national in-service training systems; Fretwell, Benus, and O’Leary (1999) for an evaluation of training programs in Hungary, Poland and the Czech Republic; and NEI (2001) for an evaluation of training programs in Bulgaria.

2. This section draws heavily from Bidani et al. (2004, 2005), where all the tables with the background statistics discussed here may also be found.

3. This section draws heavily on Rong (2002).

4. In 1999, laid-off workers in textile enterprises directly affiliated to the central government totaled 600,000, 400,000 in coal mining, 200,000 in armaments, and 200,000 in machinery enterprises. These figures are taken from the presentation entitled ‘Situation of Laid-off Workers in State Enterprise and Policies on Securing their Basic Living Standards and Promoting their Re-employment’ by the Labor Bureau at the Labor Market Policies Seminar in Beijing in May 1999.

5. Survey Report on Employment Situation in Wuhan, 1997, mimeo. The statistics refer to 1996.

6. Presentation by the Shenyang Municipal Labor Bureau on ‘Forcefully Implementing Re-employment Project – Organizing and Facilitating Redundant Workers for Re-employment’ at Labor Market Policies Seminar in May 1999.

7. In an earlier version of this paper, we also examined the effect of training on wages (Bidani et al. 2005). Based on the comments and suggestions of a referee, however, we exclude the training–wage analysis here and, hence, focus exclusively on the employment analysis.

8. To impose common support, the propensity score methods exclude extreme (in terms of their propensity score) observations. See the notes to Table 2 for details.

9. To conserve space, the results discussed here and in the remainder of this section are not reported here; they are available upon request.