劳动力结构与产业结构调整

Labor Force Composition and Structural Change

根据家庭不变量信息将"城镇住户调查数据"匹配成长时期的个体层面面板数据后发现,不同产业的劳动者构成存在明显差异。统计结果证实,劳动者在产业间的转移概率与个体特征有关。理论模型进而说明,劳动者在产业间的转移决策是个体特征的函数,并可通过multinomial-logit模型估计。实证结论表明,性别和受教育程度是劳动者产业间转移的关键影响因素,当前工作性质还将影响已参加工作劳动者的转移决策,产业结构调整也将在需求面和供给面上引导劳动者在产业间转移。以上结论在更细致的行业层面上和不同时期内具有稳健性。对中国经济结构转型进行微观考察有助于为制定与劳动力结构相匹配的产业政策和选择不同的产业结构调整驱动力提供参考。

Multi-sector neoclassical growth models up to date still largely remain at a highly aggregate level in explaining structural changes and inter-sectoral labor movements. By assuming a homogeneous labor force, they tend to have a limitation in revealing how individual characteristics affect structural transformations. This paper intends to reexamine these theories at a micro level of individual workers. With the 1986 - 2009 Urban Household Survey Data of China＇s National Bureau of Statistics, we match a person over years using the invariant information of his or her family. We find substantial differences in labor force compositions across sectors： the primary sector has the largest share of workers with an educational attainment of junior high school or below and male workers, whereas the tertiary sector has the largest share of workers with higher education levels and females. But the age and work experience compositions do not differ much. We establish a functional relationship between workers＇ individual characteristics and their sector-switch choices based on a theoretical model, and derive an empirically testable multinomial- logit model from the relationship. Individual characteristics are gender, age, educational attainment, danwei type, job type, and work experiences. We notice industrial characteristics can also affect workers＇ sector-switch choices, so we control each industry＇s labor share in SOE and collectively owned firms, productivity growth rate, and an additional fixed effect. We finally control the real GDP per capita of each worker＇s province according to structural change theories. The multinomial-logit model is estimated by the panel data matched from Urban Household Surveys. We find that for people who have already been working, females tend to join the primary sector with a lower probability but the tertiary sector with a higher probability~ a higher education level and vocational education encourage workers to join the tertiary sector~ people with a white collar job also tend to leave the primary sector and join the tertiary sector~ work experiences exert a nonlinear effect, which is firstly negative and then positive, on the probability to join the tertiary sector. For people who start to work for the first time, gender, age and a college or above degree are key factors to determine their sector choices. In particular, females and college graduates are more likely to join the tertiary sector, while age have a nonlinear effect, which is also negative first and positive later, on this probability. As to industrial characteristics, we find that a higher share of SOE and collectively owned firms workers discourages people to leave the primary sector and join the tertiary sector, but these effects are partly mitigated by a productivity growth in the primary sector. In the meantime, a productivity growth in the secondary sector induces people to leave the primary or secondary sector for the tertiary sector, whereas a productivity growth in the tertiary sector does the opposite. Finally, consistent with structural change theories, higher income leads workers to leave the primary sector and join the tertiary sector. With robustness checks, we find these results valid at more detailed industry levels and in different periods. This study sheds light on making industrial policies compatible with the country＇s labor force composition and choosing the proper engine to facilitate the country＇s structural changes.