Abstract: At present, the Chinese government attaches great importance to unemployment insurance and re-employment. The State Council, the Ministry of Human Resources and Social Security of China issued special documents successively in 2017 and 2019 on returning and subsidizing unemployed workers' skills upgrading. However, unlike endowment insurance and medical insurance, unemployment insurance can only be entrusted by the state and provided by unemployment insurance institutions, so the moral hazard problem is serious. If this kind of problem cannot be curbed in time, it will lead to a large amount of unemployment insurance fund consumption, and reduce the unemployment insurance fund for the re-employment engineering guarantee function. Therefore, the research in this paper is of important theoretical value and practical reference value for using economic means to prevent the moral hazard of unemployed insured workers. Most domestic and foreign literature focuses on the perspective of labor economics and studies the indirect influencing factors of the moral hazard of unemployment insurance. However, realistic unemployment insurance's specific form of moral hazard has not been studied deeply. This paper takes enterprise unemployed insured worker as the research object, combining with the latest relevant incentives of state unemployment insurance, and constructs the compound two-stage multitasking incentive model for unemployed insured workers by basing principal-agent theory which aims at the three typical moral hazards of voluntary unemployment, delayed employment and hidden employment that may occur during the unemployment period of insured workers at the two stages of the qualification review and treatment provided, and uses Matlab2017b simulation to show the relationship between the main influencing factors and optimal incentive factors. In this paper, the key ideas of complex two-stage multi-task incentive model are as follows: On the one hand, Jarque's (2010) expression of income function $S_{t}=\sum_{\tau=1}^{t} \eta^{t-\tau} e_{\tau}$ is used for reference. Among them, the effort level e_τ is further decomposed into avoiding voluntary unemployment, e₁ avoiding delaying employment e₂ and avoiding implicit employment e₃. λ stands for η, which means the influence factor of the efforts at the qualification examination stage on the efforts at the treatment disbursement stage. λ is used to combine the qualification examination stage with the treatment issue stage. The benefits brought to the unemployment insurance institution by the self-discipline efforts of the employees participating in unemployment insurance can be expressed as $S_{t}=\sum_{\tau=1}^{t} \eta^{t-\tau} e_{\tau}e_{1}+\left[\left(\lambda e_{1}+e_{2}\right)^{1-r}+\left(\lambda e_{1}+e_{3}\right)^{1-r}\right]^{\frac{1}{1-r}}$ . On the other hand, the cost function formula $C_3=\frac{1}{2}\left(e_{1}^{2}+e_{2}^{2}+e_{3}^{2}\right)+r_{1} e_{1} e_{2}+r_{2} e_{2} e_{3}+r_{3} e_{1} e_{3}$ of Tang et al(2013)is used for reference. It describes the costs incurred by the self-discipline efforts of unemployed workers to avoid three kinds of moral hazard, adds the two-stage self-discipline factor λ into the model. The cost function becomes $\frac{1}{2}\left[e_{1}^{2}+\left(\lambda e_{1}+e_{2}\right)^{2}+\left(\lambda e_{1}+e_{3}\right)^{2}\right]+r_{1}e_{1}\left(\lambda_{1}+e_{2}\right)+r_{2}\left(\lambda_1+e_{2}\right)\left(\lambda e_{1}+e_{3}\right)+r_{3}e_{1}\left(\lambda e_1+e_{3}\right)$ . The research conclusion shows: First, with the increase of the proportion of unemployment insurance expense by the insured employees when they are on the job, or the increase of the proportion of skill improvement subsidies allocated by the social security institutions, the positive incentives for insured employees' avoiding the three moral hazard self-discipline efforts should be increased; Second, with the increasing influence between self-discipline efforts of the two-stages moral hazard, the positive incentives for insured employees' avoiding voluntary unemployment self-discipline efforts and hidden employment self-discipline efforts should be increased, while the positive incentives for avoiding delayed employment self-discipline efforts should be reduced. Third, under the condition that the proportion of job stabilization returns increases, the positive incentives for avoiding voluntary unemployment and the hidden employment self-discipline efforts should be appropriately reduced, while the positive incentives for the delayed employment self-discipline efforts of insured employees should be increased. Future research can focus on the specific situation of moral hazard of unemployed insured workers and other types of compound incentive models can be further designed according to the specific situation of moral hazard of unemployed insured workers, or moral hazard and adverse selection can be combined to design incentive models.

Key words: unemployed worker, two-stage multitasking moral hazard, incentive mechanism

摘要： 现有文献大多侧重于劳动经济学视角,针对失业保险道德风险的间接影响因素进行研究。并未针对现实的失业保险道德风险问题深入研究。论文基于委托代理理论,以失业参保职工为研究对象,结合国家失业保险的最新政策,针对企业参保职工失业过程中,在资格审查和待遇发放两个阶段,自愿失业、延缓就业和隐性就业三种最可能发生的道德风险行为,构建了失业参保职工两阶段多任务激励模型,并借助Matlab2017b数值仿真展示了主要影响因素与最优激励因子之间的关系。研究得出了若干防范失业参保职工失业道德风险的有益结论和建议。

关键词: 失业职工, 委托代理理论, 两阶段多任务道德风险, 激励机制