经理人显性的货币薪酬具有一定的内在结构，不考虑权益性薪酬，经理人的年薪一般由基本薪酬、绩效薪酬和超额薪酬构成，不同的组成部分拥有不同的效用。基本薪酬是无风险薪酬，与经理人劳动成果无关，提供安全的功能；绩效薪酬是与业绩挂钩的薪酬，属于风险薪酬；经理人的自利本能和委托人的有限监管能力形成了超额薪酬，是需要被限制的薪酬。在职消费是指经理人工作中的消费。文章区分经理人薪酬的内在结构，研究其与隐性薪酬在职消费的关系。为保证研究结论的可靠性，文章采用全样本和小样本两套数据进行检验。全样本数据利用断点回归的思想。假设略微盈利和亏损的公司经理人才能相同，略微盈利公司经理人获得绩效奖金，而略微亏损公司经理人只收获基本薪酬。文章选取ROA亏损和盈利2%以下的样本按照行业、年份和资产规模进行配对，略微亏损样本中高管薪酬低于与之相配对的盈利样本中高管薪酬的薪酬数据，作为高管基本薪酬的代理变量。同时使用高管的薪酬与企业业绩之间的薪酬业绩敏感性指标，作为绩效薪酬的代理变量。薪酬前三高管的薪酬总额对相应公司特征进行回归后的残差，作为超额薪酬的代理变量。另外，根据《高级管理人员薪酬管理办法》手工收集基本薪酬、绩效/基本薪酬比例以及超额薪酬三个薪酬结构数据，形成150个的小研究样本。通过两套数据的回归分析发现，在职消费越高，经理人的基本薪酬越低，绩效薪酬比重也越小；在职消费越高，经理人的超额薪酬也越高。基本薪酬、绩效薪酬与在职消费的关系在国有企业和民营企业间不存在显著差异。国有企业中在职消费与超额薪酬的正相关关系要显著大于民营企业。经理人薪酬收益构成复杂，对经理人的投入产出效用也各不相同。当在职消费这种隐形薪酬与货币薪酬体现为安全和激励功能时，二者可以部分替代；当表现为代理性质时，二者重叠，存在互补关系，而且代理成本叠加现象在国企中更加严重。研究结论说明在委托人代理人薪酬博弈条件下，代理人行为具有一致性。对待自利的经理人，改变薪酬模式不能减少其侵占行为。文章从货币薪酬的结构入手，分析其与在职消费的关系，具有一定的创新。

The explicit monetary compensation of managers has a certain internal structure. The annual salary of managers is generally composed of base salary, merit salary and exceeded compensation excluding equity compensation which have different functions. The base salary is a risk-free salary, which don't connect with the manager's labor results and provides a safe function; merit salary is linked to performance and belongs to risk salary; the manager's self-interest instinct and the principal's limited supervisory ability form exceeded compensation which need to be restricted. Perks refers to the consumption of managers at work. This article decides to distinguish the internal structure of manager compensation and study its relationship with implicit salary-perks. In order to ensure the reliability of the research conclusions, this paper adopts two sets of data for full sample and small sample for testing. The full sample data uses the idea of breakpoint regression. Assuming that only slightly profitable and loss-making company managers have the same ability, slightly profitable company managers receive performance bonuses, and slightly loss-making company managers only receive base salary. This article selects samples with ROA losses and profits below 2% to match them according to industry, year, and asset size. The salary of executive compensation in the slightly loss sample is lower than in the matching profit sample, then the loss sample's salary as a proxy variable for the base salary. At the same time, the performance sensitivity index between executive compensation and corporate performance is used as a proxy variable for merit salary. The regressing residual of the total compensation of the top three executives with the company characteristics is used as a proxy variable of the exceeded compensation. In addition, according to the disclosure "Senior Management Salary Management Rules", base salary, merit/base salary ratio, and exceeded compensation were manually collected to form a small sample of 150. Through the regression analysis of the two sets of data, it is found that the higher perks accompanied with the lower base salary and merit salary of the manager, and the higher exceeded compensation. If perks and monetary compensation conduct the safety and incentive function, they can be partially substitution. If they are all agency cost, there is a complementary relationship, and it's great in the state-owned enterprises. There are both substitution and complement between perks and monetary compensation which based on the internal composition of monetary compensation. The conclusions of this paper indicate that changing the salary model alone cannot reduce manager's encroachment. The innovation of this paper is to analyze the relationship between the structure of monetary compensation and perks.

F272.92;F832.51

国家社会科学基金项目（18BJY016）