随机数据缺失条件下避税行为的逆威布尔渐近估计

Inverse Weibull Asymptotic Estimation of Enterprise Tax Avoidance Under Condition of Missing Random Data

文章借鉴逆威布尔模型的基本思想,对企业避税行为进行数理推导,研究在数据缺失情况下逆威布尔模型对企业避税金额和避税频率估计参数的渐进一致性,并推导参数的置信区间。证明所构建的逆威布尔模型对企业避税行为周期的估计是渐进无偏的一致估计量,企业避税金额和企业避税频率的蒙特卡洛模拟结果具有较高的可信度。实证结果显示,样本分析期的初期大企业的避税行为高于中小企业,随着我国社会主义市场经济体制的逐步健全和完善,所有企业的避税行为在经历逐年递增阶段后开始出现下降,目前来看中小企业的避税行为要高于大型企业。

This paper borrows the basic idea of inverse Weibull model to perform a mathematical deduction on corporate tax avoidance, and studies the inverse Weibull model＇ s asymptotic consistency of estimation parameters about the amount of tax avoidance and avoidance frequency in the case of missing data. The paper also deduces the confidence interval, and proves that the constructed inverse Weibull model＇ s estimation on the period of corporate tax avoidance is an asymptotic unbiased uniform estimator, and that the Monte Carlo simulation results of the amount of corporate tax avoidance and the frequency are highly credible. The empirical results show that at the beginning of the period large enterprises＇ tax avoidance is higher than that of small and medium-sized enterprises; with the socialist market economy system gradually developing, tax avoidance behaviors of all enterprises begin to decline; at present tax avoidance behavior of small and medium-sized enterprises is bigger than that of large enterprises.