資訊不對稱下的租稅政策

此篇論文包含三篇關於在資訊不對稱下最適租稅政策的文章。第一篇文章在生命循環模型下探討最適資本稅與勞動所得稅，此模型具有高風險的人力資本投資且消費與教育投資是難以區分的特性。社會規劃者面對代理人的外生才能與內生的人力資本具有資訊不對稱性。代理人因此可以兩種方式偏移:一是藉由謊報自己的能力，二是藉由錯置自己的人力資本量。人力資本的錯置無法直接被社會規劃者觀察到，且有一些持續性的效果：早先錯置人力資本的代理人如今將會面對不同的trade-offs與不同的內在動機去誠實呈報。這正是因為所有可能偏移策略的錯綜性與分析的困難度，使得此問題變得複雜。Constrained efficient allocation的扭曲可以藉由正值的capital wedge與在生命循環早期呈現負值，晚期呈現正值的labor wedge所刻劃。這些wedges提供機制去消除因消費無法與教育花費區隔所造成的扭曲。此文中建構一簡單的稅制包含線性資本與勞動所得稅得以在市場經濟下執行constrained efficient allocation。數值模擬的結果指出平均capital wedge在所有工作期數皆為正，capital wedge為正值且隨能力遞減，平均labor wedge在早期為負值晚期為正值，且labor wedge隨能力遞增，即使在中間的工作期數，labor wedge依舊呈現負值在一些低能力的人身上。第二篇文章試圖回答以下問題：當財富在世代間的生前移轉無法被觀察到時，最適稅制需要如何調整？為了鼓勵代理人揭露自己真實的財富，研究發現遺產稅與儲蓄稅需與現今世代和下一代的勞動所得稅值呈高度相關。此結果可透過一個由邊際勞動所得稅率調整過後的modified inverse Euler equation所刻劃。給定下一代的預期勞動所得稅，當現今世代的勞動所得稅足夠低時，邊際遺產稅與儲蓄稅皆為正值。此外，當父母親的時間投資被納入考慮，政府應該補貼父母親在扶養小孩上所花費的時間成本，並增加下一代的勞動所得稅。第三篇文章研究當勞動者的才能是資訊不對稱的情況且中間財為獨占廠商生產並伴隨內生的技術變動下的最適稅制。此研究建構在動態Mirrlees框架的基礎上，採用talent-to-task assignment 模型推導最適政府政策。在此模型中，因為不同類型的勞動投入具有不完全替代的特性，使得wage premium可被任一生產部門的生產投入量所影響。為了壓縮所得差異，此研究指出，政府應該降低高才能者的勞動所得稅，並增加高端生產部門的資本所得稅與降低其研發投資補貼。另外，當無法核實的研發投資被納入考慮時，政府應該增加高才能者的勞動所得稅與高端部門的資本所得稅，並減少低才能者的勞動所得稅與低端部門的資本所得稅作為回應。除此之外，此研究指出政府應該對所有才能與所有部門都降低勞動所得稅與資本所得稅，並提高研發投資的補貼以矯正因獨占所造成的扭曲。此模型亦闡述個人的職涯選擇將會如何被政府政策所影響。

關鍵字

最適資本與勞動所得稅；資訊不對稱；人力資本累積；遺產稅；私人生前移轉；最適研發政策；內生技術變動

並列摘要

This dissertation contains three articles about optimal taxation under asymmetric information. The first paper studies optimal capital and labor income taxes in a lifecycle model with risky human capital when consumption is indistinguishable from education expenses. The planer faces asymmetric information regarding agents’ exogenous abilities and endogenous human capital. Agents can thus deviate in two ways: by misreporting their ability and by mis-investing in their human capital. Mis-investing in human capital is not directly observed by the planner, and has persistent effects: an agent who has invested the wrong amount of human capital in the past will face different trade-offs today, and different incentives to report truthfully. It is precisely this complexity in the possible deviation strategies, and difficulties in characterizing them, that makes the problem so complex. Distortions to constrained efficient allocations are characterized by a positive capital wedge and a labor wedge that is negative early and positive later in the life cycle. These wedges serve as mechanisms to eliminate the distortion to consumption due to its indistinguishability from education expenditure. We construct a simple tax system of linear capital and linear labor income tax rates to implement the constrained efficient allocations in a decentralized economy. Simulation results suggest that the average capital wedge is positive in all working periods, with the capital wedge being positive and regressive in skills, and the average labor wedge is negative in early periods and positive in later periods, with the labor wedge being negative for lower-skill types even till mid-working periods and being progressive in skills. The second paper addresses the following question: How an optimal tax system to adjust when the inter vivo transfers between generations are not observable? In order to encourage agents to reveal their true wealth, we find that the optimal estate tax rate and the saving tax rate are highly related to the value of the labor income tax rate of the current and future generations. This result can be characterized through a modified inverse Euler equation that is adjusted by the marginal labor income tax rate. Given the expected labor income tax of future generation, the marginal estate and saving tax rate are positive when the labor income tax of current generation is sufficiently low. In addition, when the parental time investment is considered, the government should subsidize parental time in raising children but tax the labor income tax of their children. The third paper studies the optimal taxation with asymmetric information on workers’ talent and monopoly firms of intermediate goods with endogenous technical change. A task-to-talent model is used to derive the optimal government policies based on the dynamic Mirrlees framework. In this model, the wage premium can be affected by the amount of each sector’s production inputs because of the imperfect substitutability of each type’s labor effort. To compress the wage inequality, this study suggests that governments reduce the labor taxes of high talent agents and increase capital tax and reduce R D subsidy of top sector in production. Also, when non-verifiable R D investment is considered, governments should increase high talent’s labor tax and top sector’s capital tax and reduce low talent’s labor tax and bottom sector’s capital tax in response. Besides, to correct the distortion caused by monopoly, this study suggest the government should decrease labor tax and capital tax and increase R D investment subsidy for all types or sectors. The model also illustrates how individual’s occupational choice is affected by government’s policy.

並列關鍵字

Optimal capital and labor taxes ； Asymmetric information ； Human capital accumulation ； Optimal estate tax ； Private inter vivo transfer ； Optimal R D policy ； Endogenous technical change

參考文獻

Ales, L., M. Kurnaz, and C. Sleet (2015) Technical change, wage inequality and taxes. American Economic Review 105(10), 3061–3101.

Google Scholar

Cremer, H. and P. Pestieau (2001) Non-linear taxation of bequests, equal sharing rules and the tradeoff between intra- and inter-family inequalities. Journal of Public Economics 79, 35-53.