國家行為一致性和信賴保護 ─以國家行為之性質為中心─

Consistency and Legitimate Expectations -Based on the Perspectives of Government Action-

陳偉仁(Wei-ren Chen)

指導教授 ： 許宗力

國立臺灣大學/法律學院/法律學研究所/碩士(2010年)

https://doi.org/10.6342/NTU.2010.01452

摘要

國內文獻針對信賴保護的討論已經呈現爛熟的現象，惟援引美國法者或侷限於其討論之議題，援引德國法者則過分注重法理推演和信賴保護個別要件的操作；真正困難的問題，即如何權衡私人信賴利益和國家政策變遷需求，則始終未見全面性的回應。本文凸顯釋字第五二五號解釋援用撤銷廢止行政處分模型的不當之處，而指出應以各種國家行為性質的不同作為切入角度(第一章)，在綜覽美國法上關於一致性與合法期待的討論之後，分析此處各種衝突的價值衡量，並作為基本的判準(第二章，主要援用L. H. Levinson以及R. Dworkin的架構)；再以國家行為一致性為中心，並設計可能的監督機制，進一步檢討抽象國家行為(第三章，主要援用J. E. Fisch的Equilibrium Approach以及A. F. Cohen, & J. Ravitch之理論)、具體國家行為(第四章，主要援用Reich的New property理論)、以及司法行為(第五章，主要參考英美法上adjudicative retroactivity之討論)。根據前述判準，分析我國法上各個憲法機關面臨信賴保護問題時應有的思考方向，並作結論(第六章)。
本文研究發現如下：如同大多數憲法明文所保障的基本權利一般，在國家公益足夠正當化其干預要求的時候，就算是既得權也必須受到限制。既得權尚且如此，更遑論連法定要件都尚未齊備的信賴利益。但是，不同於一般的基本權，關於既得權和信賴利益的問題，還多了一層法安定性保障的考量；同時，國家管制也添加了一筆「避免既得權和對於過去的信賴利益，阻礙社會進步和改革」的籌碼，來正當化其所加諸人民的限制。
如同Dworkin所言，期待的保障並非終極目的，但是至少對於期待的干預，原則上是不可欲的。相較於Dworkin願意給予期待的擔保實在過於孱弱，本文仍願根據「對於溯及既往的不利推定」和「對於期待的干預原則上不可欲」兩項基本的原則，而主張：除非人民明知法律體系存有漏洞並加以利用，否則信賴和期待，原則上均應受到國家的保障；就前開事項，國家應負舉證責任。本文的依據如下：國家藉由難以預見的公益考量，而要求已經善意投入成本的少部分人民獨自負擔規範或事實變遷的成本，本身就是一件不公平的事情。關於信賴和公益之間的衡量若要更進一步細緻化，其關鍵在於舉證責任和風險分擔的考量。
凖此，若人民不能預見道德或公益等管制動機，則政府事後單方地要求人民獨自負擔管制的後果，並無意義。至於人民能否有預見的可能性，則牽涉到其所擁有的資源、社會地位、政府資訊公開的機制，以及法律諮詢制度的完善程度。而這些因素正是國內對於信賴保護，較少關注的制度性議題。

關鍵字

信賴保護 ； 釋字第五二五號 ； 一致性 ； 合法期待 ； 溯及既往 ； 比例原則

並列摘要

The principle of legitimate expectations is going more crucial in modern society because of the incompatibility between the people’s prediction of governmental action in order to arrange their life and the development of policy which essentially seems not to be stable forever. The aim of the research is to demonstrate Taiwan judicial review of legitimate expectations went astray among specific cases because of the failure to construct a rational and stable pattern to protect legitimate expectations. By pointing out the mistakes of “Administrative Act Revocation Approach” utilized in Interpretation No. 525, this article suggests another model, that is, the principle of proportionality with support of Retroactivity Pattern, for rule-changing review.
In details, there are two suggested models in order to improve the predictability and to strengthen reasonable reliance on legislative.
First, categorization according to different nature of the regulation is necessary. In public law area, in which the relationship is public-private, criminal law cannot be with retroactive effect, which is banned by ex post facto. In addition, laws which are not criminal but have punishment cannot be permitted with retroactive effect, unless the government proves that the regulated party deliberately took advantage of the loophole of law so as to violate “the law”, which did not exist but should have been existent. Besides, if the government makes law with retroactivity on purpose of the compelling interest, the compensation should be appropriated. Finally, laws which are neither criminal nor punishment, such as entitlements and subsidies, can have retroactive or prospective effect according to the facility for policy-making. Since these issues belong to policy concern, political responsibility should also be considered as a checking mechanism.
As regards the civil code, thinking that the relationship is private-private, Congress can form the novel relationship if it will. However, the infringement of vested rights should pass the judicial scrutiny, and the political responsibility should be taken even though the legislative is undergoing reform of society.
Furthermore, the legislative choice of translation clause and remedy should be coped with more carefulness. As a matter of policy, Congress has taken varying approaches when establishing the effective dates of statutes. All approaches are in accordance with the purpose of Congress. For instance, when Congress in the United States adopts statutes changing the federal tax law, it has often adopted one or another of the following effective dates: January 1 of the year of enactment; the date the proposal was initially announced; the date on which legislative committees tentatively agreed on the measure; January 1 of the year following the year of enactment; a multi-year phase-in of the change; "grandfather" rules, exempting items of income and expense arising from transactions entered into before the date of enactment. However, as indicated by Interpretation No. 668, "grandfather" provision is apparently not the best solution for translation period for not only the failure to achieve reform task but its disturbance to the stability of law if litigations are raised after several years of the amendment.
In the arena of Executive, the revocation or abolishment of administrative act can be dealt with Administrative Act Revocation Approach, whereas the alteration of regulations and interpretations is a different issue. Regulations, which are authorized by law and as supplement to law, and interpretations, which give the definition of what law is, are both ruling in a general way. Hence, the constitutionality of regulations and interpretations should be decided in an abstract way, i.e., as the facility of judicial review, the principle of proportionality with support of Retroactivity Pattern is a better option. Moreover, with a practical understanding that in our increasingly complex society, replete with ever changing and more technical problems, Congress simply cannot do its job absent an ability to delegate power under broad general directives. Therefore, judicial review must be lenient and deferent to agencies in fast-change regulation context if we still hope that agencies can respond immediately and act themselves flexibly. Besides, it is possible that the harsher the courts ask, the less standards and information agencies will publicize because they incline to circumvent the criticisms that they disobey their own rules.
As regards Judicial Review and Other Remedies, when the Honorable Justices use whatever approaches for legal change review, eventually they encounter the difficulty stemming from the balancing test, which is dealt under the traditional approach, the principle of proportionality. Indeed, the court complicated the issues raised from legitimate expectations. The modified principle of proportionality, by which the court will raise or lower the standard of review according to different given circumstances, is enough to legal change review. By virtue of applying appropriate standard to review, the categorized object of legislative and the flexibility of agency action can be maintained.
Not all regulations must be reviewed before Constitutional Court. Even though laws and regulations are affirmed by the Honorable Justices, it does not follow that the laws and regulations are constitutional in application. As long as the application of retroactive rules is unfair and unreasonable in specific cases, the district and appeal courts remain their power to intervene and relieve for keeping legitimate expectations safe. On the other hand, the jurisprudence must realize the judicial condemnation is not the only way to defend legitimate expectations, resorting vote and political responsibility is also a possible approach of check and balance. Finally, public officials engaged in the development of policy should still take the responsibility for their decisions even though the legal evolution at issue passes the judicial scrutiny.

並列關鍵字

legitimate expectations ； reliance ； retroactivity ； legal change ； Interpretation No. 525

參考文獻

林秀雄，〈剩餘財產之分配與不真正溯及—最高行政法院九十年度判字第六七一號判決評釋—〉，《台灣本土法學雜誌》，第30期，2002/1，頁61-77。

洪家殷，〈論信賴保護原則之適用—司法院大法官釋字第五二五號解釋評析〉，刊於：《台灣本土法學雜誌》，第27期，2001/10，頁39-55。

陳愛娥，〈信賴保護原則在撤銷授益性行政處分時的適用—行政法院八十六年判字第一二三二號判決評釋—〉，刊於：《台灣本土法學雜誌》，第8期，2000/3，頁15-29。

黃俊杰，〈稅捐優惠之憲法基礎及信賴保護〉，刊於：《台大法學論叢》，第31卷第6期，2002/11，頁183-267。

黃俊杰，〈解釋函令變更與信賴保護——行政法院八十八年度判字第三九○七號判決評釋〉，刊於：《台灣本土法學雜誌》，第20期，2001/3，頁1-13。