Remedy vs. Time: Proposed Retroactivity Rules for China’s “Recording and Review” Process
When, if ever, should a decision invalidating a piece of legislation also undo the actions already carried out in accordance with that legislation?
Welcome to a special issue of NPC Observer Monthly, a (mostly) monthly newsletter about China’s national legislature: the National People’s Congress (NPC) and its Standing Committee (NPCSC).
Today, exercising my editorial prerogative, I’d like to discuss a recent coauthored article of mine concerning “recording and review” (R&R) [备案审查]—China’s system of parallel processes for resolving legislative conflicts. The R&R process operated by the NPCSC through its Legislative Affairs Commission (LAC), in particular, also functions as China’s sole mechanism for ex post constitutional review.
R&R is a key focus of my work at Yale Law School’s Paul Tsai China Center. As I have documented on this Substack and elsewhere, there have been remarkable developments in the law and practice of R&R over the last decade. In December 2023, for instance, the NPCSC enacted landmark legislation to codify important prior R&R reforms and to introduce novel elements to improve the process.
That legislation did not answer every question that may arise in R&R, however. One such question concerns retroactivity: When, if ever, should a reviewing body’s decision to invalidate a document also undo the actions—in particular, administrative or criminal punishments—already carried out in accordance with that document? This issue is important yet complex. Too broad a retroactivity rule could destabilize the law and create chaos, whereas one that’s too narrow would not do justice to those who have been harmed by unlawful documents and could dissuade the public from making use of R&R, ultimately undermining the mechanism itself.
At one point, the LAC proposed a blanket rule of nonretroactivity, only to later drop the proposal in late 2019. It acknowledged that the proposal was premature and that retroactivity is a complicated matter that requires a “careful” balancing of the various interests at stake. Resolving the issue has remained a top priority for recent leaders of the LAC’s R&R Office, who continue to emphasize the need for further research, including on foreign practices.
Indeed, many other constitutional review systems have also had to confront the retroactivity issue. The United States Supreme Court, for example, has grappled with the retroactive effect of judicial decisions (including in constitutional cases) in a long line of cases going back to the 1880s. Since then, the Court has twice changed course on retroactivity, and generations of Justices have offered varied rationales for or against different approaches. This long-running U.S. practice could offer rich lessons for Chinese scholars and policymakers as they begin to design retroactivity rules for R&R.
Yet there was no comprehensive survey of U.S. retroactivity doctrine in Chinese literature, so I thought I might try to fill the void. I thus worked with my coauthor Zhu Jiawei, a Ph.D. candidate at Tsinghua University School of Law, to trace the evolution of the Supreme Court’s retroactivity jurisprudence through 15 key cases.1 The resulting article was published last fall in the Journal of Regulations Filing and Review [备案审查研究], a publication overseen by the LAC, and its full text (without footnotes) is now available at this link. (The incumbent director of the LAC’s R&R office reads, before approving, every piece in the Journal, I was told.) The bulk of our article explains the reasoning in those 15 cases and how they have shaped the U.S. retroactivity doctrine over time—a discussion that I don’t imagine readers of this newsletter would find particularly captivating.2
Instead, I’ll focus on the final section of the article, in which we draw on the U.S. practice to propose our own retroactivity framework for China’s R&R. Though short, this part is no less important than the comparative discussion in the article, in my view. For this reason, I’ve translated this part, titled “Concluding Remarks” [余论], in full below (with my coauthor’s consent).
Our proposed framework can be summarized as follows:
A review decision rejecting a document should apply to all non-final cases, which basically refer to cases where no unappealable judgment has been rendered.
In final cases, the retroactive effect of such a decision depends on the type of dispute and document at issue:
A review decision rejecting a substantive rule of criminal law (i.e., a rule specifying what conduct constitutes a crime) should retroactively apply to final criminal cases.
A review decision rejecting a rule of criminal procedure generally should not retroactively apply to final criminal cases, unless application of that rule may have affected the conviction or the fairness of the original trial.
A review decision rejecting a document generally should not retroactively apply to final civil or administrative cases, unless the reviewing body decides otherwise after weighing the relevant factors. But the party whose request for review led to that review decision should nonetheless be allowed to retroactively benefit from it.
If you are interested in the full discussion, please read on.
Unlike the word-for-word translations I normally provide for legal documents, the following translation includes both substantive and stylistic modifications intended to make the excerpt more accessible to a lay English-speaking audience, especially one unfamiliar with the relevant U.S. caselaw. I’ve reformatted the footnotes as if the excerpt were a standalone article; new footnotes are signed “CW.”
V. Concluding Remarks
In sum, the U.S. federal doctrine on the retroactivity of judicial decisions has evolved through roughly three stages. Before the 1960s, federal courts followed the declaratory theory3 and engaged in fully retroactive decisionmaking. Then, during the Warren Court era, out of concern for the law’s stability, they embraced selective prospectivity4 by weighing the relevant constitutional rights and reliance interests. And since the 1980s, the courts have returned to retroactivity in non-final cases, based on an overall consideration of the federal courts’ constitutional role, the fairness of “treating similarly situated litigants similarly,” and the doctrine of stare decisis. But in criminal cases on collateral review,5 the courts have generally declined to apply new decisions retroactively based on the principle of finality. In Danforth [v. Minnesota (2008)], however, the U.S. Supreme Court appears to have re-embraced the declaratory theory, holding that its newest interpretations of the Constitution bind past conduct, while stressing that its retroactivity jurisprudence is primarily concerned with the question whether remedies are available to the party harmed by a constitutional violation.
In short, the U.S. Supreme Court currently rejects a case-by-case approach to retroactivity and instead generally favors full retroactivity in non-final cases and a unified doctrine of “nonretroactivity as the rule, retroactivity as the exception” in final cases. At the same time, the U.S. federal doctrine on the retroactivity of judicial decisions overall takes a remedial approach—from which China could draw lessons as it begins to develop retroactivity rules for R&R. In fact, China has already demonstrated a similar approach to retroactivity issues in several recent R&R cases, including the case of seconding procurators to other procuratorates.6 In addition, the idea of human-rights protection reflected in the remedial approach should be an inherent normative component of [China’s] retroactivity rules. At this stage, we should design separate retroactivity rules depending on whether a case is final. When a review decision has been made but the relevant problematic legislation has yet to be amended, repealed, or annulled, the review decision generally should retroactively apply to all non-final cases,7 based on the need to preserve the uniformity of the legal system. In final cases, by contrast, we should generally adopt the approach of “nonretroactivity as the rule, retroactivity as the exception,” which can be further refined as follows:
First, where, upon review, the substantive rules of criminal law at issue in final criminal cases (chiefly the Supreme People’s Court’s and Supreme People’s Procuratorate’s interpretations of the Criminal Law [刑法]) are found to violate superior legislation, the relevant decisions should apply retroactively. Specifically, Article 253, item 3 of the PRC Criminal Procedure Law [刑事诉讼法]—which allows for retrying [a final case] when “there was truly error in the application of law by the original judgment or ruling”—should, as a normative matter, encompass the foregoing situation, in light of the prohibition of Article 37 of the Constitution on “[u]nlawful detention, or the unlawful deprivation or restriction of a citizen’s physical liberty by other means.”8 Therefore, in the case of local judicial documents on unlawful practice of medicine9 disclosed by the [LAC’s] 2018 annual report, for example, defendants who were criminally punished under the normative documents at issue should be allowed to seek retrial on the basis of [the LAC’s decision].
Second, where, upon review, a rule of criminal procedure at issue in a final criminal case is found to violate superior legislation, the relevant review decision generally should not apply retroactively. But the case should be retried if the relevant rule may have affected the conviction or the fairness of the original trial, given R&R’s function in protecting human rights. Take, for instance, the case of conditional arrests and the case of seconding procurators to other procuratorates, disclosed respectively by the 2017 and 2022 annual reports. Although the [LAC] ruled that the relevant procedures violated superior legislation in both cases, because [the violations were not so serious] as to affect the fairness of the trials or the substantive outcomes of the cases, retrials should not be allowed if not otherwise permitted by the rules of criminal procedure.
Third, in a civil or administrative dispute (or case) where the statute of limitations has expired or that has become final, [the party who may benefit from a new review decision] generally should not be allowed to file suit, apply for administrative reconsideration, or seek retrial on that basis, based on the need to maintain the law’s stability—except in cases where the reviewing body decides to allow retroactive application based on an overall consideration of various factors. Key factors can include the relevant [private rights] at issue, the public interest, and the social impact [of retroactive application].10 In addition, the reviewing body should clearly explain the specific factors it has considered in the relevant review opinion. In contrast to the U.S. model of judicial review, it is primarily the legislatures that conduct review in China’s R&R system, so granting the reviewing body such discretion [to determine whether a review decision should be retroactive] will not raise concerns that courts are intruding on legislative power, wherein lies the approach’s legitimacy.
In designing the foregoing rules, a special case should be taken into account: review decisions that fail to satisfy the reviewing body’s multifactor [discretionary] test should, as an exception, still retroactively apply to the parties to originating cases [原因案件]11 that have become final to ensure that they receive the relief and incentives they deserve. Looking overseas, aside from the Warren Court’s selective retroactive application of new rules only to the parties to originating cases, countries like Austria that carry out abstract constitutional review also grant special preferential treatment to the parties to such cases by allowing for retroactivity. At a time when China’s R&R is transitioning from a “duck paddling in water” [鸭子凫水]12 to a “ship riding the waves,” such preferential retroactivity not only would solve the kind of dilemma faced by the petitioner in the Pan Hongbin case, where he obtained no actual relief despite a favorable decision by the reviewing body,13 but would also help boost citizens’ enthusiasm for requesting review. Granted, such preferential treatment would lead to some degree of inequity because other similarly situated parties (i.e., those who [have not requested review through R&R and] would be barred by our proposed rules from seeking retrial) would not similarly benefit from retroactivity. But this exception would only afford the parties to originating cases a special [procedural] opportunity [to have their cases retried], without necessarily [substantively] changing the original rulings. Besides, our proposed framework would grant the reviewing body ample discretion to balance the interests in individual cases and to decide to universally apply certain review decisions retroactively, thereby alleviating the inequity problem to some extent. Under the new Decision on Improving and Strengthening the System of Recording and Review [关于完善和加强备案审查制度的决定], “[w]hen finding that regulations or judicial interpretations contravene the Constitution or laws or have issues of constitutionality or legality” in their work, local courts “may report them up through the levels to” the Supreme People’s Court for the latter to submit requests for review [to the NPCSC]. By connecting the judicial process to R&R for the first time, this provision shows there is now a statutory basis for the parties to originating cases to potentially benefit from their successes under R&R.14
In brief, as Director of the Office for the Recording and Review of Regulations of the NPCSC Legislative Affairs Commission Yan Dongfeng has said, the rectification of a normative document should generally have no retroactive effect on any concrete case it implicates, but if the case involves a party’s important rights, then the issue should be handled carefully and prudently. Consistent with this basic principle, this Article makes a first attempt at constructing the foregoing retroactivity framework, based on our review of and observations on the U.S. doctrine on the retroactivity of federal judicial rulings. Retroactivity is, moreover, an issue that has wide-ranging implications for the whole R&R system. To develop a system of retroactivity rules not only requires improving the R&R system itself, such as by strengthening reason-giving,15 reforming the way in which cases are disclosed,16 and improving the case-guidance system, but also requires making corresponding changes to the legislation in areas such as litigation procedures, administrative procedure, and state compensation.17 The relevant issues await further discussions among members of the academy and those who administer R&R in practice, in order to explore R&R-related institutions that are suitable for China’s reality by in part “drawing on the achievements of foreign law-based governance.”18
These cases are (in order of their appearance in the article): Norton v. Shelby County (1886); United States v. Schooner Peggy (1801); Chicot County Drainage District v. Baxter State Bank (1940); Linkletter v. Walker (1965); Stovall v. Denno (1967); Chevron Oil Co. v. Huson (1971); United States v. Johnson (1982); Griffith v. Kentucky (1987); Teague v. Lane (1989); Edwards v. Vannoy (2021); American Trucking Ass’ns v. Smith (1990); James B. Beam Distilling Co. v. Georgia (1991); Harper v. Virginia Department of Taxation (1993); Reynoldsville Casket Co. v. Hyde (1995); Danforth v. Minnesota (2008).
But if you think you might, I recommend checking out these articles: Richard S. Kay, Retroactivity and Prospectivity of Judgments in American Law, 62 Am. J. Comp. L. Supp. 37, 45–62 (2014); Stephen J. Hammer, Note, Retroactivity and Restraint: An Anglo-American Comparison, 41 Harv. J.L. & Pub. Pol’y 409, 413–26 (2018); Samuel Beswick, Retroactive Adjudication, 130 Yale L.J. 276, 293–98 (2020).
Most commonly associated with the eighteenth-century English jurist William Blackstone, the declaratory theory holds: “[J]udges have no power to constrain the retroactive nature of their judgments because judging is an inherently backward-looking exercise. The judge’s role is to resolve disputes that arose in the past, and in so doing, the judge cannot escape making determinations on what the law was (past-tense) that governed.” Beswick, supra note 2; at 130. —CW
“Selective prospectivity” means “apply[ing] a new rule in the case in which it is pronounced, then return[ing] to the old one with respect to all others arising on facts predating the pronouncement.” James B. Beam Distilling Co. v. Georgia, 501 U.S. 529, 537 (1991) (opinion of Souter, J.). —CW
Collateral review, such as federal habeas corpus proceedings, means judicial review of a criminal conviction after direct appeals (up to the U.S. Supreme Court) have been exhausted. —CW
This case is summarized in Part III of this post. Very briefly, the LAC decided that a procurator seconded to another procuratorate cannot appear in court on the latter’s behalf without first receiving a new appointment from the local people’s congress overseeing that other procuratorate. The Supreme People’s Court and the Supreme People’s Procuratorate later announced that criminal defendants cannot seek retrial (whether their case is final or not) on the basis of that LAC decision alone—that is, the new rule announced by the LAC would have prospective effect only. —CW
Here, “non-final cases” encompass not only cases that, in a typical sense, are in progress and not yet final, but also cases like civil disputes that are still within the statute of limitations and administrative disputes that are still within the limitations period for seeking administrative reconsideration or filing an administrative lawsuit. For the latter category of cases, the parties should be allowed to seek administrative reconsideration or file suit based on a review decision in their favor.
See 梁洪霞 [Liang Hongxia], 论合宪性审查溯及力的规范模式与裁量方法 [On the Normative Model of and Discretionary Approach to the Retroactivity of Constitutional Review], 当代法学 [Contemp. L. Rev.], no. 5, 2022, at 99, 108.
At issue in this case were two documents issued by the criminal justice authorities of Zhejiang Province, under which nonmedically necessary prenatal sex discernment was punishable as the crime of unlawful practice of medicine in some circumstances. Upon review, the LAC decided that those documents were functionally judicial interpretations that only the Supreme People’s Court and the Supreme People’s Procuratorate may promulgate. —CW
In addition, many scholars have also proposed the factors that should be part of the reviewing body’s overall consideration. See, e.g., Liang, supra note 8, at 105–07; 孙波 [Sun Bo], 论规范性文件备案审查结果的溯及力 [On the Retroactive Effect of the Results of Recording and Review Normative Documents], 政治与法律 [Pol. Sci. & L.], no. 1, 2021, at 107, 115–17.
Unlike in the U.S. context [where an “originating case” has been used to refer to a case where a court announced a new rule of constitutional law], in the context of R&R, an “originating case” refers only to a final court case in which a party believes that a normative document applied in the case requires rectification, so recommends review of the document through R&R. See 梁洪霞 [Liang Hongxia] & 刘方洲 [Liu Fangzhou], 人权保障视域下备案审查溯及力制度的完善 [Improving the Retroactivity System for Recording and Review from the Perspective of Protecting Human Rights], 中共青岛市委党校 青岛行政学院学报 [J. Party Sch. CPC Qingdao Mun. Comm. | Qingdao Admin. Inst.], no. 1, 2023, at 100, 104.
“鸭子凫水” is the preferred official metaphor for the state of R&R before 2017, when the NPCSC began disclosing R&R data and cases to the public. The idea is that, before 2017, R&R did operate in practice but was invisible to the public, the same way that a duck’s feet remain hidden underwater while it swims. —CW
See 梁洪霞 [Liang Hongxia], 备案审查的人权保障功能及其实现路径——潘洪斌案的再思考 [The Function of Recording and Review in Safeguarding Human Rights and the Path to Realizing It: Rethinking the Pan Hongbin Case], 人权 [Hum. Rts.], no. 2, 2020, at 64, 65. [In that case, Pan’s e-bike was seized by Hangzhou’s traffic police and then towed back to its place of registration (a different city) at his own expense, in accordance with a municipal regulation that the LAC later ruled unlawful at Pan’s request. Because at that point Pan had already lost his administrative lawsuit against the traffic police and exhausted all appeals, and because the courts refused to reopen the case on the ground that the relevant provision’s repeal was not retroactive, he never received the compensation he was due.]
See generally 梁洪霞 [Liang Hongxia], 权利救济视角下备案审查与司法诉讼的制度衔接——以《备案审查决定》新增条款为基础 [The Institutional Linkage Between Recording and Review and Judicial Proceedings from the Perspective of Rights Remedy—Based on the New Provisions in the Decision on Recording and Review], 人权法学 [J. Hum. Rts. L.], no. 3, 2024, at 57. —CW
See 李雷 [Li Lei], 人大备案审查结论存疑的优化机制探究 [Research on the Mechanisms for Optimizing Doubts About the R&R Conclusions of the People’s Congresses], 法学 [Law Sci.], no. 12, 2021, at 55, 63–65.
See Liang & Liu, supra note 11, at 103; 黄明涛 [Huang Mingtao] & 张梦奇 [Zhang Mengqi], 全国人大常委会备案审查决定的先例约束力 [The Precedential Binding Effect of the NPC Standing Committee’s R&R Decisions], 苏州大学学报（法学版） [J. Soochow U. (L. Ed.)], no. 4, 2022, at 1, 11.
See Liang, supra note 8, at 107–09.
习近平 [Xi Jinping], 坚定不移走中国特色社会主义法治道路 为全面建设社会主义现代化国家提供有力法治保障 [Fully Commit to a Path of Socialist Rule of Law with Chinese Characteristics and Provide Powerful Legal Support for Building a Modern Socialist Country in All Respects], 求是 [Seeking Truth], no. 5, 2021, at 4, 8.