The Use of Demonstration Projects to Advance Criminal Procedure Reform in China THE USE OF DEMONSTRATION PROJECTS TO ADVANCE CRIMINAL PROCEDURE REFORM IN CHINA Thomas Stutsman Author Robert L. Bernstein Fellow in International Human Rights, Vera Institute of Justice; Visiting Scholar, Harvard Law School. I would like to thank Ira Belkin, Maggie Lewis, Liu Sida, Carl Minzner, and Katherine Zhao for reviewing earlier drafts of this article. I would also like to thank Megan Golden, Guo Zhiyuan, Jim Parsons, and Monica Thornton, who in addition to helping shape this article, have been immensely influential in my understanding of justice reform in both the United States and China. Finally, I would like to express my gratitude to the Bernstein Family for its generous support of the Robert L. Bernstein Fellowship in International Human Rights. INTRODUCTION 335 I. THE ROLE OF ACADEMIC RESEARCH IN CRIMINAL PROCEDURE REFORM AND THE DEVELOPMENT OF DEMONSTRATION PROJECTS 338 A. The Role of Academic Research in Criminal Procedure Reform 339 B. Approaches to Academic Research and Criminal Procedure Reform 340 C. The Scope of Demonstration Projects on Criminal Procedure Reform 343 D. An Overview of Two Representative Demonstration Projects 344 1. Interrogation Procedures 344 2. Bail and Discretionary Non-Prosecution 346 II. THE ADVANTAGES OF USING DEMONSTRATION PROJECTS TO PROMOTE CRIMINAL PROCEDURE REFORM 348 A. Test a New Approach 349 1. Effectiveness 349 2. Feasibility 350 3. Additional Supporting Measures 353 B. Objective Information About Criminal Justice Practice 354 C. The Debate over Reform 355 D. Stakeholders' Interests and Opposition to Reform 357 E. Non-State Criminal Justice Actors' Influence over Criminal Procedure Reform 358 III. THE CHALLENGES OF CARRYING OUT DEMONSTRATION PROJECTS 359 A. Cooperation of Criminal Justice Institutions 359 B. Academic Resources and Capacity 361 334 Columbia Journal ofAsian Law, Vol. 24, No. 2 (2011) CONCLUSION 362 APPENDIX 1: SCHOLAR-LED DEMONSTRATION PROJECTS ON CRIMINAL PROCEDURE REFORM 364 The Use of Demonstration Projects to Advance Criminal Procedure Reform 335 INTRODUCTION American legal scholars and criminal justice officials widely recognized that bail systems in the United States were deeply flawed in the early 1960s.' Judges released defendants primarily according to their ability to post bail rather than their risk of fleeing the jurisdiction or committing another offense before trial.' Poor defendants were more likely to be detained before trial-as well as lose their jobs, be convicted of an offense, and be sentenced to prison as a result of their pretrial detention-simply because they were poor.' But despite widespread recognition of these injustices, few reforms had been shown to be effective or feasible. In 1961, the Vera Foundation (now the Vera Institute of Justice) initiated the Manhattan Bail Project. A research team, organized by a philanthropist and a magazine editor troubled by the failings of the bail system, collaborated with a criminal court in New York City to test reforms to bail procedures.' The researchers collected and carefully analyzed data on the effects and feasibility of implementing these reforms.' They then co-sponsored and presented the results of the project at a conference attended by numerous high-level officials.6 Based in part on the results of the project-the first of its kind in the United States-the federal government and many states passed bail reform laws aimed at increasing the fairness of pretrial release decisions by making them more sensitive to the suspect's actual risk of fleeing the jurisdiction or committing an offense prior to trial, rather than simply the seriousness of the alleged crime and the suspect's ability to secure the cash or property necessary to post bail. Forty years later and ten thousand miles away, Chinese legal scholars and criminal justice officials widely recognized coerced confessions as one of the most pressing problems facing the justice system. Police officers' use of torture 'See VERA INST. OF JUSTICE, PROGRAMS IN CRIMINAL JUSTICE REFORM TEN-YEAR REPORT 1961- 1971, at 21(1972). 2 Id. at 22. 'Id. at 22-23. 4 Id. at 12, 23. s Id. at 27-32. Id. at 33. ' Paul B. Wice, Release on Own Recognizance (ROR), in 4 ENCYCLOPEDIA OF CRIME AND PUNISHMENT 1372, 1373 (David Levinson ed., 2002). 8 See, e.g., NMJA i [Zhou Yijunl, -L i, 4 " K, ;Z fn $at 31 it WIit k " AZ [Speeches from a Conference on the She Xianglin Wrongful Conviction Case], o206 $2, filJ*ti)Mift [EAST CHINA CRIM. L. REV.] 267, 280-81 (reprinting a number of speeches identifying coerced confessions as a major problem in China and a primary cause of wrongful convictions); $-id [Li Zhongcheng] et al., I hilf 2003 I ,it [Summary of the 2oo3 Annual Meeting of the China Law Society's Procedural Law Research Association], rI M [CHINA LEGAL SCI.], no. 1, 2004 at 186, 187 (listing coerced confessions as one of the main targets of reforms). On the attitudes of the Columbia Journal ofAsian Law, Vol. 24, No. 2 (201) to obtain confessions had led to the deaths of numerous suspects,' as well as the imprisonment or execution of an untold number of innocent citizens.10 Scholars and officials put forth a variety of explanations for the prevalence of coerced confessions" but offered few concrete and feasible solutions. general public toward the use of torture in obtaining confessions, see #11M [Lin Lihong] et al., Ei- ) (ig RA#) [Survey on the Public Awareness of Coerced Criminal Confessions (Part I: General Public)], W [L. REV.], no. 4, zoo6 at 117, 133-34 (finding that nearly half of respondents believed the use of torture to obtain confessions is "common and frequently occurs" in China and that the majority of respondents recognized that coerced confessions are illegal but would tolerate the use of coercion in cases involving serious or egregious crimes). In a related study, Lin et al. found that two thirds of surveyed police officers in Henan and Hubei believed that state actors who torture suspects into confessing are prosecuted and convicted only if one of three conditions is met: (i) higher-level officials order the prosecution, (2) the coerced confession produces a strong public response or other major repercussions, or (3) reliable evidence clearly proves that the tortured suspect is innocent. #-JI [Lin Lihong] et al., jhu ik i li4R (Ta & ) [Survey on the Public Awareness of Coerced Criminal Confessions (Part II: Police)], 8 WTiA [L. REV.], no. 5, 2oo6 at 123, 134. 9 See )#9 [Zhuang Huining], #1JigiM -M F'KItJM iL [Why Torture Methods Persist], li J t ffIJ [NEWS OUTLoOK WKLY.1, July 2, 2001, at 28, 28 (noting that an official investigation in Heilongjiang, Hubei, Inner Mongolia, Shaanxi, Tianjin, and Zhejiang identified twenty-one suspects who had been tortured to death during interrogations between 1997 and 1999). "oSee IR* . RWA [He Jiahong & He Ran], J [Evidentiary Issues in Wrongly Decided Criminal Cases], ilkMAN [TRIB. POL. ScI. & L.], Mar. 2oo8, at 3, 10- 12 (analyzing the relationship between coerced confessions and wrongful convictions). For examples of wrongful convictions that resulted in the execution of an innocent citizen, see k&k [Zhao Ling], "4WATK" :i5 i "& A" A AIFF4fili [Nie Shubin Case Still Unresolved; To Prevent Cover-Up, Public Calls for Thorough Investigation], M )f R [SouTHcN.coM] (Mar. 24, 2005), http://www.southcn.com/ weekend/commend/20050324ooo6.htm (discussing the wrongful execution of Nie Shubin); 9 I-T- A A IT EfU, 9 4M A x XR [Man Executed for Rape and Murder; Real Killer Caught 9 Years Later], EH iF11T. [SANXIANG CITY EXPRESS], Aug. 6, 2009, at A9 (discussing wrongful execution of Huge Jiletu); J 17 *rHICKiJE 9 [Teng Xingshan Wrongfully Executed, Found Not Guilty 17 Years Later], R iW A$1J A [DEMOCRACY & LEGAL SYS. NET] (Oct. 13, 2007), http://www.mzyfz.com/index.php/cms/ item-view-id-363-page-1. "See, e.g., fiVId [Li Yunzhaol, filiR R.Il 4&IMffi' [Research on Causes of and Strategies Against the Persistence of Torture in Obtaining Confessions], 2002 k [J. PEKING U.] (SPECIAL ISSUE) 154, 154-56 (identifying a range of causes of coerced confessions, including the presumption that suspects are generally guilty, hatred of criminals, the absence of an effective exclusionary rule for confessions obtained through torture, the lack of the right to remain silent, poor professionalism, insufficient funding, poor investigative technology, and the infrequency with which police officers known to have used torture to obtain a confession are punished); T-401 [Wang Zhenchuan], M- ffi iRi9 fig i iRA (il WH M 336 The Use of Demonstration Projects to Advance Criminal Procedure Reform 337 Two legal scholars at the China University of Political Science and Law, Fan Chongyi and Gu Yongzhong, attempted to develop a practical solution to the problem of coerced confessions. They collaborated with local procuratorates in Beijing, Henan, and Gansu to test three potential reforms: videotaping interrogations, audiotaping interrogations, and allowing defense lawyers to observe interrogations. After collecting and carefully analyzing data on the effects and feasibility of implementing these reforms, they presented their findings at a conference attended by numerous high-level officials." The Supreme People's Procuratorate then consulted this project-one of the first of its kind in China-as it expanded pilot regulations mandating the recording of interrogations in cases investigated by procurators, 3 thereby taking a significant step toward reducing the number of coerced confessions in China.' [Curing the Chronic Ill of Coerced Confessions--On the Causes, Harms, and Remedies], A R&& [PEOPLE'S PROCURATORIAL SEMIMONTHLY], no. i, January 2oo6 at 15, 16-18 (attributing the prevalence of coerced confessions to factors such as a long tradition of using torture to obtain confessions in China, poor enforcement of suspects' and defendants' procedural rights, lack of effective sanctions faced by state actors who use torture to obtain confessions, budgetary and logistical constraints, outdated investigative technology and techniques, rigid performance evaluation systems, intense pressure from party and higher-level department bodies to solve "major" cases, poor professionalism of police officers, and the importance courts place on confessions in adjudicating criminal cases). " For transcripts from the presentation of findings and the ensuing discussion, see flit iH n TrEF ik$ *ZiET Iff -- M if gi t i ti A , :R R 2$1J1 ila'R [ EMPI RICAL RESEARCH ON INTERROGATION PROCEDURE REFORM--EXPERIMENTATION WITH LAWYERS- ON-SITE, AUDIOTAPING, AND VIDEOTAPING IN SUSPECT INTERROGATIONS] 201-59 (RIB P), *0k1 [Fan Chongyi & Gu Yongzhong] eds., 2007) [hereinafter EMPIRICAL RESEARCH ON INTERROGATION PROCEDURE REFORM]. 3 In 2005, the Supreme People's Procuratorate passed pilot regulations requiring procurators to record their interrogations of suspects. Sup. People's Procuratorate, AR RV iJ H IR V RA RJ AFR IM (iiTh) [People's Procuratorate Regulations on the Implementation of Real-Time Audiovisual Recording of Suspect Interrogations (Trial)] (promulgated.by the Sup. People's Proc., Dec. 15, 2oo5, effective Dec. 15, 2005), available at http://www.pkulaw.cn/fulltext-form.aspx?db=chl&gid= 147147. Id 2006, the Supreme People's Procuratorate elaborated on the implementation of these regulations by announcing a three-step plan for reform that mandated the recording of interrogations in all cases investigated by procurators nationwide beginning in October 2007. Att, MYNA1 [Yang Weihan & Lii Xueli], i A l J S.JR t3-B IF l f it ft iZE ( 1, Mf * R [XINHUA NET] (Mar. 11, 2006), http://news.xinhuanet.com/misc/2oo6-03/ui/content_4289692.htm; see also 3E Mvi A [Wang Xinyou], ilik: L l it # [Supreme People's Procuratorate: All Interrogations Video-Recorded in Full; Recordings Have Not Revealed the Use of Illegal Investigation Methods], SUP. PEOPLE'S PROCURATORATE (Nov. 14, 2007), http://www.spp.gov.cn/sitezoo6/2007-1-14/ooozn6o95.html (reviewing the implementation of the pilot regulations). 14 Email from Fan Chongyi, Professor, Department of Law, China University of Political Science and Law, to Thomas Stutsman, Robert L. Bernstein Fellow in International The Manhattan Bail Project and the "three systems" interrogation reform project described above played strikingly similar roles in promoting specific justice reforms. They may share another similarity as well. The Manhattan Bail Project, as the first in a long line of projects that used rigorous empirical methods to test justice reforms in practice, pushed forward the development of evidence-based policy in the U.S. justice system. The "three systems" project may herald a similar development in China. This article assesses the development of scholar-led demonstration projects on Chinese criminal procedure reform. Over the past decade, Chinese scholars have carried out a number of demonstration projects, an applied form of empirical research characterized by the organized implementation and systematic evaluation of a new practice in the justice system to assess the merits of adopting the reform on a larger scale. My review of these projects, as well as the major advantages and challenges of this approach to advancing reform, shows that demonstration projects are feasible and hold significant promise for the future of Chinese criminal procedure reform. I argue that although comparative and theoretical research play an important role in guiding criminal procedure reform, demonstration projects have a number of unique and complementary qualities that can be exploited by scholars to more effectively promote reform. Part I of this article outlines the role of academic research in Chinese criminal procedure reform and reviews demonstration projects that have been led by Chinese scholars. Part II draws on these demonstration projects to discuss several advantages of this approach to criminal procedure reform. Part III describes two significant challenges to carrying out demonstration projects. I conclude by offering a few thoughts on the future role of demonstration projects in Chinese criminal procedure reform. I. THE ROLE OF ACADEMIC RESEARCH IN CRIMINAL PROCEDURE REFORM AND THE DEVELOPMENT OF DEMONSTRATION PROJECTS I begin this Part by describing the role of academic research in the reform process and the main types of research conducted by Chinese legal scholars. I then outline the scope of demonstration projects that have been carried out on Chinese criminal procedure reform over the past decade. Finally, I briefly summarize two well-known examples of how legal scholars have employed this emerging research approach to simultaneously fill significant gaps in the scholarly literature and promote criminal procedure reform. Human Rights, Vera Institute of Justice (Mar. 31, 2011) (on file with author). Wang Zhenchuan, Executive Vice President of the Supreme People's Procuratorate, attended the conference and specifically referred to the value of the project in an article in People's Procuratorial Semimonthly. Wang, supra note ii, at 20. 338 Columbia Journal ofAsian Law, VOL 24, No. 2 (2on1) The Use of Demonstration Projects to Advance Criminal Procedure Reform 339 A. The Role of Academic Research in Criminal Procedure Reform Criminal procedure reform in China is frequently described as a dynamic top-down and bottom-up process comprised of central authorities and local criminal justice institutions. " Central authorities provide the guiding principles, goals, and foci of reform.'6 They lay out general areas of reform in central-level policies and reform plans," and mandate the implementation of other reforms by enacting or revising national legislation and supporting regulations.'8 Local criminal justice actors, in turn, must exercise the significant discretion accorded to them to implement national laws and regulations and translate general policies and goals into local practice. Information on the practical problems local criminal justice actors encounter in implementing mandated reforms and testing potential reforms in practice-as well as the concrete approaches they develop to overcome these challenges-is then filtered up to higher-level policymakers to consider in their future reforms. 9 " See, e.g., rE _2 K [Zuo Weimin], + -"- M n1 ii) + A [S-86 : RE -'- -R IA [Judicial Reform in China at a Crossroads: Reflections and Prospects], fIMI [MOD. L. Sa.], Nov. 2008, at 6o, 61-63 (suggesting that central authorities, local criminal justice actors, and legal scholars are the primary sources of justice reforms in China). t Id. at 63. 17 Instructions from central authorities may involve policy announcements relating to criminal justice in general or directives on more specific subareas. See, e.g., Sup. People's Ct., ARR V 4i4E1l (2009-2013) [Third Five-Year Reform Plan of the People's Court (2oo9-2o3)], Mar. 17, 2009, SUP. PEOPLE'S CT. GAZ. (P.R.C.) (identifying general areas and goals of reform); 144 [Wang Doudoul, io P 1 H & 1H a R I iiN Ai IT f 01i i Ir f L * & [Beginning Oct. i All Courts to Implement Comprehensive Standardized Sentencing Reform], 8$#i Hi [LEGAL DAILY], Sep. 17, 2010, at I, available at http://www.legaldaily.com.cn/index-article/content/20lo-o9/16/content 2291o99.htm?node=595 5 (an example of the Supreme People's Court instructing lower courts to test sentencing reforms). For a discussion of the institutional and ideological dynamics of implementing criminal policy changes, see Susan Trevaskes, The Shifting Sands of Punishment in China in the Era of "Harmonious Society", 32 L. & POL'Y 332, 344- 53 (2010) (discussing the development of the "balancing leniency and severity" policy in criminal justice). 18 See, e.g., A WE f$iff [Criminal Procedure Law of the People's Republic of China] (promulgated by the Nat'l People's Cong., effective Jan. i, 1980, amended Mar. 17, 1996), STANDING COMM. NAT'L PEOPLE'S CONG. GAz. (P.R.C.) [hereinafter CPL]; iR R AfAi ((A jg i) g a Uff [Supreme People's Court Interpretation on Certain Questions Concerning the Implementation of the People's Republic - of China Criminal Procedure Law] (promulgated by the Sup. People's Ct., Sept. 2, 1998, effective Sept. 8, 1998), SUP. PEOPLE'S CT. GAz. (P.R.C.), available at http://www.dffy.com/faguixiazai/ssf/200311/ 20031109201918.htm. '9 Zuo, supra note 15, at 62-63. Provincial-level authorities, of course, also play an important role in the reform process. See ?)ME [Fang Baoguo], U B -- 2 filt Ag JIJ ifliW tiE 10 E! 9! [Reality Has Already Occurred--On Local Criminal Columbia Journal ofAsian Law, Vol. 24, No. 2 (2011) In addition, legal scholars play a vital role in criminal procedure reform through their research. Leading criminal procedure scholar Zuo Weimin argues that legal scholars, along with central authorities and local criminal justice actors, constitute the three main forces in criminal procedure reform.2 0 Zuo points out that while legal scholars tend to be somewhat idealistic and frequently draw on foreign models to advocate immediate and fundamental reform," their recommendations contribute to new ways of thinking about criminal procedure and the values to be realized through reform-both of which are critical in guiding reform." Additionally, central-level criminal justice agencies frequently consider ideas developed by scholars alongside the practical needs of their constituent offices to pass important reforms,23 and local criminal justice institutions selectively draw on legal scholarship to develop and try out reforms.24 B. Approaches to Academic Research and Criminal Procedure Reform The type of research on which influential scholars base their recommendations for reform has evolved in recent years. Fan Chongyi explains that two methodological approaches have dominated the study of Chinese criminal procedure during the reform era: hermeneutic analysis and comparative analysis." Hermeneutic analysis-a form of textual analysis that can be used to interpret the provisions and underlying principles of the Criminal Procedure Law (CPL) and other relevant laws-was the primary approach to studying criminal procedure in the 198os and early 199os, while Evidence Rules in China], i$(MitM (TRIB. POL. Sci. & L.], May 2007, at 41, 41-45 (describing the evidentiary rules and reforms developed by various provincial authorities). 20 Zuo, supra note 15, at 61-62. 21 Id. at 64. 22 Id. at 63- 23 Id. ' Id. Legal scholars can also exercise a relatively direct influence on criminal procedure reform by interacting with state actors to design reforms or draft legislation. Professor Chen Guangzhong, for example, played a critical role in shaping the 1996 revisions to the CPL. Tgge [Bian Jianlin], Vj3 A rP: V gljSW SW A [Chen Guangzhong: Leader in New China's Criminal Procedure Law], [M ITJ~f V iA [CHINA TRIAL NEWS MONTHLY], May 2008, at 30, 32-33. According to Zuo Weimin, legal scholars have become even more actively and directly engaged in promoting criminal procedure reform since 2003, which is reflected by their newfound willingness to make recommendations for reform directly to central policymaking bodies. Zuo, supra note 14, at 62. 25 3-' [FAN CHONGYI], fl)$ TV [PHILOSOPHICAL THOUGHTS ON CRIMINAL PROCEDURE] 229-31(2010). 34o0 The Use of Demonstration Projects to Advance Criminal Procedure Reform 341 comparative analysis has dominated the study of criminal procedure since the mid-199os. 6 Zuo Weimin holds a similar view on the state of Chinese criminal procedure research and argues that criminal procedure scholarship is currently dominated by two related approaches: the comparative approach and the normative approach. Scholars who employ the comparative approach generally look at the justice system of a developed country-often the United States'-to identify differences between that country and China.'8 They then proceed to speculate whether the foreign practice is suitable for China, and if so, argue that China should adopt the foreign practice.29 Scholars who employ the normative approach to legal scholarship, on the other hand, use (frequently abstract) principles or values such as human rights or democracy to evaluate the Chinese justice system and suggest reforms they believe would better realize these principles or values." Regardless of methodological approach, the vast majority of Chinese scholarship on criminal procedure over the past thirty years has been theoretical. ' Chinese legal scholars rarely collect empirical data or use empirical evidence to support their claims.2 Moreover, many studies that 26 Id. 27 Zuo, supra note 15, at 67. 28)-a K lZuo Weimin 1, M A 0114-99Mft ifif ittiliAE S-M ff: 3ifIYE Miftit [Paradigmatic Transformation and China's Criminal Procedure System Reform-A Discussion on Empirical Research], i MM [CHINA LEGAL Sd.], no. 2, 2oo9 at u8, u8-19. 2 Id. 30 Id. at 19. Normative research is generally more abstract and theoretical than comparative research, which typically describes foreign laws and practices. I therefore refer to the two main currents of Chinese criminal procedure research as "comparative" and "theoretical" throughout this article. 3' FAN, supra note 25, at 238; i3 [Chen Guangzhong], Rkfft 30 4Mffi]iffi6& * [Accomplishments ofthe Science ofCriminal Procedure Law in 30 Years ofReform], fQ RMt [MOD. L. Sci.], Jan. 2009, at 141, 145 (noting that Chinese legal scholars have traditionally emphasized abstract theory at the expense of empirical investigation, which has resulted in a sharp divergence between academic understanding of the justice system and actual practice). 32 Empirical research constitutes only a tiny fraction of Chinese legal scholarship. A January 4, 2011 search of Amazon.cn for "empirical research" (WiiEWR) (shizheng yanjiu) books written by a Chinese author in the "law books" category generated one result for 2000, thirteen results for 2004, and twenty results for 2009. In the China Academic Journal Database (http://china.eastview.com), a mere 0.12% (61/50613) of law articles listed "empirical research" as a keyword in 2009. This number is, however, higher than in 2000 and 2004, when 0.04% (11/27847) and 0.04% (16/39676) of articles, respectively, listed "empirical research" as a keyword. When this search was further limited to "criminal law" and "procedural law and the judicial system" law articles, there was one result for 2000, four results for 2004, and twenty-three results for 2009. Columbia Journal ofAsian Law, Vol. 24, No. 2 (2011) purport to be empirical simply review a few relevant cases or base conclusions on anecdotal evidence rather than systematically collected data.33 In recent years, a handful of scholars have increasingly recognized the benefits of using empirical methods to study the justice system. The growth of publications on empirical legal studies over the past decade reflects not only a readiness to advocate for more empirical research, but also to dedicate the considerable time and resources necessary to carry it out. This gradual trend toward empirical research can be seen in both general legal scholarship and research aimed at promoting concrete criminal procedure reforms. Indeed, several prominent legal scholars have argued that empirical research holds immense potential for the study of Chinese criminal procedure. Although empirical research continues to play a limited role in Chinese scholars' advocacy for criminal procedure reform,36 an important development is taking place. A small cadre of elite scholars has come to appreciate the potential of collaborating with criminal justice institutions and using empirical methods to test potential reforms in practice. Scholars have designed and led a number of demonstration projects despite the fact that they require close collaboration with criminal justice institutions and are generally more complex and expensive than non-empirical comparative or theoretical research.3 1 Fan Chongyi and Gu Yongzhong were at the forefront of this development. Their first interrogation reform project, which tested the effectiveness and feasibility of defense lawyers attending custodial interrogations of criminal suspects in Zhuhai and Beijing between 2002 and 2004, was the first major 3 FAN, supra note 25, at 238. The small but increasing number of books and articles related to empirical research, described supra note 32, illustrates the gradual development of empirical legal studies in China. 3 E.g., FAN, supra note 25, at 31-32 (arguing that empirical research could reshape the study of criminal procedure and is essential to understanding and implementing criminal procedure norms); Zuo, supra note 28, at 121-25 (discussing several advantages to using empirical research in justice reform). 3 Chen Guangzhong identifies four major collections of suggested revisions to the CPL published by scholars between 2003 and 2oo8. Chen, supra note 31, at 146. The authors of these collections, however, provide little empirical evidence to support their recommended revisions. ' The Vera Institute of Justice's seven-step approach to promoting reform hints at the level of involvement required to use demonstration projects to advance criminal procedure reform: (i) identify opportunities for innovation and secure funding to design a reform, (2) conduct additional background research and carry out preliminary planning activities, (3) design a reform and develop a detailed plan to test the reform in practice, (4) plan how to evaluate the effectiveness and feasibility of the reform, (5) collaborate with the relevant criminal justice organizations to implement the reform, (6) analyze the evaluation data and report the findings, and if the reform is successful, (7) work to institutionalize the reform on a larger scale. See JIM PARSONS ET AL., iA! -5. itk: 3 j2ri 6Jfll MO liii Uf ff r [EXPERIMENTATION AND REFORM: EMPIRICAL METHODS FOR IMPROVING JUSTICE SYSTEMS] 11 (WLl1, [Guo Zhiyuan] trans., 20o6). 342 The Use of Demonstration Projects to Advance Criminal Procedure Reform 343 scholar-led demonstration project on criminal procedure reform. 8 From 2004 to 2oo6, Chen Guangzhong, the leading scholar and principal architect of the 1996 revisions to the CPL,3 9 supervised two demonstration projects: Zuo Weimin's project on live witnesses4o and Song Yinghui's project on bail and discretionary non-prosecution. 4 ' Fan, Gu, Zuo, and Song have continued to pioneer this approach to advancing criminal procedure reform, which has attracted attention not only from other scholars, but from the Supreme People's Court and Supreme People's Procuratorate as well.4 C. The Scope of Demonstration Projects on Criminal Procedure Reform Chinese legal scholars have now employed demonstration projects to test a wide range of criminal procedure reforms. These reforms have involved diverse actors in the criminal process-from the police officers who investigate alleged offenses to the court adjudication committee members who effectively decide "important" or "difficult" cases. They have also addressed numerous stages of the criminal process-from the initial custodial interrogation of suspects through sentencing. Scholars have utilized a wide array of methods to produce empirical data to evaluate these reforms, such as reviewing case files, observing judicial proceedings, distributing questionnaires, and conducting, interviews. Appendix i, which lists a selection of notable demonstration 38 FiiK [Gu Yongzhong], M JiRHff' i E (iOSA) [Lawyer-on-Site During Suspect Interrogation (Pilot Test)], in f1J * W Y BJE $iTII: M iRJ H i'iJE (if Sl) [AN EMPIRICAL STUDY OF CRIMINAL PRETRIAL PROCEDURE REFORM: LAWYER-ON-SITE DURING SUSPECT INTERROGATION (PILOT TEST)] 192 (RER [Fan Chongyi] ed., zoo6). 3 Bian, supra note 24, at 32 (describing Chen Guangzhong's role in the revision process and stating that sixty-five percent of his recommended revisions were adopted in the CPL). 40 A [Zuo Weimin] et al., iiEi i [Pilot Research Report on Witness Court Appearance in Criminal Cases], in 1igfJ6flj iziTIJ iiEWY [EMPIRICAL STUDIES ON THE OPERATIONAL MECHANISMS OF CRIMINAL LITIGATION IN CHINA] 301 (tr -P K [Zuo Weimin] et al. eds., 2007). 4 *X7 , fil [Song Yinghui & He Ting], 5Ar, Ai) d S [Research Report on Reform of Bail and Discretionary Non-Prosecution of Minors] (2006) (unpublished report, on file with author). 42 See '4Zdi [Jiang Anjie], M fff+)F: IffJ JS2tfA " WiiA'lif" ItPIP [Sentencing Procedure Reform: "Wuhu Pilot Case" Released], M$IJPI [LEGAL DAILY], Sep. 30, 201o, at 9, available at http://www.legaldaily.com.cn/zbzk/content/2010-o9/30/content_23o6853.htm (demonstration project led by Professor Chen Weidong with the support of the Supreme People's Court); fi Flg -MWiS$1JJl iMEN [PROCEDURAL REFORM RESEARCH SUMMARY: CONCLUDING REPORT ON PLEA BARGAINING SYSTEM RESEARCH] ( K-Vff [Zhang Zhihui] ed., 2010) (demonstration project led by the Director of the Supreme People's Procuratorate Institute of Procuratorial Theory). Columbia journal ofAsian Law, Vol. 24, No. 2 (2011) projects that have been led by Chinese scholars, illustrates the versatility and growth of this method of conducting research and advancing reform.4 3 D. An Overview of Two Representative Demonstration Projects The preceding Section illustrated Chinese scholars' growing interest in using demonstration projects to advance criminal procedure reform. In this Section, I summarize two comparatively well-conducted projects on important areas of reform: Fan Chongyi and Gu Yongzhong's "three systems" interrogation reform project and Song Yinghui's bail and discretionary non- prosecution reform project. 1. Interrogation Procedures Chinese police officers' use of coercion to obtain confessions has contributed to several high-profile wrongful convictions in recent years. She Xianglin and Zhao Zuohai, each of whom served over a decade in prison for murders that never occurred, are two prominent examples of Chinese citizens whose lives were destroyed by convictions founded in part on coerced confessions. 4 Although it is impossible to meaningfully estimate the prevalence of coerced confessions, legal scholars and officials widely view the use of coercion to obtain confessions as one of the most pressing issues facing the Chinese justice system. 45 Fan Chongyi has led a series of three demonstration projects aimed at reducing coerced confessions. Although each project uses a slightly different approach to testing potential reforms, all three projects are premised on the idea that making interrogations more open and transparent will increase the accountability of investigating officers, thereby reducing the likelihood that they will use coercion to obtain confessions. Fan led the second of these projects with Gu Yongzhong in 2005.6 The primary goal of this project was to test the feasibility of "three systems" of 4 This list is based on a thorough literature review and discussions with numerous researchers involved in Chinese criminal procedure reform. It does not include projects that are either in their early stages of development or for which detailed information is not publicly available or otherwise accessible to the author. "For media accounts of their cases, see z11.11JA [Yun Shancheng], JA ##:;7 ; MIS i$1i 0 WIM'IN [Looking at Defects in China's Criminal Procedure from the She Xianglin Case and Other Miscarriages of Justice], ' [). CHINESE PEOPLE'S PUB. SECURITY U.], no. 6, 2005 at io8; kLf'4: AL'WOT [Zhao Zuohai: I Have Been Wronged], NEWS.163.COM (May 19, zoo), http://focus.news.163.com/10/o519/ 14/6729NT320001SM 9 .html. 45 See Zhou, supra note 8, at 28o-8i. 46 mgp1 7, [Fan Chongyi & Gu Yongzhong], iiiiJ kg Aflii'j, 2 JR20tJf (if) 9It Afl W [Summary Report on the Experiment for Establishing a 344 The Use of Demonstration Projects to Advance Criminal Procedure Reform 345 reducing coerced confessions: videotaping, audiotaping, or permitting a lawyer to attend custodial interrogations.4 ' Fan and Gu collaborated with public security bureaus in three disparate locations-Beijing, Jiaozuo (Henan province), and Baiyin (Gansu province)-to capture some of China's immense diversity and thereby more effectively assess whether the reforms could be implemented on a nationwide scale.' 8 The researchers ensured the availability of the necessary audiovisual equipment and defense lawyers and then collaborated with the police to select cases to include in the study.49 The participating police officers provided suspects with four options for interrogation: (1) request a lawyer to attend the interrogations, (2) videotape the interrogations, (3) audiotape the interrogations, or (4) proceed according to the "traditional" interrogation method (i.e., unrecorded and without a lawyer present).o A total of 265 suspects were given a choice of interrogation method.9 These 265 suspects comprised four groups: three experimental groups (i.e., lawyer-on-site, audiotaping, and videotaping groups) and a comparison group of suspects who chose the "traditional" interrogation method." A second comparison group consisted of 117 suspects who were not offered a choice of interrogation method and were interrogated according to the "traditional" method." The researchers collected and analyzed data on the characteristics of the suspect, the offense he or she was suspected of committing, and his or her choice of interrogation method. 54 They also interviewed suspects after their interrogations to inquire about the reasons they chose a particular interrogation method and their satisfaction with the method they selected." In addition, they distributed questionnaires to participating police officers and conducted interviews to explore the officers' views on interrogation reform and their experiences participating in the project.56 System of Lawyer-on-site, Audiotaping, and Videotaping During Suspect Interrogations], in EMPIRICAL RESEARCH ON INTERROGATION PROCEDURE REFORM, supra note 12, at 3, 8. 47 Id. at 3-5. 4' Id. at 6-7- 4 Id. at 8. 'o Id. at 8-9. " Id. at 9. 52 Id. 53 Id. 5 Id. at 9-19. Characteristics examined included the suspects' age, gender, education level, profession, and residency status. 5 Id. at 19-25. 56 Id. at 29-32. 346 Columbia Journal ofAsian Law, Vol. 24, No. 2 (2011) The researchers found that not a single suspect in the study reported being coerced into confessing," and that audiotaping, videotaping, or allowing a lawyer to attend interrogations did not meaningfully reduce the likelihood that suspects would confess.'8 The research team also found that a suspect's choice of interrogation method was associated with both the suspect's personal characteristics and the crime he or she was suspected of committing.59 Finally, the researchers found that the police officers who participated in the project generally supported interrogation reform and held positive views toward the "three methods" approach to reforming interrogation procedures 6 0-both of which suggest the feasibility of large-scale reform. 2. Bail and Discretionary Non-Prosecution The CPL grants police and procurators discretion to treat persons arrested for minor offenses with leniency at several junctures of the criminal process. One way in which police can exercise this discretion is to release a suspect on bail if he or she secures a "guarantor" who will ensure that the suspect does not flee the jurisdiction, tamper with evidence, interfere with witnesses, or miss court hearings.6 1 Procurators may similarly exercise their discretion by declining to prosecute cases in which the circumstances are minor and punishment is unnecessary. " Nevertheless, provisions on bail and discretionary non-prosecution have not been effectively translated into practice. Police rarely release suspects before trial in cases where there is sufficient evidence to secure a conviction,63 and procurators rarely exercise 5 Id. at 33. 58 XJl AW [Project Leaders' Report and Review of Experimental Videotaping Materials], in EMPIRICAL RESEARCH ON INTERROGATION PROCEDURE REFORM, supra note 12, at 209, 213 (transcript of Gu Yongzhong's oral presentation on the project). 59 Fan & Gu, supra note 46, at 9-19. Id. at 29-30. CPL, supra note 18, arts. 53, 56. For provisions on bail, see generally id. arts. 50-58. 62 Id. art. 142. 6, See M Ar [Tang Liang], P $ -f9' iE)} f [Empirical Analysis of Pretrial Detention in China], fill$I1 [CRIM. L.], no. 7, zoo at 29, 31 (studying two urban districts and finding that 22.6% of suspects in cases transferred to the procuratorate for prosecution by the police were granted bail); see also fi _P.R [Zuo Weimin], 0M IMATP: iliTi f1)A&}f [Release on Bail During Investigation: A Functional Analysis Based on Empirical Evidence], +3+&* [PEKING U. L.J.], no. 3, 2007 at 339, 341- 43 (investigating the use of bail in three public security bureaus and finding that police generally release suspects on bail because of insufficient evidence to continue detaining the suspect rather than to protect the suspect's interest in being free before trial). The Use of Demonstration Projects to Advance Criminal Procedure Reform 347 their power of discretionary non-prosecution.64 In 2005, Song Yinghui collaborated with police and procurators in Yongkang County, Zhejiang Province to carry out a demonstration project on expanding the use of bail and discretionary non-prosecution for juveniles.6 5 The goals of Professor Song's project were threefold: (i) evaluate the effects of expanding the use of bail and discretionary non-prosecution in cases involving juvenile offenders, 6 (2) examine the various factors that influence the effectiveness and feasibility of bail and discretionary non-prosecution, 6 and (3) use the results of the demonstration project to recommend revisions to the CPL.6 8 The researchers began by thoroughly reviewing the case files of past cases involving juvenile suspects in Yongkang County to develop screening tools to evaluate juvenile offenders' suitability for pretrial release and discretionary non-prosecution. 69 They then collaborated with police officers and procurators to test the screening tools in cases involving juvenile suspects. There were 103 participants in the bail experiment. The research team or the police applied the bail screening tool to seventy of these suspects; the remaining thirty-three suspects were handled by the police using their traditional criteria to form a comparison group."o There were fifty-five participants in the discretionary non-prosecution experiment. The research team applied the discretionary non-prosecution screening tool to twenty-five of these suspects; the other thirty suspects formed the comparison group and were handled by procurators according to their usual style.7 ' Use of the bail screening tool nearly doubled the number of juveniles released on bail: forty-seven percent of juveniles whose bail decisions were made according to the bail screening tool were released on bail, while only twenty-four percent of juveniles in the comparison group were released on 64 See 1 [Li Jianling], M. ifi f [Study of the Discretionary Non- prosecution System], 1(MR [J. NAT'L PROSECUTORS C.], Aug. 2009, at 123, 124 (analyzing criminal cases handled by procuratorates in two cities in Shandong province and finding that only twenty-three of the io,n8 criminal cases handled in 2007 were disposed of through discretionary non-prosecution). 65 Song & He, supra note 41. Id. at 3. 67 Id. 68 Id. 69 Id. at 4-5. The screening tools developed by the researchers included factors such as whether the suspect (i) had an established residence or close relatives in the area, (2) was a first offender, (3) played a minor role in the offense, (4) voluntarily surrendered to the authorities, (5) took the initiative to return the fruits of the crime or compensate the victim, and (6) would be able to continue his or her education or secure employment after being released. Id. at 5. 70 Id. at 4. 7' Id. at 5. Columbia Journal ofAsian Law, Vol. 24, No. 2 (2011) bail." Despite the larger number of juvenile offenders released before trial, juveniles selected for pretrial release using the bail screening tool were less likely to reoffend than juveniles selected for pretrial release by the police using their own criteria. While twenty-five percent of juveniles in the comparison group who were granted bail committed another offense while released on bail, only six percent of juveniles who were granted bail according to the bail screening tool committed another offense while released.' Use of the discretionary non-prosecution screening tool similarly increased the number of cases disposed of through discretionary non- prosecution and did not appear to negatively impact public safety. Fifty-six percent of suspects whose cases were handled through the experimental procedures were granted discretionary non-prosecution, while no suspects in the comparison group were given discretionary non-prosecution.7 ' None of the suspects who were granted discretionary non-prosecution in accordance with the screening tool were found to have reoffended at the time of the researchers' follow-up investigation. 5 II. THE ADVANTAGES OF USING DEMONSTRATION PROJECTS TO PROMOTE CRIMINAL PROCEDURE REFORM The influence of legal scholars' comparative and theoretical research on the 1996 revisions to the CPL indicates the enormous impact that these types of research can have on criminal procedure legislation.76 According to Zuo Weimin, these two approaches continue to dominate Chinese legal scholars' research aimed at promoting criminal procedure reform.' Nevertheless, the uneven implementation of the CPL over the past fifteen years suggests that enacting sweeping reforms without testing them beforehand may not be the most effective way to change criminal justice practice. Demonstration projects are a crucial yef underutilized complement to comparative and theoretical research on criminal procedure reform.' 8 In Part III, I draw on past demonstration projects to highlight several of their benefits in promoting criminal procedure reform. I begin by discussing the main advantage of using demonstration projects rather than comparative or theoretical research for advocating reforms: their ability to test a reform on a 7 Id. at 6. 7 Id. at 7. ' Id. at 1o. 7 Id. at io-u. 76 See supra notes 20-24 and accompanying text for a description of legal scholars' role in justice reform. 7 Zuo, supra note 28, at n8-19. 78 See supra note 32 for the percentage of law journal articles that list "empirical research" as a keyword. 348 The Use of Demonstration Projects to Advance Criminal Procedure Reform 349 small scale before it is implemented at the provincial or national level. I then review four other advantages of demonstration projects. Demonstration projects can: (1) produce objective information about criminal justice practice, (2) focus the debate over reform, (3) reduce opposition to reforms by showing that stakeholders' interests will not be substantially harmed, and (4) provide an avenue for non-state participants in the criminal process to influence reform. 9 A. Test a New Approach As explained in Part I, Chinese criminal procedure reform is both a top- down and a bottom-up process. Countless criminal justice institutions across the country are experimenting with a wide range of reforms. Demonstration projects allow legal scholars to take part in this dynamic source of reform by conducting research that answers three critical questions about a potential reform: whether it is effective, whether it is feasible, and what supporting measures are necessary to successfully implement it on a larger scale. 1. Effectiveness A reform flawless in theory may fail in practice. Criminal justice institutions and the participants in the criminal process are complex-the effect a reform will have in practice is difficult to predict without testing it beforehand. Reforms that have not been tested in practice may even exacerbate the problems that they were designed to address. One well-known example is the pretrial reforms effected by the 1996 revisions to the CPL. An attempt to minimize the phenomenon of judges "deciding the case before trial" (AGlJ}Ei$) (xianpan houshen), these revisions limit the amount of evidence that procurators can submit to the court before trial. Ironically, whereas before the 1996 revisions the defense could generally access the wide range of materials submitted to the trial court by procurators, defense lawyers now struggle to obtain sufficient information to prepare a defense. 8 " For a related discussion of the benefits of using demonstration projects to promote criminal procedure reform in China, see MEN [Guo ZHIYUAN], iAA- &A: +i ME [EXPERIMENTATION AND REFORM: THE CHINESE EXPERIENCE] (forthcoming Aug. 2011). Id. (noting that the revisions virtually eliminated "the defense's only opportunity to see the prosecution's evidence before trial"); HUMAN RIGHTS IN CHINA, EMPTY PROMISES: HUMAN RIGHTS PROTECTIONS AND CHINA'S CRIMINAL PROCEDURE LAW IN PRACTICE 33-35 (2001) (noting that ambiguous wording in the CPL and the requirement that defense lawyers have access to "major evidence," as opposed to all evidence, have given authorities broad discretion to withhold evidence). Ira Belkin, currently the Program Officer for Law and Rights in the Ford Foundation's Beijing office, describes these pretrial reforms as "a classic case of the rule of unintended consequences." Ira Belkin, China's Criminal Justice System: A Work in Progress, 6 WASH. J. MOD. CHINA 61, 8o It is, of course, better that a reform fails as part of a small-scale demonstration project rather than after it is implemented at the provincial or national level. Minimizing the harm of failed reforms is particularly important for criminal procedure because failed reforms may pose a risk to public safety. If the bail screening tool used in Song Yinghui's bail and discretionary non- prosecution project had been ineffective, it may have resulted in the release of offenders who would flee the jurisdiction and reoffend. If videotaping, audiotaping, or allowing a defense lawyer to observe interrogations appreciably reduced the ability of police to obtain truthful confessions, thereby leading to fewer guilty suspects being convicted, these offenders may have committed additional crimes when they would have otherwise been safely incarcerated. Although a non-negligible risk to public safety accompanies many demonstration projects, this risk is far greater when reforms that have not been tested in practice are implemented on a broad scale. 2. Feasibility Poor implementation of the CPL has plagued the Chinese justice system.81 Despite the lofty intentions of progressive academics and policymakers who participated in the 1996 revisions to the CPL, many of the rights granted to suspects and defendants are not consistently enforced in practice. For example, the robust rights to counsel and to prepare a defense described in the CPL are highly circumscribed in practice.8 The disconnect between formal law and actual practice may be especially pronounced due to the lack of institutional guarantees necessary to compel criminal justice actors to comply with criminal procedure legislation.8 4 (2000). For a background discussion on the marginalization of defense lawyers, the difficulties they face in securing the cooperation of state actors, and the importance of guanxi, see generally Ethan Michelson, Lawyers, Political Embeddedness, and Institutional Continuity in China's Transition from Socialism, 113 AM.J. Soc. 352 (2007). 81 For further discussion on the deficiencies of the revised CPL and its implementation, see generally HUMAN RIGHTS IN CHINA, supra note 8o (examining the problems of implementation with regard to judicial infrastructure, role of lawyers, pretrial detention, and evidentiary procedures) and Zuo, supra note 15, at 67 (stating that numerous CPL provisions based on adversarial reforms are "dead letters"). 82 See HUMAN RIGHTS IN CHINA, supra note 8o, at 2 (concluding that the implementation of the CPL has departed substantially from both the letter and the spirit of the law, with state actors refusing to follow the law and aggressively exploiting its loopholes). 83 Id. at 24-43 (discussing the limited role that defendants and defense lawyers are able to play in the criminal process and defense lawyers' difficulties collecting evidence and gaining access to their clients despite CPL provisions authorizing them to access certain case materials collected by the prosecution and to meet with their clients). N See, e.g., id. (discussing ubiquitous violations of defendants' right to counsel); ttiT [Li Fenfeil, Altn 1 1IJJ$@)ihM [Curbing Criminal Procedural Violations 350 Columbia Journal ofAsian Law, Vol 24, No. 2 (2on1) The Use of Demonstration Projects to Advance Criminal Procedure Reform 351 Demonstration projects can reduce the gap between formal law and its enforcement by providing empirical data on the feasibility of potential reforms. I address two aspects of feasibility below: criminal justice actors' acceptance of the reform and resources required for its implementation. The uneven implementation of the CPL suggests that reforms that are not sensitive to the social, political, and institutional context in which they are to be implemented run the risk of being nullified by the line-level criminal justice actors responsible for their implementation.8 Through Procedural Sanctions], 9** [JURISTS REV.], no. 1, 2009 at 95 (arguing that the prevalence of procedural violations is due to the absence of effective procedural sanctions under the current CPL regime); I*it [Chen Ruihual, Mflj$WFY AlHIFJ 10Wff3 [A Preliminary Study on the Failings of Criminal Procedure], FPW * [CHINA LEGAL SCd.], no. 6, 2007, at 141 (attributing the prevalence of procedural violations in part to the absence of concrete procedures for dealing with state actors' violations of defendants' rights); IlM [Tang Lei], i- l i [On the Classification and Exclusion of Illegally-Obtained Evidence], % ff [Soc. Sci. RES.], no. 6, 2001 at 91 (suggesting that formal laws prohibiting coerced confessions are ineffective because defendants have no effective remedies for violations of these provisions). For a recent example of the disparity between formal law and actual practice, see Paul Mooney, Chongqing Execution Raises Political Spectre, S. CHINA MORNING POST, Oct. 3, 201o, at 6 (discussing the Fan Qihang case, in which a confession was admitted into evidence despite the defendant's claim-supported by powerful evidence-that it was obtained through torture and must therefore be excluded under recently passed regulations that require the exclusion of confessions obtained through torture). 8 But line-level criminal justice actors' reluctance to accept new practices is not a phenomenon unique to China. Drawing on' her more than ten years of experience designing and supporting demonstration projects as Director of Planning and Government Innovation at the Vera Institute of Justice, Megan Golden suggests as follows: People naturally fear and resist change. This is not a sign of bad intention or hostility or excessive self-interest, it is a natural human tendency. Significant justice reforms often require people to change what they do every day in their jobs and how they interact with others. .. . [A] lot of resistance to justice system reforms comes from this natural reaction. Golden further explains this phenomenon in the context of a demonstration project on restorative justice being carried out in Brooklyn, New York: For example, in 2oo9, Vera launched Common Justice, a restorative justice program for victims and young adult defendants involved in violent crimes. The program brings those responsible for violence together with the people they harmed in a carefully structured way to come up with a resolution that addresses the underlying circumstances, holds them accountable for their actions, and addresses victims' needs. The resolution is an alternative to incarceration. We work in close partnership with the District Attorney's office, which must approve participation. When we introduced the idea to the prosecutors, they were skeptical. However, when they saw the research it was based on, the operations plan, and our risk management plans, they agreed to Columbia journal ofAsian Law, Vol. 24, No. 2 (2011) Understanding the extent of this resistance to a potential reform is vital to ensuring that the proposed reform is feasible. Demonstration projects generate concrete and objective information on line-level criminal justice actors' acceptance of a reform and, in effect, the probability that it will be implemented if entered into law. Some demonstration projects suggest that a potential reform is likely to be implemented in practice; others indicate that the potential reform would likely fail in practice.8 In either scenario, demonstration projects provide important feedback that can be used to design and enact reforms that will actually be implemented in practice and meaningfully advance their intended goals. Insufficient resources have also contributed to the poor implementation of reforms laid out in the CPL and related laws and regulations. 8 A reform based on foreign practice or abstract theory may require more resources than available to the criminal justice actors responsible for its implementation. Demonstration projects can address this issue by providing detailed data on participate in a carefully designed pilot. Although the program is still in its early stages, many of the prosecutors who have been involved have found it a more productive and satisfying way to resolve some cases of violence because it truly addresses victims' needs and the circumstances that led to the violence. Email from Megan Golden, Director of Planning and Government Innovation, Vera Institute of Justice, to Thomas Stutsman, Robert L. Bernstein Fellow in International Human Rights, Vera Institute of Justice (Jan. 23, 2011) (on file with author). 8 See, e.g., PROCEDURAL REFORM RESEARCH SUMMARY: CONCLUDING REPORT ON PLEA BARGAINING SYSTEM RESEARCH, supra note 42, at 3 (finding that the project's experimental plea bargaining procedures were successfully implemented by project participants in general). 8See, e.g., Mig [Chu Fumin], ih firJl Zfl- 111MO{ & n$jp @igg* ippf fl j } lt~f1i'i~" ]4 IP [Between Experimentation and Implementation-"Bail System Reform and the Expansion of the Role of Defense Lawyers" Project Report], in Jfij$2FiPit [CRIMINAL LAW REVIEW] 86, 93 (MM, ft [Chen Xingliang] ed., 2009) (relating that line- level prosecutors were generally unwilling to implement the bail procedures designed by the research team or the leadership of the procuratorate). 8 See, e.g., Wang, supra note n, at 17 (identifying underfunding of the police as a contributing factor in the prevalence of coerced confessions in China); Lin et al., The Public Awareness of Coerced Criminal Confessions (Part II: Police), supra note 8, at 130 (claiming that many practitioners in the Chinese justice system believe that underfunding of the police contributes to the use of torture to obtain confessions). For a discussion on sources of funding for public security institutions and the impact of budgetary restraints on policing, see Murray Scot Tanner & Eric Green, Principals and Secret Agents: Central versus Local Control Over Policy and Obstacles to "Rule of Law" in China, 191 CHINA Q. 644, 659-60, 667-68 (2007). 352 The Use of Demonstration Projects to Advance Criminal Procedure Reform 353 the amount of resources required to implement a particular reform and by testing alternative, more cost-effective reforms that advance the same goals.8 Fan Chongyi's most recent interrogation reform project, which tests the feasibility of audiotaping or videotaping interrogations in five locations across China, exemplifies the use of demonstration projects to evaluate the cost of reforms.9o In his live witness project, Zuo Weimin analyzed the length of oral testimony-a proxy for judicial resources-of witnesses who testified in court." Fan Chongyi and Gu Yongzhong's "three systems" project illustrates how demonstration projects can reveal unexpected or subsidiary cost savings: defendants whose confessions were recorded were less likely to recant their confessions in court or accuse police of misconduct, which in turn reduced the costs necessary to investigate and respond to accusations of coerced confessions.2 3. Additional Supporting Measures Related to demonstration projects' ability to test both the effectiveness and feasibility of a reform is their ability to identify necessary supporting measures. A major reform often requires a number of supporting measures to be effective. Some supporting measures, such as establishing a system of presentence reports to support the implementation of sentencing guidelines, are costly and time-consuming. Other supporting measures, such as allowing a suspect to call a lawyer soon after being detained, may appear minor but are critical to the success of a much more significant reform. Demonstration projects reveal the real-world challenges to successfully implementing a reform by testing the reform in practice. The results inform the development of necessary supporting measures to maximize the effectiveness and feasibility of the primary reform, thereby helping to prevent widespread implementation problems if the reform is enacted on a larger scale.93 SSee, e.g., :EU, KV1 [Wang Jing & Zhang Kai], 8jffiai ggA'iTfiJ ., i S$IJA I (i4 9) Xi l 9. 1 [Establishing a System of Lawyer-On-Site, Audiorecording, and Videorecording During Suspect Interrogation (Experiment): Project Report from Haidian], in EMPIRICAL RESEARCH ON INTERROGATION PROCEDURE REFORM, supra note 12, at 71, 104 (comparing the average duration of interrogations when a lawyer was present, when the interrogation was audiotaped, and when the interrogation was videotaped). 90 4ti1 [Du Meng], Jtf'Jii -i Ai h idif [Procedural Reforms Banning Coerced Confessions May Be Written into the Criminal Procedure Law], $l[M I 1 [CHINA COURTS] (July 13, 2010), http://www.chinacourt.org/html/article/2o1oo7/ 13/417863.shtml (interview with Fan Chongyi). 9 Zuo et al., supra note 40, at 338-39. i S .- {R { 2 & 1% (r i it [Research Team Presentations and Question-and- Answer Session], in EMPIRICAL RESEARCH ON INTERROGATION PROCEDURE REFORM, supra note 12, at 216-17. 9 Song Yinghui, for example, details reforms in seven different areas to support the implementation of victim-offender mediation: (i) state compensation for crime victims, B. Objective Information About Criminal Justice Practice A thorough understanding of criminal justice practice is critical to effectively reforming criminal procedure for two reasons. First, an accurate understanding of the prevalence or severity of an alleged problem can be used to focus reform efforts on issues most in need of reform." Second, a nuanced understanding of criminal justice practice can ensure that reforms address the underlying causes of the problems that they are intended to resolve. Basing reforms on empirically unsupported assumptions can easily lead to ineffective reforms. Before Zuo Weimin's demonstration project on live witnesses in criminal trials, for example, many scholars believed that the low rate at which witnesses testified in court was primarily due to their inability or unwillingness to appear in court. 9 Zuo's thorough empirical research, however, found that the primary reason witnesses rarely appear in court is not because they are unable or unwilling to do so,96 but because procurators and judges rarely request that they appear in court and judges seldom grant defense lawyers' requests to call a witness.97 A reform intended to increase the rate of in-court testimony would therefore be more likely to succeed if it was aimed at procurators and judges rather than directly at witnesses. (2) pretrial detention, (3) community corrections, (4) performance evaluation systems, (5) supervision of victim-offender mediation, (6) criminal justice funding, and (7) mediator training. *%I4 [Song Yinghui], )IJSE$fUW iTW [Summary Report of Empirical Research on Criminal Mediation], in MJ W M V ifE Tf [EMPIRICAL RESEARCH ON CRIMINAL MEDIATION] (TA7 [Song Yinghui] et al. eds., 2010) 3, 32-36. 9 Using empirical methods to investigate local criminal justice practice can also illuminate alternative approaches that local-level criminal justice actors have developed to address a particular problem. See, e.g., 5WW [Song Yinghui] et al., S -M Y li J1W# "&MA'i)JE" 44iA. R [Survey Report on Criminal Mediation by "Commissioned People's Mediators" in District Y of City S], in. EMPIRICAL RESEARCH ON CRIMINAL MEDIATION, supra note 93, at 77-84 (evaluating one city district's use of people's mediators to conduct victim-offender mediation). 9 See, e.g., iLEFA [Xu Zhi], 7 4J tiEA f.$1JiaE0, A8 @@ [Legislative Ideas for a Working System of Witness Court Appearance], 3 [M AR K # * * [J. CHINESE PEOPLE'S PUB. SECURITY U.], no. 5, 2004 at 88, 88-90 (arguing that witnesses are generally unwilling to testify in court and attributing their reluctance to legal shortcomings and cultural attitudes). 96 In Zuo's study, eighty-three percent (44/53) of witnesses called to testify appeared in court. Zuo et al, supra note 40, at 309. Fan Chongyi has also found that witnesses often appear in court when asked to do so. 2A3L [Fan Chongyi], fJ-S$1M &y$RJAiff i [Research Report on Criminal Trial Procedure and Pretrial Procedure Reforms], in fJ $ A J Y & * i * 'm [RESEARCH REPORT ON CRIMINAL TRIAL PROCEDURE REFORMS] ( 3 [Fan Chongyi] ed., 2oo8), 1, 3-4 (finding that witnesses appeared in thirty-one percent (5/i6) of cases in which the court implemented concrete procedures for notifying witnesses to testify at trial). " Zuo et al., supra note 40, at 304. 354 Columbia Journal ofAsian Law, Vol. 24, No. 2 (2on1) The Use of Demonstration Projects to Advance Criminal Procedure Reform 355 Chinese scholars have used a range of empirical methods to investigate criminal justice practice over the past decade. One approach is to review case files. Two illustrative examples are Zuo Weimin's analysis of cases to examine the rate at which witnesses testify in court, and Song Yinghui's review of cases handled during the year preceding his victim-offender mediation project." Another approach is to interview or distribute questionnaires to participants in the criminal process. Song Yinghui's method of collecting a sample of local victim-offender mediation practices by traveling to a number of provinces to distribute questionnaires and distributing questionnaires to criminal justice officials attending conferences or trainings at the National Judges College, the National Procurators College, or the Chinese People's Public Security University illustrates an innovative approach to collecting questionnaire data on criminal justice practice.o Contrasting cases handled through experimental procedures with a comparison group of cases handled according to regular procedures during the same time period can also generate useful data on criminal justice practice. Yet another Song Yinghui demonstration project-his bail and discretionary non-prosecution project- shows how evaluating a contemporaneous comparison group alongside an experimental group can provide useful data on current criminal justice practice.' C. The Debate over Reform Ira Belkin suggests that one major obstacle to criminal procedure reform in China is that arguments for and against particular reforms are repeated over and over but fail to persuade policymakers who do not already hold similar views.o2 Demonstration projects provide concrete data that support or weaken a policy position, which can in turn advance the debate over reform.o 3 98 Id. at 302-04. 99 Song, supra note 94, at 8 n.2. o Id. at 5. Song & He, supra note 41, at 6-12 (comparing bail and discretionary non-prosecution decisions between the experimental group and the comparison group). 1o2 Telephone Interview with Ira Belkin, Program Director for Law and Rights, Ford Foundation (Jan. 27, 2011). '03 In the context of U.S. criminal justice policy, prominent criminologist Joan Petersilia argues that "If there is no research, there are no objective data to guide decision makers. With no objective data, politics will continue to dominate decisions, and money will be invested in unproven programs and public safety will be jeopardized." Joan Petersilia, Influencing Public Policy: An Embedded Criminologist Reflects on California Prison Reform, 4 J. EXPERIMENTAL CRIMINOLOGY 335, 345 (zoo6). Jim Parsons, Director of the Substance Use and Mental Health Program at the Vera Institute of Justice, similarly suggests that empirical evidence produced by demonstration projects may be "the only fixed point in ideological and fiscal debates that typically center on rhetoric and political expediency." Email from Jim Parsons, Director, Substance Use Columbia Journal ofAsian Law, Vol. 24, No. 2 (201) As mentioned earlier, concerns for public safety are often raised in debates on criminal procedure reform-especially when the reform would expand the rights of the accused. Demonstration projects shed light on whether this anxiety is well-founded, reducing the likelihood that unfounded beliefs about increased crime will halt efforts to expand suspects' and defendants' rights."o4 In Fan Chongyi and Gu Yongzhong's "three systems" project, the impact on police officers' ability to fight crime was one of the main arguments by line- level police officers who opposed interrogation reform.' The empirical data produced by the demonstration project refuted this basis for opposing interrogation reform: videotaping, audiotaping, or allowing a defense lawyer to be present during interrogations did not meaningfully influence suspects' willingness to confess.os Another common argument raised by opponents of reforms that expand the rights of the accused is that the proposed measures are prohibitively expensive and time-consuming. Demonstration projects can show that implementing these reforms will not require a substantial amount of additional resources. 0 7 Similarly, demonstration projects clarify the underlying reasons for both support for and opposition to proposed reforms. Discussing the relationship between empirical research and criminal law, leading U.S. criminologist Michael Tonry states as follows: Even concerning the relatively impervious issues like capital punishment... , evidence sharpens the debates and clarifies what the issues really are. It is important to show and know, for example, that no credible empirical evidence suggests that capital punishment is an effective deterrent of homicide. If the evidence is absent or unclear, then people who support the death penalty for ideological reasons, but feel uncomfortable saying so, can hide behind claims that they support capital punishment as a means to saving innocent victims' lives. If the evidence clearly shows that death does not deter, then the and Mental Health Program, Vera Institute of Justice, to Thomas Stutsman, Robert L. Bernstein Fellow in International Human Rights, Vera Institute of Justice (Jan. 25, 2011) (on file with author). 104 According to Megan Golden, demonstration projects are especially effective in allaying fears over increased risk to public safety that might otherwise derail positive reforms. Email from Megan Golden, supra note 85. '05 Fan & Gu, supra note 46, at 30. to6 Project Leaders' Report and Review of Experimental Videotaping Materials, supra note 58, at 213. ' See supra notes 90-92 and accompanying text (discussing how demonstration projects can estimate the amount of resources necessary to implement a reform). 356 The Use of Demonstration Projects to Advance Criminal Procedure Reform 357 debate must be made in the moral terms that really motivate 108opponents and supporters. This benefit of empirical evidence may be particularly important in China. Zuo Weimin argues that criminal justice institutions strongly consider their institutional interests in evaluating potential reforms and sometimes block reforms by invoking a subterfuge (e.g., cost or risk to public safety) rather than disclosing the real reason for their opposition to the reform (e.g., that it would harm their institutional interests). ' Demonstration projects produce empirical evidence that can undercut opposition to a reform that criminal justice actors motivated primarily by institutional self-interest claim will adversely affect public safety or the administration of justice. By reducing the effectiveness of pretexts to justify support of or opposition to reforms, empirical evidence focuses the debate over criminal procedure reform and increases the accountability of institutional actors involved in the reform process. D. Stakeholders' interests and Opposition to Reform As noted above, criminal justice institutions jealously guard their interests in the policymaking process and vigorously oppose reforms they believe would harm their interests."o The preceding Section described how demonstration projects can focus the debate over criminal procedure reform. The ability of demonstration projects to replace assumptions about the implications of a reform with informed opinions based on actual experience facilitates reform in another way as well: it can show that the perceived threat of a reform to a criminal justice institution's interests is exaggerated,"' thereby weakening political opposition to a reform based on institutional concerns. In fact, the empirical data may even create an additional source of support."2 'os Michael Tonry, "Public Criminology" and Evidence-based Policy, 9 CRIMINOLOGY & PUB. POL'Y 784, 794 (2010). '0 Zuo, supra note 15, at 67. Several illustrative factors that influence the decisions of criminal justice institutions are available resources, caseloads, and performance evaluation systems. See infra notes 122-126 and accompanying text. no Id. ". Identifying how a reform may impact various criminal justice institutions' interests is, in fact, an important aspect of planning and evaluating demonstration projects. Email from Megan Golden, supra note 85 ("[A] properly planned demonstration project will have gotten input from all stakeholders and addressed their concerns to the extent possible."). "' Id. (stating that U.S. criminal justice actors who initially oppose a reform frequently find that the reform tested through a demonstration project makes their job easier or more satisfying). Columbia Journal ofAsian Law, Vol. 24, No. 2 (2011) Fan Chongyi and Gu Yongzhong's "three systems" project illustrates the ability of demonstration projects to reduce opposition to reforms by using empirical evidence to change perceptions of institutional interests. Many police officers who participated in the project initially opposed lawyers observing interrogations because they believed the presence of a lawyer would discourage suspects from confessing. But as these officers participated in the project they gradually found the presence of a lawyer not only does not impede confessions, but actually facilitates them."' This type of empirical data can be expected to reduce investigatory agencies' political opposition to similar reforms. E. Non-State Criminal Justice Actors' Influence over Criminal Procedure Reform A number of Chinese scholars have argued that non-state participants in the criminal process (e.g., defendants, victims, and witnesses) and the general public have little influence over the course of criminal procedure reform." Some of these scholars suggest that the lack of effective avenues for public participation has led to numerous reforms that facilitate the exercise of power by criminal justice institutions while neglecting the needs and interests of non-state participants in the criminal process." Empirical data on the views and attitudes of non-state participants in the criminal process gives them an indirect yet tangible voice in criminal procedure reform."6 Even simple observations of these participants' behavior in the criminal process can facilitate reforms that respond to their needs and interests. In Fan Chongyi and Gu Yongzhong's "three systems" project, suspects were given the opportunity to choose from several possible types of interrogation procedures."' Their choices shed light on both the need for reform and the degree to which they favored particular reforms. This type of empirical data provides crucial information for policymakers who wish to consider the needs and goals of non-state participants in the criminal process. "3 Wang & Zhang, supra note 89, at io6. "4 See Zuo, supra note 15, at 67. "' See id. "6 See Email from Jim Parsons, supra note 103 ("Research data can also give voice to those who have less power (in the sociological sense). By providing information that captures and describes the experiences of those who are otherwise powerless (such as children in contact with the courts, defendants without legal representation or crime suspects in police custody) research can highlight their experience in ways that they could not do alone."). " Fan & Gu, supra note 46, at 8-9. 358 The Use of Demonstration Projects to Advance Criminal Procedure Reform 359 Ill. THE CHALLENGES OF CARRYING OUT DEMONSTRATION PROJECTS Part II discussed several of the many ways in which demonstration projects complement comparative and theoretical research on criminal procedure reform. In this Part, I briefly discuss two challenges to using demonstration projects to advance criminal procedure reform: securing the cooperation of criminal justice institutions and overcoming a lack of knowledge and experience related to conducting empirical research."' A. Cooperation of Criminal Justice Institutions Demonstration projects require scholars to collaborate with criminal justice institutions to test a reform and collect the data necessary to evaluate its effectiveness and feasibility. Securing and maintaining the cooperation of these institutions is critical to the success of a demonstration project. Yet participating in demonstration projects can be time-consuming, expensive, and politically risky for overworked and underfunded criminal justice institutions. Chinese scholars have drawn on their prior relationships with state officials to secure partnerships with criminal justice institutions. As in the United States,"' Chinese legal scholars' accumulated guanxi is an invaluable resource for securing the necessary support and approvals to carry out a demonstration project."' But even a hard-won agreement to participate in a "8 Two additional impediments to carrying out demonstration projects are that they require significantly more time and funding than non-empirical comparative or theoretical research. Zuo Weimin states that empirical research takes far more- possibly ten times as much-time and energy to carry out than theoretical research. Interview with Zuo Weimin, Professor of Law, Law School of Sichuan University, in New York, N.Y. (Feb. 6, 2011). Ira Belkin notes that grants awarded by the Ford Foundation for demonstration projects on Chinese criminal procedure reform typically range from $150,000 to $200,000, with grantees and their partners also supplying substantial in-kind contributions and occasionally additional funding from other sources. Telephone Interview with Ira Belkin, supra note 102. Cf Northwest University of Politics and Law, Ford Foundation Grants, FORD FOUNDATION, http://www.fordfoundation.org/grants/grantdetails?grantid=8929 (last visited Aug. 26, 2011) (stating that the Ford Foundation recently awarded a $148,813 grant "[flor the Criminal Law Scientific Research Center to conduct a pilot project providing free criminal defense services to indigent defendants in the city of Xian"). Belkin does, however, point out that obtaining funding is generally a less significant obstacle than securing the cooperation of local criminal justice institutions or overcoming a lack of knowledge and experience related to carrying out demonstration projects. Telephone interview with Ira Belkin, supra note 102. "' Email from Megan Golden, supra note 85. 0 Securing the necessary cooperation is, of course, considerably easier if the project is supported by a central-level agency. Telephone interview with Ira Belkin, supra note 102. Examples include Chen Weidong's sentencing project, which was supported by the Supreme People's Court, Jiang, supra note 42, and Zhang Zhihui's plea bargaining Columbia Journal ofAsian Law, Vol. 24, No. 2 (2011) demonstration project may not translate into actual cooperation. In Fan Chongyi and Gu Yongzhong's first interrogation reform project, the pilot site had to be moved to another location because the head of the public security bureau who had agreed to host the project had changed his mind the last minute.12' In addition to securing the support of the leaders of partner institutions, scholars must ensure the cooperation of line-level staff. Line-level personnel may pursue different interests and respond to different incentives than the leadership of a criminal justice institution. In Chen Ruihua's bail reform project, the actual bail hearing procedures used by line-level staff substantially deviated not only from the procedures designed by the scholars leading the project, but also from the procedures adopted by the leadership of the procuratorate participating in the project.12 2 This line-level departure from established procedures can change the nature of the tested reform as well as reduce the validity of the project results. A range of factors can impact the cooperation of criminal justice actors in demonstration projects. One factor is changes in local or national criminal justice policy. In one of Song Yinghui's bail reform projects, for example, a recently issued "strike hard" (PFT) (yanda) policy impeded the researchers' goal of testing reforms that would increase the rate at which suspects are released on bail.'" Another factor is heavy caseloads. In Song Yinghui's victim- offender mediation project, procuratorates with lighter caseloads tested victim-offender mediation reforms more fully and in more cases than procuratorates with heavier caseloads.' The most common reason line-level criminal justice actors do not fully cooperate in demonstration projects, however, appears to be the performance evaluation systems (t WMJb1J) (kaoping jizhi) that play a critical role in determining the annual bonuses and career advancement of criminal justice personnel.'"2 Chinese scholars involved in project, which was supported by the Supreme People's Procuratorate, PROCEDURAL REFORM RESEARCH SUMMARY: CONCLUDING REPORT ON PLEA BARGAINING SYSTEM RESEARCH, supra note 42. Scholars' participation in these projects also ensures that their findings reach a receptive policymaking audience. '2' Du, supra note 90. 22 For instance, although the researchers in Chen Ruihua's bail reform project had hoped to randomly assign cases to the experimental group and the comparison group, the procurators participating in the project selected cases in a non-random way that minimized the negative impact that using the experimental procedures in certain types of cases might have on their performance evaluations. Chu, supra note 87, at 92-94. 123" fiJf3 [He Ting], SW 9iiE iIff +l Z 427 8 [Experimental Methods in Legal Empirical Research], m Ve$ ' F$R- . [J. NATL PROSECUTORS C.], Apr. 2010, at 77, 81. 124 Id. at 81. 12 See, e.g., A 39 [Ma Mingliang], Ir M i 9A : WF Rft $1J F [Study on Judicial Performance Appraisal System], + [H]In [JUST. CHINA], (9KiZ# [Zhang Wenjing] ed., 2009) (arguing that existing evaluation systems encourage police to conceal evidence 360o The Use of Demonstration Projects to Advance Criminal Procedure Reform 361 carrying out demonstration projects frequently lament the refusal of line-level staff to fully implement potential reforms that would negatively affect their performance evaluations.26 B. Academic Resources and Capacity Academic resources on empirical methods to study the legal system and legal scholars capable of conducting empirical research were virtually non- existent in China until recently. Relevant books or articles on empirical methods were scarce, and few legal scholars had experience carrying out empirical research, much less demonstration projects. Foreign universities and research organizations helped to satisfy this initial need by providing training and technical assistance, and continue to support Chinese criminal procedure reform by helping Chinese scholars develop the capacity to carry out demonstration projects.127 Yet Chinese legal scholars' capacity to conduct empirical research remains weak overall. Only a favorable to suspects and undermine several important goals of the criminal justice system); 5iP/fi [Wu Jianxiong], j k FJ ) A [Rethinking About the System of Procuratorial Business Assessment and its Reconstruction], VA3 A [L.J.] 2007 (laying out some of the advantages and disadvantages of using performance evaluation systems to assess procuratorial work). See also Carl F. Minzner, Riots and Cover-Ups: Counterproductive Control of Local Agents in China, 31 U. PA. J. INT'L L. 53 (2009-2010) (discussing the immense influence of performance evaluation systems over the behavior of criminal justice actors and arguing that these systems both impede the protection of citizens' rights and contribute to a range of governance problems). 126 See, e.g., Chu, supra note 87, at 93 (finding that because prosecutors are evaluated in part on their arrest, indictment, and conviction rates, they have powerful incentives not to release suspects who have already been detained); Song & He, supra note 41, 13 (finding that performance evaluation systems are an important reason that procurators tend to prosecute even in cases where prosecution is unnecessary). 127 The Vera Institute of Justice has supported demonstration projects on Chinese criminal procedure reform by helping senior scholars design and implement demonstration projects, hosting the annual Vera Fellowship in Justice Research and Innovation for junior scholars dedicated to using empirical research to advance reform, co-authoring a book with China University of Political Science and Law Professor Guo Zhiyuan on how to use demonstration projects to promote reform, and a range of other activities. For background information and a list of projects, see Legal Reform in China, VERA INST. OF JUSTICE, www.vera.org/project/china (last visited Aug. 26, 2on). The Harvard Kennedy School's Program in Criminal Justice Policy and Management has also played a crucial role in developing Chinese scholars' capacity to carry out demonstration projects. See China: Empirical Research & Criminaljustice Policy Reform, HARVARD KENNEDY SCH., http://www.hks.harvard.edu/programs/criminaljustice/research -publications/china-empirical-research-criminal-justice-policy-reform2 (last visited Aug. 26, 2on). tiny fraction of legal scholars have conducted empirical research,28 and scholars involved in demonstration projects on criminal procedure reform are concentrated at several law schools and one state research organization." The lack of knowledge and experience related to carrying out demonstration projects, which require a very different skill set than translating foreign legal materials into Chinese or critiquing a legal theory, remains one of the most significant obstacles to using demonstration projects to advance Chinese criminal procedure reform.' 30 Nevertheless, the capacity of a small group of legal academics has rapidly increased over the past decade. These scholars have published a significant body of empirical research on criminal procedure,' as well as several books on empirical legal research methodology."' These resources have set the stage for criminal procedure scholars across the country to explore an interest in empirical legal research and possibly lead demonstration projects in the future. CONCLUSION Demonstration projects are an important but underutilized complement to comparative and theoretical research on criminal procedure reform. Although scholars must overcome a range of challenges to carry out demonstration projects, the potential benefits of collaborating with criminal justice institutions to use empirical methods to evaluate reforms in practice are immense. Demonstration projects are critical to ensuring that reforms aimed at protecting the basic rights of participants in the criminal process are not only enacted, but effective and implemented in practice. t28 The importance of experience for designing, carrying out, and evaluating demonstration projects should not be underemphasized. According to the authors of a leading U.S. textbook on program evaluation, "nothing teaches how to do evaluations as well as direct experience in designing and running actual evaluations." PETER H. ROSSI ET AL., EVALUATION:A SYSTEMATIC APPROACH, at ix (7th ed. 2004). ' The lead researchers of the demonstration projects listed in Appendix I come from only five law schools and one state research organization: Beijing Normal University (Song Yinghui), Beijing University (Chen Ruihua), China University of Political Science and Law (Bian Jianlin, Fan Chongyi, and Gu Yongzhong), Renmin University (Chen Weidong), Sichuan University (Zuo Weimin), and the Supreme People's Procuratorate Institute of Procuratorial Theory (Zhang Zhihui). 130 Telephone interview with Ira Belkin, supra note 102. '' Many of these scholars are listed in Appendix I. ' These works include n2- Y [Bai Jianjun], i [THE PRACTICE OF EMPIRICAL LEGAL RESEARCH] (2oo8) and ' WiEW & [THE PRACTICE OF EMPIRICAL LEGAL RESEARCH] (*r i' [Song Yinghui] et al. eds., 2009). Chen Ruihua also devotes significant attention to empirical research in his highly practical book on the research process. MJP [CHENRUIHUAI, it&*f WrAM [WAYS OF LEGAL STUDY] 187-98 (2009). 362 Columbia Journal ofAsian Law, Vol. 24, No. 2 (2on1) The Use of Demonstration Projects to Advance Criminal Procedure Reform 363 While this article began on an optimistic note, it closes with a dose of realism. Legal scholars play an important role in Chinese criminal procedure reform, but criminal justice policymaking is a fundamentally political process. Empirical evidence produced by demonstration projects-like comparative or theoretical research-is unlikely to be the sole or even primary basis for policymakers' decisions regarding criminal procedure reform. 1 Some demonstration projects, such as Fan Chongyi and Gu Yongzhong's interrogation reform projects, may have a substantial impact on the evolution of criminal procedure in China. Other projects may have little or no impact on the course of reform. Michael Tonry has pointed out that the influence of empirical evidence on criminal justice reform "varies with time, space, and subject as well as with how astutely innovators operate within political and bureaucratic settings."' Although methodologically rigorous demonstration projects hold significant promise for the future of Chinese criminal procedure reform, they are only a tool. Legal scholars must play a leading role not only in using empirical methods to test potential reforms, but also in acting as policy entrepreneurs to maximize the impact of their findings.'35 13 See ROSSI ET AL., supra note 128, at 369-421 (suggesting that evaluation research influences policy through a complex political process that must account for the demands of many different stakeholders); Michael Tonry & David A. Green, Criminology and Public Policy in the USA and UK, in THE CRIMINOLOGICAL FOUNDATIONS OF PENAL POLICY 485, 485-510 (Lucia Zedner & Andrew Ashworth eds., 2003) (arguing that criminological research must pass through various "filters" such as the prevailing paradigms and ideologies of punishment, political and bureaucratic considerations, and official inertia to influence policy). 13 Tonry, supra note io8, at 788-89. 1 See Michael Mintrom, Policy Entrepreneurs and the Diffusion of Innovation, 41 AM. J. POL. SCI. 738, 739 (1997) (defining policy entrepreneurs as "people who seek to initiate dynamic policy change ... through attempting to win support for ideas for policy innovation ... [by] identifying problems, networking in policy circles, shaping the terms of policy debates, and building coalitions."); see also Petersilia, supra note 103 (discussing how criminology scholars can use their research to advance justice reform). Columbia Journal ofAsian Law, Vol. 24, No. 2 (2011) APPENDIX 1: SCHOLAR-LED DEMONSTRATION PROJECTS ON CRIMINAL PROCEDURE REFORM Dates of Project Lead Impementationect Rea s Issues Investigated Locations Implementation Researchers 2002-2004 Fan Chongyi Interrogation Haidian District, Beijing; 136 Gu Yongzhong Zhuhai, Guangdong 2004 Chen Weidong Pretrial discovery Shouguang, Shandong137 2004-2005 Zuo Weimin In-court testimony by wit- Chengdu, Sichuan138 nesses ("Live Witnesses") 2004-2006 Bian Jianlin Pretrial discovery Dongying, Shandong 2005 Fan Chongyi Interrogation Haidian District, Beijing; Gu Yongzhong Jiao-zuo, Henan; Baiyin, Gansu140 2005 Song Yinghui Bail and discretionary Yongkang, Zhejiangl 41 non-prosecution 2005-2006 Fan Chongyi (1) In-court testimony by Dongying, Shandong witnesses; (2) Judges' out- of-court collection of evidence; (3) Defense lawyers' role in adjudication; (4) Presiding judges' ($PJ-1k) role in adjudication; (5) Adjudication Committees' role in 136 Gu, Lawyer-on-Site During Suspect Interrogation (Pilot Test), supra note 38. ' 1J4 EE [Chen Weidong], i [Theoretical Exposition of the Shouguang Experimental Discovery Model], UJE$PJ [SHANDONG JUST.], no. 1, 2005 at 4; 2- #HAIB [Wang Chunhua & Lu Dongxiao], 8 AJilfiiEEKTh$'J Il) m 4-- N*A M 8IMHJS ijE ISJR AI A [Thoughts on Establishing a System of Discovery: The Shandong Shouguang City Court's Experiences Testing Discovery in Criminal Cases], A K ij [PEOPLE'S JUDICATURE], nO. 5, 2005 at 61. 8 Zuo et al., Pilot Research Report on Witness Court Appearance in Criminal Cases, supra note 40. 13 1 31 [ Bian Jianl in], ) :F 'EA fl7- & A&M IVR- T KiiE 9-3T-TjA Al1 I 4R3aly [Report on the Dongying City People's Procuratorate Evidence Discovery Pilot Project] (2006) (unpublished report,.on file with author). 140 EMPIRICAL RESEARCH ON INTERROGATION PROCEDURE REFORM, supra note 12. '' Song & He, Research Report on Reform of Bail and Discretionary Non-Prosecution of Minors, supra note 41. 142 Fan, Research Report on Criminal Trial Procedure and Pretrial Procedure Reforms, supra note 96, at i-n. 364 The Use of Demonstration Projects to Advance Criminal Procedure Reform 365 Dates of Project Lead Dmpementatrojeon Rea s Issues Investigated Locations Implementation Researchers adjudication; (6) The form and publication of court judgments 2006-2007 Chen Ruihua Bail Pingyi, Shandong 2006-2007 Song Yinghui Bail Qinhuangdao, Hebei 1 44 2007-2008 Song Yinghui Victim-offender Nanjing, Jiangsu; mediation Shijiazhuang, Hebei; Wuxi, Jiangsu 1 5 2008 Chen Weidong (1) Interrogation; Liaoyuan, Jilin 1 46 (2) Detention conditions 2008-2009 Fan Chongyi Interrogation Ningbo, Zhejiang; Wuxi, Jiangsu; Guangzhou, Guangdong; Wuhan, Hubei; Pu'er, Yunnan147 2008-2009 Zhang Zhihui Simplified procedure Chongqing; Shijingshan (plea bargaining) District, Beijing; Chengde, Hebei; Wuhan, Hubei; Wuxi, Jiangsu; Suzhou, Jiangsu; Shangyu, Zhejiang; Wuyuan, Jiangxi1 8 2008-2010 Bian Jianlin Victim-offender Chaoyang District, mediation Beijing1 49 2009 Zuo Weimin Sentencing Sichuan Provincelso 14 Chu, Between Experimentation and Implementation, supra note 87. 1445 f, fil M [Song Yinghui & He Ting], IRfQ iiELWRam- [Empirical Research Report on Bail] (2007) (unpublished report, on file with author). 145 Song, supra note 93, at 3-36. 146 E [Chen Weidong], Q4TPfiMA$11a-Tffr [Research Report on Detention Center Inspection Systems], &#WR) [CHINESE J.L.], no. 6,2009, at 3. 17 This project is referenced in Du, supra note 90. 148 PROCEDURAL REFORM RESEARCH SUMMARY: CONCLUDING REPORT ON PLEA BARGAINING SYSTEM RESEARCH, supra note 42. 1 Th 2$1J [Bian Jianlin & Li Ming], )fli4figi'-9ErTiA-lATM EPEAK & Ul~igAAi -i-ff f [Victim-Offender Mediation and Diversion: A Pilot Project at the Chaoyang District People's Procuratorate in Beijing] (zoo) (unpublished report, on file with author). 150 2K [Zuo Weimin], +[HJlfflJ i'k*: iXI-1iEit [Sentencing Procedure Reform in China: Misunderstandings and Corrections], 32 & *Wr [CHINESE J.L.] 149 (2010). 366 Columbia Journal ofAsian Law, Vol. 24, No. 2 (2011) Dates of Project Lead Locations Implementation Researchers 2009-2010 Zuo Weimin Criminal defense "D" County, Sichuan's' 2009-2011 Gu Yongzhong Criminal defense Haidian District, Beijing152 2010 Chen Weidong Sentencing Wuhu, Anhui 53 "' Email from Zuo Weimin, Professor, Department of Law, Law School of Sichuan University, to Thomas Stutsman, Robert L. Bernstein Fellow in International Human Rights, Vera Institute of Justice (Dec. 1, 2010) (on file with author). '5 Email from Gu Yongzhong, Professor, Department of Law, China University of Political Science and Law, to Thomas Stutsman, Robert L. Bernstein Fellow in International Human Rights, Vera Institute ofJustice (Jan. 15, zon) (on file with author). ' Jiang, supra note 42.