Where Should Criminal Policy Research in Taiwan Head?

Subgroup 5 of the National Conference on Judicial Reform will discuss the sub-topic "Establishing Research Mechanisms/Institutions, Utilizing Judicial Statistics, and Improving the Quality of Judicial Practical Operations" on May 11. The Crime Prevention Research Center of the Academy for the Judiciary under the Ministry of Justice submitted a report, suggesting that more resources should be invested in itself, its staffing expanded, and its administrative rank elevated to a level-3 central government agency directly subordinate to the Ministry of Justice, thereby enhancing the research capacity and influence of criminal policy. While research certainly requires resources, simply mentioning the need to infuse more resources without reviewing past problems and bottlenecks in criminal policy research is unlikely to truly resolve the issues.

Recommendations[1]

All criminal justice-related databases (including police administration, legal affairs, and judiciary) should release blank forms and coding methods under open licenses to allow the public to inspect whether the methods of collecting and coding statistical data should be reviewed and improved.

After balancing the promotion of knowledge development, public interest, and personal privacy, criminal justice-related databases should be released in a reasonable manner. The released content should at least include database files (in different formats such as SPSS, STATA, EXCEL/TSV, etc.) and codebooks. For databases involving microdata, the release methods can refer to the practices of databases in the public health field.

Establish a Specialized Center for Criminal Policy Research under the Research Center for Humanities and Social Sciences of Academia Sinica, and permit the research center's personnel to make reasonable use of the databases.

The transparency of the organizational operations and decision-making of the Crime Prevention Research Center should be strengthened.

Invite researchers from the social sciences familiar with quantitative research methods to assist in reviewing the problems in the "Crime Conditions and Analysis" report, exchange ideas, and establish basic requirements and directions for improvement in quantitative analysis. Opportunities for cross-disciplinary cooperation should also be sought during this process, such as in statistics, economics, psychology, sociology, public health, and related fields.

Invite researchers from the social sciences familiar with quantitative research methods to assist in reviewing the effectiveness of the punitive criminal policy, and reports should be submitted. If effectiveness cannot be reviewed, reasons should be explained along with how rigorous effectiveness evaluation methods should be established in the future. Cross-disciplinary cooperation should likewise be sought for this part.

Openness of Statistical Databases

The openness of criminal justice-related databases is crucial for criminal policy research. Taiwan has long lacked rigorous criminal policy evaluations. One reason is that researchers with research capabilities find it difficult to access official raw databases and are unclear about what data are available for analysis; another reason may be that the scope and categories of data collection in official databases are problematic, rendering the data insufficient for analysis.

Currently, Taiwan's criminal justice-related statistics are either selected by officials from statistical items they deem important for public release (such as statistical yearbooks on police administration, legal affairs, and the judiciary) or made available for public query online (such as the Police Statistics Query Network and Advanced Query for Legal Statistics). However, such methods of releasing statistical data suffer not only from incomplete content and arbitrary merging of statistical categories, but they are also not provided as database files, making it impossible for researchers to read files using statistical software (such as SPSS or STATA) for operation and execute various statistical procedures. Currently, a data release method that better meets researchers' needs is similar to the Judicial Yuan's practice for the "Survey on Public Perception of the Judiciary". This database is included in the Survey Research Data Archive (SRDA) of the Center for Survey Research under Academia Sinica's Research Center for Humanities and Social Sciences, which publicly releases database files, questionnaires, and codebooks. However, this database is relatively distant from criminal justice.

Currently, other criminal justice-related databases have not been released in an open manner. To view the coding methods or use these databases, one must currently compete for research projects from relevant units, yet only a very small number of people can successfully secure such research projects. Such a situation is actually unfavorable to research development and the clarification of the current status.

To have good criminal policy research, official authorities must release databases under open licenses and make them available to researchers and the public in a reasonable manner. Therefore, this article proposes the following recommendations:

All blank forms and coding methods of criminal justice-related databases (including police administration, legal affairs, and the judiciary) should be released under open licenses.

Databases do not generate data automatically; they inevitably require someone to log the data or programs to capture data. Social sciences generally refer to this data-logging process as "coding." Some forms filled out by citizens and public officials may actually constitute preliminary data collection, which may then be directly logged into a database (such as online forms) or re-coded by someone before being logged.

In order to understand the scope and categories of data collection in criminal justice-related databases, it is recommended that all blank forms and coding methods of the databases be released, with the release of coding methods referencing the standard release format of general database codebooks. For the preparation of codebooks, Academia Sinica's SRDA can be consulted; this is not difficult and does not involve the infringement of personal privacy. Once blank forms and coding methods are released, researchers interested in this subject can help confirm whether the scope and categories of data collection are problematic and submit recommendations to the competent authorities.

After considering the balance between promoting knowledge development, public interest, and personal privacy, criminal justice-related databases should be released in a reasonable manner.

What requires more careful consideration regarding the scope and method of release is the database file itself. The release of database files involves whether researchers can conduct analyses using the data. All database release methods should include database files compatible with common statistical software in different formats such as SPSS, STATA, EXCEL/TSV, etc., and a codebook is naturally an essential element.

Databases can be divided into two categories: microdata and macrodata. Microdata refers to data where the individual is the unit, such as an inmate's entry into prison, covering gender, birth year, date of entry, and correctional facility. Macrodata, on the other hand, consists of databases aggregated from smaller units, such as the number of male inmates serving definitive sentences of over one year in a specific correctional facility, the number of female inmates serving definitive sentences of over one year, total male inmates, total female inmates, and so on. Taiwan's criminal justice-related agencies are more familiar with the release of macrodata; however, the macrodata categories they release do not necessarily match research needs, nor are the release formats database files, which is disadvantageous to researchers.

Regarding the release of macrodata databases, it is purely a matter of whether statistical categories meet research needs and does not involve personal privacy. How to release this part should be open to public solicitation of opinions from researchers after all database coding methods are published. Additionally, one may refer to the database release methods in the National Archive of Criminal Justice Data (NACJD). In addition to databases, the NACJD website also lists what research has been conducted utilizing the databases, from which one can trace back to understand how databases are utilized.

Regarding the release of microdata databases, it more heavily involves balancing the promotion of knowledge development with personal privacy. However, data of this kind is extremely important for evaluating the effectiveness of treatments received by individuals entering the criminal justice system, especially records of recidivism. Regarding how microdata-related databases should be released, they are roughly divided into direct release to the public and restricted users. The former requires more complete de-identification and other privacy protection practices; the latter may involve partial de-identification or the provision of raw data depending on user permissions, coupled with the imposition of obligations of confidentiality. Taiwan has more experience with database usage rules in public health related to microdata, because such data involve personal privacy such as medical behaviors and diseases, which can serve as a reference. The Government Open Data License, Version 1.0 may also be used as the licensing terms. In short, the direction should be toward openness, striving to avoid a situation where only a handful of applicants for government research projects can use the data.

Establishment of Research Institutions

Establish a Specialized Center for Criminal Policy Research under the Research Center for Humanities and Social Sciences of Academia Sinica, and permit its personnel to make reasonable use of criminal justice-related databases.

In criminal policy research, the soundness of relevant databases and their openness to researchers outside the government are of vital importance. Considering that the government may have many reservations about opening data to the general public or researchers, leaving the ultimately released data fragmented, it is recommended that at least Academia Sinica is also a government agency with an exalted academic status; thus, the use of criminal justice-related databases should be made available to Academia Sinica researchers as unrestricted as possible. However, different personnel should still have varying access restrictions to the databases, downloads of files should be logged, and obligations of confidentiality regarding personal data must be upheld. Academia Sinica may also participate in discussions or provide opinions on how relevant units determine database coding methods and reasonable methods of use.

Academia Sinica currently has a Research Center for Humanities and Social Sciences engaged in cross-disciplinary research. Criminal policy research itself possesses an interdisciplinary nature, potentially intersecting across sociology, psychology, economics, and even public health (such as drug policy). Coupled with database usage, a Specialized Center for Criminal Policy Research could be established under Academia Sinica's Research Center for Humanities and Social Sciences to elevate the importance of criminal policy and strengthen the reasonable use of databases.

Strengthen the transparency of the organizational operations and decision-making of the Crime Prevention Research Center.

The report by the Crime Prevention Research Center of the Academy for the Judiciary, Ministry of Justice (hereinafter referred to as the "Crime Prevention Center") argues that the state should invest more resources in the Crime Prevention Center, expand its staffing, and elevate its administrative rank to a level-3 central government agency directly subordinate to the Ministry of Justice, thereby enhancing research capacity and influence in criminal policy. However, intra-governmental think tanks of this nature may suffer from a lack of independence, easily degenerating into entities that endorse policies rather than making policy recommendations and evaluations based on evidence. In fact, one of the reference targets proposed by the Crime Prevention Center itself, the National Institute of Justice (NIJ) in the United States, has faced such criticism. The U.S. National Research Council (NRC) critiqued the operations of the NIJ in 2010, pointing out problems such as insufficient independence of the NIJ, conducting research based on political mandates, an inadequate scientific research culture, and lack of transparency, while putting forward reform recommendations. The NIJ subsequently issued a statement of self-evaluation and organizational adjustment in 2011. The Crime Prevention Center's report glosses over these issues, demanding only the injection of personnel and resources.

Given Taiwan's circumstances, it is probably difficult to expect strong independence from an in-house government Crime Prevention Center. Under such circumstances, this article argues that the research center should at least prioritize the transparency of its organizational operations and decision-making. For example, how is the research direction of the center determined? How are research project subsidies decided? Which researchers applied for subsidies, and what were the reasons for rejecting or approving the subsidies? Data collected through research (especially questionnaire surveys) should be reasonably made public, and the final outcomes of research projects should also be made public, with their subsequent utilization subject to scrutiny. For instance, did the research findings genuinely translate into policy implementation? What are the policy's effectiveness evaluations? How frequently is the research cited by other researchers? Related recommendations on transparency can also reference the report issued by the NRC in 2010.

Strengthening Criminal Policy Research

Invite researchers from the social sciences familiar with quantitative research methods to assist in reviewing problems in the "Crime Conditions and Analysis" report.

This article suggests that before further expanding its establishment, the Crime Prevention Center should thoroughly review the problems in the "Crime Conditions and Analysis" reports it previously produced. Looking at past editions of "Crime Conditions and Analysis," one can see that these reports contain almost no analysis, consisting merely of piled-up statistical figures.

Rather than spending manpower and financial resources every year piling up these figures, it would be better to invite researchers from the social sciences familiar with quantitative research methods to propose analytical recommendations based on open data in comparison with past "Crime Conditions and Analysis" reports. There are not many experts in quantitative methods within Taiwan's crime prevention community; some senior researchers are accustomed to conducting research using focus interviews combined with foreign literature collection, which is far from sufficient. Under the approach of conducting criminal policy research through empirical methods,[2] the proficiency and mastery of quantitative research methods are crucial keys. To strengthen and stimulate the development of criminal policy research, researchers in fields such as statistics, economics, psychology, sociology, and public health who are familiar with quantitative research methods and interested in criminal policy should be invited to jointly review how the "Crime Conditions and Analysis" report can be improved. Conducting rigorous quantitative research requires data that are not limited to currently released criminal justice statistics categories, which is why the report invariably suffers from a lack of analysis. Through data opening and inviting pan-social science researchers to jointly review the "Crime Conditions and Analysis" report, establishing basic requirements and improvement directions for quantitative analysis is an important step in attracting more talent to invest in this field.

Invite researchers from the social sciences familiar with quantitative research methods to assist in reviewing the effectiveness of the punitive criminal policy and help establish rigorous effectiveness evaluation methods.

Regarding the review of the effectiveness of punitive criminal policy, strictly speaking, it should fall within the scope of reviewing the "Crime Conditions and Analysis" report; however, because evaluating the effectiveness of this policy is of high importance, this article specifically singles it out for discussion. A segment of the public harbors an inexplicable sense of security toward punishment, as if the mere existence and intensification of punishment solves problems, thereby halting the evaluation and reflection on policy effectiveness and creating the illusion that they no longer need to face problems. Drug policies and drunk driving policies both exhibit this kind of situation. Unfortunately, under the current trend of populist policies, policy declarations in criminal policy that are impossible to achieve and act like drinking poison to quench thirst—such as "drug-free society," "zero tolerance," and "resolutely cracking down"—continuously emerge.

This is not to say that punishment is completely ineffective; in fact, current research has already shown that the existence of punishment itself has a deterrent effect. The problem lies in the marginal deterrent effect of punishment, the side effects brought by punishment, and whether other alternative approaches might exist. For a literature review on this part, one can consult the article written by Chung Hung-pin, "A Review of Empirical Research on the General Deterrent Effects of Punishment". Regrettably, Taiwan currently has no rigorous effectiveness evaluation whatsoever regarding the long-term development of punitive criminal policy. Professors Tsai Te-hui and Yang Shih-lung are highly influential scholars in the crime prevention community, and they have conducted a study titled "A Study on the Deterrent Effects of Punitive Criminal Policy on Recidivism", but their research methodology is flawed. The research by Professors Tsai Te-hui and Yang Shih-lung merely distributed questionnaires to judicial personnel and inmates, asking them whether they "feel" punitive criminal policies have a deterrent effect and what makes for a stronger deterrent effect, entirely omitting recidivism rates. In research terms, however, one should examine how policies influence people's subjective perceptions and how those subjective perceptions subsequently influence recidivism. The research by Professors Tsai Te-hui and Yang Shih-lung only isolated the subjective perception segment without addressing the subsequent impact on recidivism. This article cannot ascertain the reasons for adopting this research method, except to say that this is not a rigorous policy evaluation method.

Examining policy effectiveness from the indicators of outcome evaluation is an extremely important component, yet Taiwan's criminal policy pays no heed to this practice. This article suggests inviting researchers from the social sciences familiar with quantitative research methods to assist in reviewing the effectiveness of the punitive criminal policy and submit a report. Even if policy effectiveness cannot be evaluated, reasons should be stated, alongside proposals on how to establish rigorous effectiveness evaluation methods in the future.

Only by opening up data so that more people with diverse expertise can invest in criminal policy research and put forward rigorous policy analyses can we truly halt punitive legislative directions that lack evidentiary foundations yet are driven by the need for a sense of security; only then can we review the waste of government resources and infringements on human rights caused by past practices of punitive criminal policy, moving toward the establishment of a social safety net that respects human rights and is pragmatic and effective.

Notes

[1] The ideas in this article stem from the concept of "evidence-based" judicial reform research advocated by Attorney Huang Chih-hao. This concept was proposed by Attorney Huang based on his psychological science research and legal practical experience during our interview with him. The writing of this article benefited from discussions with Chung Hung-pin, a Ph.D. student in the Faculty of Law at Humboldt University of Berlin; Kao Shih-hsuan, a student at the Interdisciplinary Graduate School of Law at National Taiwan University; and Assistant Professor Tsai Po-fang from the Graduate Institute of Medical Humanities at Taipei Medical University, to whom we hereby express our gratitude.

[2] In fact, sound criminal policy research can never emphasize quantitative methods alone; qualitative methods and theoretical contemplation are both vital. However, criminal policy research in Taiwan relies too heavily on literature reviews and focus groups, and this practice is more suited for preliminary opinion collection and problem clarification. Taiwan's reflection on punitive criminal policy has already discussed a great deal in theory, but what kind of effect the overall policy has ultimately produced remains unclear. This part still has to be explored through quantitative methods.