Investigation Management: Reform of Police Performance Metrics

Image source / Taiwan Police Trade Union Facebook Fan Page

Advocacy Points

The unscientific and unreasonable performance-metric system should be reviewed to prevent frontline police officers from resorting to improper enforcement practices or suffering from overwork. The approaches are as follows:

Police agencies should establish reliable and valid performance evaluation mechanisms aimed at measuring management conditions and workplace issues, without directly tying them to personnel promotions.

New performance evaluations should incorporate indicators related to working conditions and procedural justice, abolish performance systems or items that are excessively trivial, sweeping, unexecutable, or generate vicious competition, and the amendment process should incorporate the participation of frontline police officers.

Police agencies should make public their performance evaluation criteria and calculation methods for public scrutiny. Furthermore, a research and management team combining frontline units and scholars should be established to review and assist in improving the performance system.

Supplementary measures should include a police union and a whistleblower system to strengthen internal resistance against unreasonable performance demands and unlawful orders.

Introduction

Current discussions on police officers' non-compliance with due process in law enforcement mostly focus on the issue of officer competence. However, visits conducted by our association with internal members often reveal opinions such as: "Under performance pressure, there is no way to care about all of that." In addition, performance pressures indeed tend to drive police officers toward high-risk enforcement methods such as vehicle pursuits, which frequently make the news when they result in casualties among officers and the public. Alternatively, due to performance pressures, officers may choose enforcement methods tainted with legality doubts, such as the proliferation of consent searches, arbitrary standards for reasonable suspicion stops, and violations of the principle of good faith in traffic violation inspections through covert photography, ambushes, etc.

Furthermore, the performance system has in fact become a primary factor for supervisors to exploit grassroots personnel. Even ignoring basic rewards and punishments, ordinary workplaces still exhibit improper leave control due to performance issues—such as denying leave to those with poor performance—and officers being more readily assigned to so-called "undesirable and demanding" duties. Psychological stress arising from performance issues, which has even led to suicide (occurring frequently among low-to-mid-level managers such as police station directors and branch team leaders), is also a major issue.

Performance issues even become a factor in some court judgments where judges "sympathize" or are "forced by public sentiment to sympathize" with law enforcement officers, preventing them from examining the legality of enforcement patterns from an objective, third-party perspective. Consequently, much police and criminal procedure jurisprudence faces backlash for being "at odds with practical realities." However, what the police circle calls "jurisprudence failing to match practice"—because strict jurisprudence imposes greater limitations on enforcement—should actually be translated as "jurisprudence conflicting with the policy direction of the performance system." Yet, the performance system itself lacks legitimacy; rule-by-performance neither aligns with our country's concept of a constitutional state nor allows the entire police enforcement apparatus to undergo proper reform.

In other words, the current performance system has actually eroded and even destroyed our frontline enforcement model, but this system lacks complete legitimacy, creating a collective connivance wherein "the police are helpless, scholars cannot intervene, and the courts sympathize yet remain unable to be objective."

Our country's police performance system was originally created to reform past internal promotion chaos. Today, however, it can be seen that while performance metrics—compared to the relatively rigid personnel systems of the past—do somewhat mitigate arbitrary promotions and reassignments, subjective and arbitrary reassignments or promotions still persist within police agencies. Furthermore, it is an open secret that promotions and performance often bear no apparent connection. Instead, this situation paradoxically causes mid-to-high-level supervisors to become even more superstitious and obsessed with performance figures, explicitly or implicitly signaling frontline personnel to ignore due legal process.

That is to say, if we wish to change the current state of police enforcement, the normalization of the performance management system is a necessary task.

As for our country's current performance evaluation and management system, commonly used performance items can be categorized into two operational models:

Establishing Benchmark Scores and Detailed Scoring Methods
The problem with this approach is that the determination of relevant benchmark scores is opaque, and disputes over performance scoring methods frequently arise. Such evaluation methods are often uniform standards formulated by the central government, yet they simultaneously serve as a means for the central authority to control local regions. The standards are actually inconsistent and give rise to sudden increases in workload driven by the fluctuating standards of the central government, meaning the ruling authorities.

Calculating Weekly Incident Occurrence and Clearance Benchmarks, Followed by Periodic Reviews
Related review methods suffer from the same problem. Because social phenomena are complex and random, occurrence and clearance numbers cannot actually be accurately evaluated. Yet, it is common to see evaluations of these figures calculated down to two decimal places (e.g., "the number of burglaries occurring in this district shall be lower than 10.53 cases"), clearly demonstrating that such evaluations are fictional numbers detached from reality and science. Misapplications of relevant formulas (or the complete non-existence of evaluation formulas) even result in benchmark values that defy common experience and are extremely difficult to achieve.

Both performance evaluation items and management systems mentioned above are frequently criticized for ignoring regional characteristics, manpower, and other factors, relying merely on crudely determined figures. Moreover, due to the erroneous interpretation of "treating projections as targets," methods such as concealing cases, rejecting cases, or taking shortcuts frequently appear in order to meet projected figures whose essence is clearly "uncertain." Under such circumstances, "projected figures"—indicators that inherently should contain margins of error—turn out to be infallible and therefore require no adjustment.

Performance System Reform Proposals

Rationalization of the Performance System
By analogy with the principles of administrative planning, prospects and evaluations should be put forward, project goals established, and, in particular, the legality of the plans underlying performance examined.

Performance is essentially an evaluation of the effectiveness of administrative plans. However, our country's practice has consistently failed to review, research, or discuss the effectiveness or prospects of the plans or projects behind such performance; many plans lack discussion and research regarding their policy connotations and effects. For example, current anti-drug plans essentially boil down to setting performance targets and demanding everyone meet them. Yet, in practice, it remains inconclusive whether achieving performance should be interpreted as "vigorous police enforcement" or "good results in our country's drug prevention," because they are fundamentally contradictory. Our country's police agencies actually do not know what their working goals are, nor can they handle this contradiction.

On another level, formulating authorities should evaluate relevant plans to see whether the enforcement tools provided by current laws are actually practicable, and whether they will cause police officers to take desperate risks and choose illegal or dangerous enforcement methods. Evaluations and research on this aspect have historically been absent from relevant plans.

Making performance definitions effective; units setting performance should explain the definitions and objectives of these indicators.

Our country's performance setting often produces extremely stringent items, such as subdividing "drugs" into "juvenile drugs," and further subdividing juvenile drugs into "juvenile internet drugs." Yet, in reality, the stricter the item, the harder it is to achieve, and many performance items are formulated merely out of demand without any serious discussion of "whether they can possibly be achieved." Is there really a necessity to make such fine classifications? Taking the previous example, whether juvenile drugs need to be independent is debatable, but if "juvenile drugs" are actually controlled, "juvenile internet drugs" can naturally be brought under control as well—so why add this subcategory? Part of this stems from superiors' lack of understanding regarding emerging concepts.

The evaluation of our country's criminal investigation efficiency actually suffers from management difficulties caused by overly granular items. Furthermore, during final scoring, controversies are frequently so numerous that fine categorization is simply abandoned. This stems from a failure to assess the effectiveness of indicators beforehand and a failure to examine whether various sub-indicators can be simplified and merged. If subcategories are to be discussed, qualitative analysis and interpretation of relevant data should be conducted ex post facto, rather than narrowing down items at the very outset, which prevents normal performance evaluation from being achieved afterwards.

Scientization of Performance Evaluation Benchmark Data
This part primarily discusses indicator items used to forecast and evaluate the "future," such as "occurrence numbers/clearance numbers." To date, no police agency has publicly explained the sources, calculation benchmarks, or formulas of these performance figures, leaving frontline personnel suspicious of the scientific validity of the benchmarks and unable to verify them.

Therefore, police agencies should make public the relevant calculation benchmarks and methods, and open themselves up to criticism.

However, it is even more likely that certain performance indicators lack any scientific calculation and evaluation benchmark at all. If that is the case, under such unscientific evaluations, are such evaluations or benchmark values even necessary?

Current evaluation methods, such as those used for traffic performance, employ "regression analysis." However, the premise of regression analysis is that a correlation exists between the independent variable and the dependent variable. Yet, "time itself" does not necessarily correlate with the occurrence of traffic accidents or violations. Therefore, one cannot help but doubt whether police agencies themselves understand the nature of the evaluation tools they use.

Additionally, as mentioned earlier, evaluations of related performance empirically fail to account for differences in regional characteristics, relying merely on the analysis of past data, meaning their evaluative effectiveness still requires review.

Grassroots Personnel Should Have Channels for Direct Participation and Questioning of Performance Indicators
Regarding performance appraisals, frontline enforcement personnel are actually more familiar with on-site conditions and variations, and may hold more concrete assessments of plan feasibility and indicator effectiveness. Part of the problem with our country's performance management system lies in its inability to grasp detailed conditions nationwide. Current research on performance management also suggests that performance should be discussed with or even agreed upon by frontline personnel, and mechanisms should be provided for grassroots personnel to voice reflections and feedback when performance appears unreasonable.

Through exchanges with other civil servants, the Taiwan Police Trade Union discovered that certain administrative agencies hold review meetings allowing grassroots personnel to directly question whether performance indicators are reasonable, complete with grievance remedies if they are unsatisfied with the discussion results. When discussing this with those civil servants, they found the rigidity and unquestionability of the police agency's performance system to be unbelievable.

Quality Management of the "Performance System" Itself
Reviewing and inspecting the quality of performance itself to eliminate arbitrary indicators and unscientific benchmark values.

Practical units and scholars should be convened to separately establish a research and management team dedicated to our country's performance system. This team would conduct comprehensive discussions and quality management concerning our police plans and performance system, examine the effectiveness and feasibility of each indicator, and review the formulas used to calculate benchmark values—essentially establishing a substantive "performance measurement system" to reduce the arbitrariness and dictatorial problems plaguing current performance planning.

At this level, we can also introduce the academic system of conducting ex-ante and ex-post qualitative research on performance itself, concretizing the functions of performance indicators to discard the obsolete and introduce the new.

Excessive evaluation items represent excessively high costs; the central government should control the quantity of performance items.

One of the problems with our country's performance system is the excessive number of performance items. We should actually control how many performance systems currently exist in our country. Too many performance items only create management difficulties because current manpower cannot possibly effectively execute so many types of performance, and repetitive evaluations often occur between different performance areas. These can actually be unified and streamlined.

Furthermore, regarding the bloating of performance due to competition among agencies at all levels, the central government should exercise unified control and jointly supervise whether various units excessively exaggerate performance indicators.

Implementation of Rolling Revisions for Performance
Current performance management actually employs a "rolling revision" method, which involves adjusting performance baselines. However, the specific methods for such rolling revisions remain unclear, and these revisions fail to address the continuous exaggeration of performance figures by various units, frequently resulting in upward revisions.

These practical problems can be resolved through qualitative methods or by evaluating external indicators of other social conditions for revisions. For obvious "overheating" (exaggerated and unreasonable performance figures), criminological research methods can also be used for correction, or even strategic control, such as canceling evaluations. Because, as mentioned above, overheated performance often stems from excessive demands for performance figures, which conceals enforcement risks or even illegal situations. From a management perspective, it would be better to abandon such policies.

Elimination of indicators or systems that are foolish, uncontrollable, self-contradictory, or even obstructive to work.

One problem with current performance indicators is the failure to distinguish between "controllable factors" and "uncontrollable factors," frequently setting performance requirements that are practically impossible to execute or control—such as the case in Taipei that made the news: "jurisdictions without electronic gaming arcades must still have electronic gaming arcade crackdown performance."

Additionally, there are indicators that create vicious competition in practice, such as "reviews and punishments when cases within a jurisdiction are solved by other units." The purpose behind designing such an indicator is completely incomprehensible. In practice, disputes over performance allocation or claiming credit between supporting units can also lead to situations where units refuse to seek support.

New Performance Indicators
Incorporating working hours, manpower, labor conditions, and casualty statistics (including generalized death from overwork) into performance evaluations as gauges of "whether performance indicators and planning are normal."

Under conditions of abnormal performance pressure and overwork, police personnel will naturally tend to choose "shortcut" methods, such as various controversial or risky enforcement practices. Moreover, our country also utilizes abusive duty-dispatch methods, such as the abuse of spot checks and road checkpoints. Under such duty methods, performance will naturally grow, and such rolling revisions will forever "roll upward." However, this positive feedback simultaneously implies overall duty risks. Furthermore, due to psychological burnout caused by enforcement patterns, it is foreseeable that police officers will tend to use illegal or borderline methods.

Current performance indicators lack an evaluation of on-site labor conditions, and the introduction of such indicators would actually serve to test the reasonableness of performance itself. As discussed in Section 2, Item 3 regarding the rolling revision of performance, aside from simply observing whether performance figures are abnormally heating up, numerous casualties, instances of overwork, and even deaths from overwork occurring during enforcement mostly stem from performance formulations exceeding reasonable metrics. Therefore, labor conditions should be incorporated into evaluations to serve as indicators for revising or lowering performance targets.

Incorporating "procedural justice" as a performance evaluation item—externally to fulfill the ideals of a constitutional state, and internally to eradicate "cheating"—as one of the indicators for assessing whether performance is normal.

As mentioned above, alongside psychological burnout caused by overworking for performance and abusing manpower, the police circle also harbors the bad habit of viewing "walking the edge of regulations" as "ability." Outstanding performance often implies numerous controversies in duty execution. Such borderline maneuvers for the sake of performance generate an effect that raises overall performance baselines (because performance is good). This stimulus drives everyone to choose "cheating" and creates a cycle of hostility toward "individuals who question police enforcement."

Therefore, on the one hand, relevant enforcement controversies should be treated as a basis for lowering performance—breaking the cycle of "violating the law to secure performance." On the other hand, "public questioning of law enforcement" should become "an opportunity to rationalize overall police workloads" rather than merely being treated as "adding difficulties to police enforcement."

Introducing "external indicators" to establish external metrics as a basis for evaluating internal indicators and sharing responsibility.

Objective observation and evaluation indicators of social phenomena—such as analyses of economic conditions, demographic conditions, social welfare conditions, and educational conditions—mean that assessing public order in a region should involve examining various other relevant indicators alongside the crime rate to compare regional differences and comprehensively evaluate the status quo, rather than discussing only a single crime rate and dumping public order pressure onto the heads of police agency chiefs. Only then is it possible to effectively reduce the connivance or even demand for enforcement models that skirt the legal boundary.

In reality, the evaluation of public order conditions should also incorporate discussions on the effectiveness of various related units across municipal governments (bureaus of education and social welfare). Public order has never been solely a police issue, yet current local municipal government administrative practices are accustomed to pinning all problems on police agencies, which is actually detrimental to the integration of the entire crime prevention system.

Neutralization of the Performance System, Decoupling from Personnel, and the Necessity of Internal Checks and Balances
The so-called neutralization of performance and decoupling from personnel means that the impact of performance on management models should have reasonable boundaries. For example, practical operations still witness improper leave-deduction or shift-scheduling problems driven by performance. Grassroots personnel criticize that if they are already overworked yet performance is lackluster, creating even more overwork through punitive measures merely forms a negative feedback loop—making that person forever the worst-performing and most overworked. Therefore, managers themselves should possess this management consciousness, treating work improvement as a matter of measuring management conditions and workplace problems for reasonable allocation, rather than intuitively using leave as a tool to coerce performance.

Another widely criticized aspect of the current performance system is the rampant issuance of rewards and punishments. The reward and punishment system for ordinary police officers is already abnormal and overly abused compared to general civil servants. Consequently, to achieve project results, performance rewards and punishments often create unreasonable gaps. Alternatively, due to regional characteristics, certain areas inevitably have poorer performance or even none at all; being penalized simply for lacking performance due to uncontrollable regional characteristics is frequently viewed as a reason for performance unreasonableness.

From another perspective, the above problems also point to the establishment of unions and whistleblower systems, enhancing means of resistance against internal unreasonable performance demands and even unlawful orders. The longstanding problem with our country's internal police performance is that labor conditions and the legality of law enforcement have not been incorporated into performance evaluations; current police performance is built upon the foundation of "ignoring labor rights and the spirit of the rule of law."

Conclusion

Reforming the performance-metric system is a bidirectional process. It requires changing the perspectives of both the public regarding police law enforcement and the police regarding public dissent. Whether enforcement controversies or the exercise of the right to dissent are viewed as "making trouble" or as "an opportunity to re-examine police labor conditions" merits deep reflection. Based on the preceding analysis of abnormal police labor conditions, the legality of law enforcement, and the performance-metric system, we must ponder whether this can reconcile the past binary opposition of "troublemakers vs. police" and "human rights vs. public order."

As performance metrics form the basis of police management, if we wish to reform the current law enforcement conditions of police personnel, we must demand that this system acknowledge and notice that law enforcers are human beings, and highlight the fact that "law enforcers are human beings" within it. Only then will this system guide the police in the right direction.