Front page, Jobs and Training

Embracing the inevitable – using cognitive computing to recruit police

Men and women run around a dirt track.
Correctional cadets run the track at the Correctional Training Center in Galt, undated.

By Lucas Abarca, Special Agent

Editor’s note: Agent Abarca wrote this as part of his Command College class. Learn more about his experience at Command College.

Driving on empty streets, shopping in half-empty stores, living a socially distanced life seems surreal. The situation is reminiscent of a Hollywood plot, an unrealistic story. Amid the COVID-19 quarantine, though, many people have binge-watched movies while socially isolated from friends and coworkers.

In one superhero movie, the villain eliminated half of all life with a snap of his fingers. The villain described himself as “inevitable.” A similar inevitability is occurring in the criminal justice system. Like Thanos in “The Avengers,” generational changes are inevitable, and have resulted in what some have called a “workforce crisis” in the criminal justice system. Fewer young people are applying to become police officers, and more are leaving the profession, often after only a few years on the job. (PERF, 2019).

COVID-19’s economic impact will pause the workforce crisis in the near term, masking its symptoms. A tightened labor market with fewer opportunities will likely reduce agency concerns about attrition and recruiting in the short term. To remain viable after the COVID-19 pandemic, though, agencies must address systemic issues that created an inevitable workforce crisis of officer attrition and fewer candidates.

Most efforts to address workforce shortages focus on recruitment strategies to solve the crisis. Unlike a superhero movie, no “time heist” or Iron Man reverse snap will fix recruitment challenges. Changes to peace officer retention programs, however, will have the most significant long-term impact to staffing needs. Only when retention challenges are prioritized and addressed will agencies reap the rewards from improved recruitment methods. Effective retention strategies, augmented by Cognitive Computing, may forge a symbiotic relationship between recruitment and retention to resolve long term challenges.

The symptoms

According to a 2019 report from the Police Executive Research Forum (PERF), the criminal justice system is facing a recruitment and retention crisis. Fewer people are applying to become peace officers, more people are leaving the profession before reaching retirement age, and the attrition rate is increasing among current staff before reaching retirement age and service requirements.

Nationwide, applications for police officer positions have decreased by 60% from 2010 to 2019. About 8.5% of current officers are eligible for retirement, and 15.5% will become eligible within five years. The most common reason for voluntary resignations within officers’ first five years of employment is to transfer to another law enforcement agency; the second most common is to pursue a career outside of law enforcement (PERF, 2019).

Police departments are not alone in the struggle. Nationally, Departments of Corrections have higher attrition rates; as high as 53 percent in some states, and fewer applications for recruits, resulting in a 45 percent vacancy rate in some states (RAND, 2018). Many of the underlying reasons for the recruitment and retention crisis are due to generational changes.

COVID-19’s economic impact will likely result in short term recruitment gains. The economic tumult that accompanies it are transformative events, and will probably alter lifestyle and financial choices for every generation. In the short term, increased unemployment rates will drive interest in stable careers.

Millennials might seek law enforcement careers since their life experience includes three seismic events: 9/11, during which nearly all of them were under the age of 18; the 2008 Great Recession; and COVID-19. For many in Gen Z, the economic challenges of COVID-19 will completely reframe their college and work decisions. Even though interest will likely increase, without pre-screening improvements, candidate and employee engagement will result in long-term challenges (Glazer, 2020).

Understanding COVID-19’s impact on recruiting and retention will likely be similar to what occurred in previous economic crisis. During the last recession in 2009, the New York Police Department (NYPD) had a 54% increase in applications from 2007 to 2008.

The wider pool of candidates allowed the NYPD to tighten selection standards. However, due to budget shortfalls and hiring restrictions the department size shrank with attrition. When the economy rebounded, the NYPD capitalized on the increased interest and hired recruits better suited for the profession (Schmidt, 2009).

Embracing the cause

Law enforcement agencies are experiencing a generational shift. In the near future a complete change in staffing, leadership, and management will occur across the three generations now in, and entering, the law enforcement workforce:

  • Generation X (Gen X) or the senior officers and leadership, born from 1965 to 1980, are now retiring and will continue through 2030.
  • Generation Y (Millennials), current officers and future leadership, born from 1981 to 1996, will not reach retirement age until 2031 at the earliest, and younger Millennials will be working into the 2050s.
  • Generation Z (Gen Z) or the rookie officers, born from 1997 to 2012, began their law enforcement careers in 2018. As with the rise and fall of all generations, they will assume leadership roles and Millennials will retire.

What can we learn from generational changes in law enforcement? In 2016, Gallup released a study about employee engagement and interests of the different generations. The primary difference between Millennials and Gen X is worker engagement.

Gallup found that only 29% of Millennials are engaged at work, meaning only about three in 10 are emotionally and behaviorally connected to their job. The majority of Millennials are predominantly checked out, not putting energy or passion into their jobs (Gallup, 2016).

Gen Z and Millennials expect instant gratification and personalization in the workplace. They want careers focused on their individual goals, work style, and preferences. They work for the culture and not only pay. They want an opportunity to influence their organization and need recognition from their supervisor and coaches (He, 2019).

Given these realities, how can agencies better understand employees on a personal level and keep them engaged? About 80% of Fortune 100 companies use personality assessments to build stronger, more effective teams and healthier organizations.

There are many personality models; the most popular are DISC, MBTI, Enneagram, and Big Five. The Myers-Briggs Type Indicator (MBTI) is the most widely used personality assessment in the world, with more than 3.5 million administered annually (Bajic, 2015).

Personality affects everything people do; choosing career paths, maintaining relationships, communicating with others, and spending free time. Agencies that identify an individual employee’s personality can create more effective career pathways, training opportunities, and upward mobility.

They can also match newer employees with mentors to keep them more engaged at work. The challenge for any agency to understand staff at a personal level is daunting, and why the workforce crisis seemingly inevitable. Small agencies might be able to manage such a task with talented managers and supervisors.

No snap of the fingers, however, can minimize the challenge for large agencies. For large agencies to understand employees at such a micro level is impossible without technological innovations. The solution to retain engaged employees exists through cognitive computing.

Innovation

Cognitive computing uses machine learning, natural language processing, and data mining techniques emulating the way the human brain reasons and makes decisions. A cognitive computing system processes information and considers parameters and variables similar to the way humans might choose which restaurant to eat at or which car to buy.

Cognitive systems learn and improve through every data point, interaction, and outcome, building a deep and broad knowledge base that is always up-to-date (Bhattamishra, 2019).

When cognitive systems finish their analysis, they can provide what they think is the best choice for a given problem from an array of possible solutions. This is not necessarily the right choice, however. It leaves it to the human who is using the system to decide what the right course of action is in a given situation (Loeffler, 2019).

The best-known example of a cognitive computing system is IBM Watson. In 2011, IBM Watson won a one million dollar prize by beating two of the game show Jeopardy’s all time champions. Watson uses natural language processing and machine learning to make sense of the huge amounts of data available today, including unstructured data from social media posts, blog posts, articles, reports, and enterprise system data.

Watson can even answer complex questions. For example, during Jeopardy, Watson was able to understand the dynamics of the game and responded correctly “What is Las Vegas” to the question “This town is known as ‘sin city’ & its downtown is ‘glitter gulch’” (Software Nation, 2016).

In peace officer retention, a cognitive computing system could look at various data sources, including an employee’s professional experience, personality, and past performance, and then further analyze them against the characteristics of other successful jobholders to recommend where they are likely to find success in the organization. This is not a dream of the future; cognitive assessments are being used today in other industries.

Schneider Electric, an energy management company established in 1836, created an internal talent marketplace for their 150,000 employees in over 100 countries. Within the first two months, they successfully assigned over 150 employees to new roles.

Schneider Electric used Innermobility, an artificial intelligence solution, to identify personalized opportunities for each employee based on their unique skills, experiences, and ambitions. Innermobility learns each employee’s unique professional profile and pairs it with a spectrum of available opportunities across the organization, from part-time projects to full-time positions and mentorships (Saidy, 2019).

This type of employee-job matching can be especially valuable in larger organizations. Cognitive computing can personalize its interactions with employees when it knows where employees have worked, what jobs they liked and disliked, what jobs interested them, and their personality type. Large agencies have multiple career path options for job classifications, so these systems can help match the agency’s needs to the skills and interests of employees who might otherwise be unaware the opportunity exists.

For example, a new officer or deputy might start working in patrol or a correctional facility. As new officers gain experience, they will work different roles, some suiting them better than others. The cognitive computing system will ask them to rate the different roles, identifying their interests.

For the most accurate information, the application needs job preference information from experienced staff. Through electronic inquiries, the cognitive system can evaluate all sworn and non-sworn personnel, increasing the sample size and broadening the knowledge base.
The cognitive computing system also evaluates an employee’s training record.

Agencies with a Learning Management System (LMS) that can enhance officer training with cognitive computing. When staff receive a new assignment, for example, the LMS can send them short YouTube-style videos showing how to perform the job’s unique functions. The same videos provide a means for the system to share personalized career opportunities with staff.

Learning from all existing staff, the system advertises opportunities for upward mobility, lateral career pathways, training opportunities, and matches staff with coaches/mentors. LMS will also further evaluate each individual’s interactions with agency enterprise systems and update their digital identity accordingly.

According to Josh Bersin, an expert in human resources technology, Fuel50 is one example of cloud based system that integrates with an enterprises human resource data, learning data, career architecture data, and job vacancies data.

Fuel50 also allows employees to self-assess their interests and capabilities, it matches them to jobs, and it facilitates the creation of personalized career paths and easy to navigate job models that make career management easy.

Fuel50 matches staff to mentors and identifies development programs to help staff with their career growth (Bersin, 2020). All of these capabilities provide a glimpse into the future of cognitive computing to help recruit the right people, retain them, and develop their expertise.

Inevitable coaching and supervision

Cognitive computing can also augment staff communication skills. In 2015, a software company named Crystal developed and released what they call the first Personality AI.

Personality AI takes multiple types of inputs; text samples, demographic data, real life observations, questionnaire responses, and outputs personality insights.

Personality AI accurately identifies how people like to communicate, what motivates, energizes and stresses them, and how they like to interact with others. Personality AI outputs a personality model comparable to a personality assessment like DISC, MBTI, the Big Five, and OCEAN (Skloot, D’Agostino, 2019).

This approach can allow an agency to reach employees on a personalized level, facilitating more effective supervision and coaching to develop the employee’s potential.

The cognitive computing system provides situation-specific suggestions when communicating based on a staff members digital profile. Similarly, the application sends supervisors daily prompts about their team with suggestive interactions related to work performance and behaviors the application identifies.

The application would not force interactions, but augment a supervisor’s awareness about the team. Keeping coaches and supervisors engaged with employees will improve retention and employee engagement. The engaged workforce is the most powerful recruiting tool to reach the rising generations.

Inevitable recruitment

A recent study revealed Gen Z and Millennials trust friends and family more than external sources like website reviews, news media, celebrities and athletes. The study also explained how most young Americans are interested in becoming social media influencers, with 5% of Gen Z and 48% of Millennials motivated by the opportunity to make a difference in the world (Morning Consult, 2020).

The engaged workforce of sworn and non-sworn staff can recruit their friends and family simply by sharing their experiences in social media and in person. Within a large law enforcement agency, the employee retention process advertises upward and lateral mobility available to all staff, sworn and non-sworn. Satisfied employees will likely share the opportunities with people in the sphere of influence extend to tradition and non-traditional candidates.

Long term recruiting strategies must include non-traditional peace officer candidates. The traditional pipeline of candidates from the military and families with a history of police service no longer provide the same number of recruits.

The conventional pipeline of candidates responds well to legacy recruiting strategies focused on dramatic or exciting parts of the job. Non-traditional candidates respond well to a more realistic (what the job actually entails) service-oriented message.

The non-traditional candidates are interested in the community service aspects of the job (PERF, 2019). Agencies might find success raising awareness with non-traditional candidates through outreach at high schools, community colleges, and universities.

The time-intense, multi-step peace officer selection process is ideal for the use of Cognitive Computing. The process generally follows a variation of the following; application, reading and writing assessment, physical fitness examination, oral interview, background investigation, truth verification, medical evaluation, and psychological evaluation.

The copious amounts of data gathered from and about each candidate are evaluated to determine suitability for peace officer employment. The Cognitive Computing system would process the data, and then identify common characteristics of successful candidates.

The system would classify candidates and predict their success in the selection process. The resulting score would provide organizations with a pool of candidates to process by their suitability. This assessment takes place at each stage of selection, with the highest-ranking candidates identified at each step along the way.


The cognitive computing system also would maintain regular communication with candidates via Chatbots or smart digital assistants. Millennials and Gen Z have no issue initiating conversations without a live person on the other end. Chatbots will answer questions, schedule interviews, and maintain candidate engagement. Candidates would also complete their personality assessment online, connecting their personality with employee retention data.

During the selection process, the system will advertise personalized career opportunities, share videos about the career, and initiate human intervention when it detects negative emotions. The system will continuously evaluate these interactions and adjust the candidate suitability rank when necessary (Roddy, 2019). All of the tools and information the system learns through employee retention will shape how it interacts with candidates.

First steps to the inevitable

Cognitive computing will enhance officer recruitment and retention in the near future provided agencies are willing to change their culture and embrace the new technologies. Cognitive computing requires so much computing power it is cloud based.

Agencies need their data in a secure cloud server for the application to evaluate. Cognitive computing platforms access the data through Application Programming Interface (API) allowing agency information technology departments to easily integrate the solutions with their enterprise systems.

The API’s are program specific, meaning an agency can choose to use how many services they desire. For example, An agency can use IBM Watson for chatbots now, adding personality assessments later. Fees for service are based on usage configured with monthly or annual fees.

The most critical first step requires a change to software-based recruitment, hiring, training, and retention systems. Agencies can, with minimal impact on their budget, implement new software solutions that will enable the capture of unstructured data that cognitive computing can use in the future.

Large volumes of data will improve system accuracy, large agencies will benefit first, but smaller agencies can see benefits by capturing the data early and, over time, improving accuracy.

Conclusion

With cognitive computing, recruitment and retention form a symbiotic relationship. Retention programs feed the recruiting system by linking employee behavior and personalities with peace officer candidates. The system will use the same means to market peace officer careers in a personalized manner keeping candidates informed and engaged from initial application to a tenured officer.

A broadened candidate pool will improve the system accuracy and find non-traditional candidates. The non-traditional candidate pool is the key to fixing the recruitment and retention crisis; they will diversify the agency in the short term and help bring other non-traditional candidates toward law enforcement careers in the long term. The needed culture change throughout law enforcement can accelerate ahead of historical trends and improve public sentiment overall.