The Impact of Public Support and Scrutiny on the Retention and Recruitment of Law Enforcement Officers

We will conduct two surveys. Survey 1 will be administered among sworn law enforcement officers in the state of Arizona. Survey 2 will be administered among working-age adults in Arizona, California, Nevada, Utah, Colorado, and New Mexico who are not law enforcement officers. In both surveys, respondents will answer questions about their demographic characteristics, their own work experiences, and their preferences over alternative work experiences.

2023-07-24

2023-08-25

Retention/turnover intentions among sworn law enforcement officers.

Recruitment intentions among people who are not sworn officers (i.e., their interest in pursuing a career as a law enforcement officer).

Retention/turnover intentions of law enforcement officers will be measured as Likert scale responses to three questions: (1) “To what extent do you currently have a desire to pursue a career outside of being a law enforcement officer?”; (2) “To what extent do you currently have a desire to pursue a career as a law enforcement officer in another department?”; and (3) “Please indicate how likely it is that you will be working as a law enforcement officer in Arizona in the future?” The first question gauges one’s overall retention/turnover intention. The second question gauges whether one wants to stay in the profession, albeit in a different location. The third question gauges the temporal nature of one’s retention/turnover intention (e.g., one might be very certain they’ll still be in the profession a year from now, but they might think it is very unlikely that they’ll still be in the profession five years from now).

Recruitment intentions of people who are not sworn officers (i.e., their interest in pursuing a career as a law enforcement officer) will be measured as a Likert scale response to the question: (1) “To what extent do you currently have a desire to pursue a career as a law enforcement officer?” In addition, we will also ask the following question: (2) “Based on your responses to the questions above, we can send you a curated list of law enforcement officer job postings in and around your zip code. If you would like to receive information about these job postings, please provide your email address in the box below.” The first question gauges one’s overall interest in pursuing a career as a law enforcement officer. The second question/prompt provides respondents with a binary prompt (either input their email address or not), which will gauge their immediate interest in learning about the available job opportunities in their area.

Heterogeneous responses across demographic groups (e.g., by age, race, gender, and sexual identity/orientation).

Survey 1: Responses to questions about burn-out, career ambitions, and department culture.

In both surveys, we will ask respondents about their demographic characteristics. We will estimate whether the treatment effects on the primary outcomes of interest differ based on the age, race, gender, or sexual identity/orientation of the respondents.

Survey 1: We will ask the following questions, which will be responded to using a five-point Likert scale. “I feel emotionally drained from my work.”; “Working with people all day is really straining for me.”; “I feel burned out from my work.”; “I worry that this job is numbing me emotionally.”; “I feel that I am positively influencing other peoples’ lives through my work.”; “I feel very committed to protecting and serving citizens in my jurisdiction.”; “Most of my social interactions outside of work are with other law enforcement officers.”; “I aspire to earn a promotion in a more senior position within my department.”; “I aspire to work in a specialty position within my department (e.g., detective, drug unit, traffic unit, etc.).”; “I would forcibly remove a fellow officer from a situation if their behavior violated policy (e.g., using excessive force).”; “I feel comfortable calling out fellow officers if they do not adhere to protocol.

All details of the experimental design are in the "(Hidden)" response box below.

Respondents of both surveys will be randomized into one of six different treatment cells, which will vary the exposure they receive to information about the public’s level of support for law enforcement officers. This will be done automatically via the Qualtrics block randomization feature. The treatment and control group details are as follows:
(1) Control Group - No information provided at all (i.e., a blank block).
(2) Treatment, Neutral Support - Provides information that highlights both support for and scrutiny of law enforcement officers.
(3) Treatment, High Support by Country - Provides information that highlights both support for and scrutiny of law enforcement officers. Also describes factual survey results that indicate high levels of support for law enforcement officers nationwide.
(4) Treatment, Low Support by Country - Provides information that highlights both support for and scrutiny of law enforcement officers. Also describes factual survey results that indicate low levels of support for law enforcement officers nationwide.
(5) Treatment, High Support by State - Provides information that highlights both support for and scrutiny of law enforcement officers. Also describes factual survey results that indicate high levels of support for law enforcement officers nationwide, with an emphasis on the respondent’s state.
(6) Treatment, Low Support by State - Provides information that highlights both support for and scrutiny of law enforcement officers. Also describes factual survey results that indicate low levels of support for law enforcement officers nationwide, with an emphasis on the respondent’s state.

Qualtrics Block Randomization Feature

Individual

No

Survey 1 will be distributed among all law enforcement officers in Arizona (N ~ 16,000). We anticipate a completion rate of about 20% (N ~ 3,200).

Survey 2 will be completed by ~2,000 Arizona adults; ~1,000 California adults; ~1,000 Nevada adults; ~1,000 Utah adults; ~1,000 Colorado adults; and ~1,000 New Mexico adults.

Survey 1 will be distributed among all law enforcement officers in Arizona (N ~ 16,000). We anticipate a completion rate of about 20% (N ~ 3,200). Survey 2 will be completed by ~2,000 Arizona adults; ~1,000 California adults; ~1,000 Nevada adults; ~1,000 Utah adults; ~1,000 Colorado adults; and ~1,000 New Mexico adults.

Respondents will be randomized into one of 6 treatment cells. So, we anticipate a roughly equal number of individuals from each group in each of the 6 treatment cells.

For Survey 1, with an expectation of 3,200 respondents, that would be 533 individuals per cell.

For Survey 2, with an expectation of 7,000 respondents, that would be 1,167 individuals per cell.

For responses made on five-point Likert scales. With a control group mean of 3.0, a standard deviation of 1.0, a power of 0.80, N per cell of 533, the minimum detectable effect size is 0.172.