Journal of Police and Criminal Psychology

, Volume 25, Issue 1, pp 1–16

A Comparison of Law Enforcement Divorce Rates with Those of Other Occupations

Authors

  • Shawn P. McCoy
    • Department of PsychologyRadford University
    • Department of PsychologyRadford University
Article

DOI: 10.1007/s11896-009-9057-8

Cite this article as:
McCoy, S.P. & Aamodt, M.G. J Police Crim Psych (2010) 25: 1. doi:10.1007/s11896-009-9057-8

Abstract

It is a common belief that the divorce rate for police officers is higher than that of the general population. This belief is commonly held in spite of the fact that there is no empirical research supporting such a belief. To compare the divorce rate of law enforcement personnel with the rates for other occupations, we analyzed data from the 2000 U.S. Census. The results of this analysis indicate that the divorce rate for law enforcement personnel is lower than that of the general population, even after controlling for demographic and other job-related variables.

Keywords

Law enforcement divorce
It is a common belief that the divorce rate for police officers is higher than that of the general population (Honig 2007; Kappeler et al. 2000). For example, books by Jurkanin and Hillard (2006); Territo and Sewell (2007); Wells and Alt (2005) suggest higher than normal rates of divorce for police officers, and Aamodt (2008) reported finding the following statistics on the Internet:
  • Surveys of police officers continually reflect estimates of divorce rates as high as 75%

  • Police officers…have a high divorce rate, about second in the nation.

  • Compared to national averages, police officers have been reported to have…double the divorce rate…

  • Police officers, for example, face divorce rates averaging between 66 and 75 percent.

  • The police profession has the highest rates of divorce, alcoholism, domestic violence, and suicides.

Although the notion of a higher divorce rate for police is often publicized, there seems to be little empirical support for such a notion. Consequently, this paper will begin by reviewing the research on law enforcement divorce rates and then report an analysis of occupational differences in divorce rates based on 2000 census data.

Why Would Divorce Rates be Higher in Law Enforcement?

There are two theories for why divorce rates would be higher for law enforcement personnel than for the public in general: the job itself and the personality of the people who enter the profession.

The Stress of Being a Police Officer

Policing is thought to be a highly stressful job. For example, Dantzker (2005) states:

Although many occupations share similar stressors – such as shift work, supervisory problems, and improper or poor equipment – there are those that are inherent to policing. These include responding to unpredictable situations, periods of work that range from inactiveness to extremely stressful activity back to periods of less stressing activity, instantaneous decision making, the court system, and the use of force (Dantzker 2005, page 277).

Other studies show that there is little evidence to support the idea that policing has unusually high levels of stress (Brown and Campbell 1994). Evidence showing policing to be highly stressful “often is methodologically limited or contradictory (Brown and Campbell 1994).” These two contradicting views leave much to be researched. If police officers do have a high divorce rate, can it be empirically connected to stress?

The Police Personality

A second theory for potentially higher divorce rates in law enforcement is that the people who go into law enforcement are “more prone” to divorce or marital problems than are people in other occupations. However, meta-analyses indicate that regardless of the psychological or personality test used, police applicants and police officers have normal, psychologically healthy profiles (Aamodt 2004). Thus there seems to be no support for this potential explanation.

Are Law Enforcement Divorce Rates Actually Higher?

Although it is commonly thought that the divorce rate for law enforcement personnel is above the national average, there has been relatively little research on the topic. Furthermore, the available research not only is dated, but does not offer support for the notion of higher divorce rates.

Studies investigating differences in divorce rates across occupations have generally used one of two approaches: analysis of census data or surveys.

Analysis of Census Data

There have been two studies using census data to compare divorce rates across occupations, both of which have been highly cited. Lichtenberger (1909, 1963) was one of the first researchers to publish findings on the relationship between divorce rates and occupation. Using data from the 1900 U.S. census, Lichtenberger rank-ordered 39 occupations on the basis of divorce rates. As shown in Table 1, the rank for law enforcement personnel (30th) was well below the average rank of the 39 occupations. It should be pointed out that although most sources cite Lichtenberger’s work as being published in 1963, his findings were first published in the Columbia University: Studies in History and Economics and Public Law in 1909 and reprinted in 1963; making his study much older than one would think given the often cited 1963 publication date.
Table 1

Results of Lichtenberger’s (1909; 1963) analysis of the 1900 Census Data

Rank

Order of Divorce Rate (highest to lowest)

1.

Actors, professional showmen

2.

Musicians and teachers of music

3.

Commercial travelers

4.

Telegraph and telephone operators

5.

Physicians and surgeons

6.

Barbers and hair dressers

7.

Servants and waiters

8.

Bartenders

9.

Restaurant and saloon keepers

10

Hotel keepers

11.

Tobacco and cigar factory operatives

12.

Printers, lithographers and pressmen

13.

Bookkeepers, clerks, stenographers

14.

Steam railroad employees

15.

Painters, glaziers and varnishers

16.

Bakers

17.

Laborers (not specified).

18.

Agents

19.

Salesmen

20.

Butchers

21.

Tailors

22.

Plumbers and gas and steam fitters

23.

Machinists

24.

Merchants and dealers

25.

Lawyers

26.

Bankers, brokers and officials of banks

27.

Masons (brick and stone)

28.

Boot and shoe makers and repairers

29.

Teachers and professors in colleges

30.

Watchmen, policemen, firemen, etc

31.

Manufacturers and officials

32.

Engineers and firemen (not locomotive)

33.

Miners and quarrymen

34.

Carpenters and joiners

35.

Farmers, planters and overseers

36.

Blacksmiths

37.

Draymen, hackmen, teamsters, etc.

38.

Clergy men

39.

Agricultural laborers

A more recent, but still dated, study on police divorce rates was conducted by Whitehouse (1965) using data from the 1960 census. Similar to Lichtenberger, Whitehouse (1965) found that the divorce rate for police officers was lower than the national average. Although the Whitehouse (1965) study is commonly cited, only three sentences in the entire article relate to divorce rates:

However, the United States Bureau of Census revealed that policemen and detectives have 1.7 percent of their number listed as divorced. Of the total male population fourteen years or older, 2.1 percent are divorced. However, all males in the same age distribution as police men and detectives have a divorce rate of 2.4 percent (Whitehouse 1965, page 31).

Survey Studies

Three studies have investigated occupational differences in divorce rates by surveying law enforcement personnel. All three of these studies were conducted in the 1970s.

Durner et al. (1975) administered a marital satisfaction survey to officers in three departments: Baltimore, Chicago, and Santa Ana (CA). Although the focus of the study was on marital satisfaction in general, Durner and his colleagues reported wide differences in divorce rates among the three departments. The rates ranged from Baltimore at 17 percent, Santa Ana at 27 percent and Chicago at 33.3 percent. Although these figures were interpreted as indicating high divorce rates, the article did not provide any norms with which these figures could be compared nor was the article clear regarding what these figures actually represent. That is, are they the percentage of officers who ever got divorced or are they the percentage of officers per year who got divorced?

Niederhoffer and Niederhoffer (1978) surveyed 30 police departments throughout the continental U.S., asking officers for their current status of being married, single, or divorced. The researchers received responses from 50,000 police officers, which translated into “a 10 percent sample of the total sworn police complement in the United States.” Of the 50,000 officers who responded, only 2.5% were divorced. The researches compared this percentage to the national divorce rate of 3.7% found in Statistical Abstract of the United States and concluded that the divorce rate for police officers was actually lower than the U.S. average. Because the New York City police department had such a large response (31,196), Niederhoffer and Niederhoffer computed the divorce rate on the remaining 29 departments, finding a divorce rate of 4.5%, which was slightly higher than the 3.7% found nationally.

Kroes et al. (1974) focused on 100 Cincinnati police officers and the stress they perceived on the job. Of the 100 men interviewed, 81 were married, and 5 were divorced, a divorce ratio of 6.17%. However without any national divorce ratio with which to compare this divorce rate, no conclusions can be drawn from this study.

Current Study

The purpose of the current study was to compare the divorce rates of law enforcement personnel with the divorce rates from other occupations. Consistent with the small amount of previous research, we hypothesized that the law enforcement divorce rate would be lower than that of the general working population

Method

To obtain information on divorce rates for a variety of occupations, we accessed the 2000 U.S. Census using a special program developed by the Census Bureau, Data Ferret, which can be downloaded at http://dataferrett.census.gov/dataferrettapplicationinstall.exe. Because a detailed breakdown of jobs was needed to isolate the law enforcement profession, we opted to use the 2000 People and Housing One Percent Sample Census Survey. This dataset contained both a detailed breakdown of occupations as well as employees’ current (2000) marital status.

The census collects data on 449 occupations. These occupations are divided into 23 major groups, 96 minor groups and 449 broad occupations. For police officers, the major group is Protective Service Occupations (33), the minor group is Law Enforcement Workers (33-3000), and the broad occupation is Police Officers (33-3050).

To compute the divorce rate for each occupation, we used the following formula:
$$ \left( {{\text{Separated}} + {\text{Divorsed}}} \right) \div \left( {{\text{Total}}\,{\text{Population}} - {\text{Never}}\,{\text{Married}}} \right). $$

This formula yielded the percentage of people in each occupation that had been in a marital relationship, but were no longer with their spouse.

So that we could control for demographic variables that might be related to divorce rates, we also obtained race, gender, age, and income information for each occupation. Correlational analyses revealed that divorce rates were higher for occupations with higher percentages of African Americans (r = .49) and women (r = .26), lower for occupations with higher percentages of Asian Americans (r = -.14), and lower for occupations with higher average incomes (r = -.53). Because of the significant relationships of these variables to occupational divorce rates, these data were entered into a regression to estimate what the divorce rate should be for a given occupation after controlling for the demographic (gender, race, age) and income characteristics of each occupation. The results of this regression analysis are shown in Table 2.
Table 2

Regression results

Regression Statistics

Multiple R

0.62

    

R Square

0.38

    

Adjusted R Square

0.38

    

Standard Error

0.04

    

Observations

512

    

ANOVA

     
 

df

SS

MS

F

Significance F

Regression

5

0.5246

0.1049

63.25

.0001

Residual

506

0.8395

0.0017

  

Total

511

1.3641

   

Variable

Coefficients

Standard Error

t Stat

P-value

 

Intercept

0.24871889

0.0185

13.42

0.0000

 

African American

0.26185158

0.0334

7.83

0.0000

 

Asian

-0.09441692

0.0494

-1.91

0.0565

 

Percent Male

-0.02612271

0.0071

-3.69

0.0002

 

Average Age

-0.00120629

0.0005

-2.67

0.0079

 

Ave Income

-0.00000081

0.0000

-6.54

0.0000

 

It should be noted that we also rated each of the occupations on the extent to which it involved three sources of occupation stress: shift work, overtime, and weekend work. Because none of these variables accounted for a significant amount of variability in the regression, they were removed from further analyses.

Results and Discussion

A complete listing of the 449 broad occupations and their divorce rates can be found in the Appendix. To examine occupational differences in divorce rates after controlling for race, gender, and income, we divided the actual divorce rate by the rate predicted by the regression equation. Occupations with ratios above 1.0 had divorce rates higher than expected, whereas occupations with ratios lower than 1.0 had lower rates than expected.

Table 3 shows the actual divorce rates; expected divorce rates based on race, gender, and income; the actual divorce rate divided by the national average divorce rate; and the actual divorce rate divided by the expected rate for law enforcement and corrections positions. As shown in Table 3, the divorce/separation rates for law enforcement occupations (14.47%) was lower than both the national average (16.96%) as well as the rate expected given the demographic and income characteristics of the law enforcement workers (16.35%). When law enforcement workers are broken down into broad occupations, the divorce rates for police officers (15.01%), supervisors (12.75%), detectives (12.53%), and railroad transit police (5.26%) are lower than the national average as well as the expected rate after controlling for demographics. However, the rates for animal control officers (19.02%), fish and game wardens (25.53%), and parking enforcement officers (26.25%) are higher than the national average as well as the demographically controlled expected rate.
Table 3

Actual and expected divorce rates for law enforcement and corrections occupations

Occupation

N

Divorce Rates

Ratios

Actual %

Expected %

National Norms

Expected

Law Enforcement

954,615

14.47

16.96

88.50

85.32

 Transit & railroad police

1,143

5.26

15.13

32.17

34.77

 Detectives

122,125

12.53

15.72

76.64

79.70

 Supervisors

129,970

12.75

14.98

77.98

85.11

 Police officers

675,756

15.01

17.52

91.80

85.67

 Animal control officers

14,051

19.02

17.14

116.33

110.97

 Fish & game wardens

3,885

25.53

14.12

156.15

179.79

 Parking enforcement

7,685

26.25

22.54

160.55

116.46

Corrections

492,542

21.30

21.37

130.28

99.67

 Officers

437,616

21.54

21.49

131.74

100.23

 Supervisors

54,926

19.58

20.38

119.76

96.07

All Census Jobs

 

16.35

   

Looking at corrections positions provides an interesting picture. As shown in Table 3, the rates for corrections officers (21.54%) and corrections supervisors (19.58%) are higher than the national average, but are lower than, or equal to, the demographically controlled expected rate. Thus, it appears that corrections workers have divorce rates similar to their race, gender, age, and income matched peers.

As shown in the Appendix, the five broad occupations with the highest divorce rates are dancers and choreographers (43.05%), bartenders (38.43%), massage therapists (38.22%), gaming cage workers (34.66%), and extruding machine operators (32.74%). The five occupations with the lowest divorce rates were media and communications equipment workers (0.00%), agricultural engineers (1.78%), optometrists (4.01%), transit and railroad police (5.26%), and clergy (5.61%).

This study demonstrates that the idea that divorce rates are unusually high for law enforcement workers is unfounded. Such results are not surprising as meta-analyses and other research indicates that when it comes to a variety of issues ranging from personality profiles to suicide rates to the ability to detect deception, law enforcement personnel are similar to the general public (Aamodt 2008; Honig 2007).

In using census data, one limitation of this study is that we do not know if law enforcement personnel on average get divorced more often than employees in other occupations. That is, the Census data provide a snapshot of the percentage of people that were currently divorced or separated at the time they completed the 2000 census. As pointed out by a colleague, it is possible that law enforcement personnel get divorced more often and that a better comparison would be to compare the average number of divorces per person in each occupation. Unfortunately such data are not available, and the possibility that such data would yield results different from what we found is purely speculative and without any empirical support.

Copyright information

© Springer Science + Business Media, LLC 2009