Marital Status Distribution of the U.K. Military: Does It Differ From the General Population?

ABSTRACT The U.K. media suggest that U.K. military personnel have high divorce rates; to date, these claims are not substantiated. Marital status distribution of the general population and military were compared using data from the Office for National Statistics' marital projections and a military cohort study. Overall, military personnel were more likely to be married and less likely to be divorced than were the general population. Women in the military and married military personnel younger than 30 years of age are more likely to report divorce. Military welfare services might target these groups with programs assisting marital relationships.

The U.K. media have reported anecdotally that the divorce rates of military personnel have risen over time (BBC News, 2000). However, to date there has been no research from the U.K. to support these claims. Research from the United States indicates that divorce rates among U.S. military personnel are comparable to civilians (Karney & Crown, 2007;Karney, Loughran, & Pollard, 2012;McCone & O'Donnell, 2006). Despite this, Pollard, Karney, and Loughran (2008) reported that the U.S. media have continued to raise concern; moreover, spouses of military personnel are also concerned (Karney et al., 2012). Pollard and colleagues (2008) suggested that spouses' concerns have grown from beliefs that military families are more vulnerable than comparable civilians; however, this is not substantiated by U.S. research.
Since military operations began in Iraq (2003) and Afghanistan (2001), there have been increased demands placed on U.K. military personnel (Rona et al., 2014). U.K. research suggests that marital difficulties could ensue as a consequence of military deployments in those who deploy for more than 13 months in a 3-year period (Rona et al., 2014). Other factors found to be associated with relationship difficulties among U.K. military personnel include childhood adversity, lack of support, and financial difficulties (Keeling, Wessely, Dandeker, Jones, & Fear, 2015), things that are likely to impact relationships regardless of military service. In the United States, concern has been raised about increased marital difficulties amongst military personnel; however, these concerns are not supported by divorce figures (Karney & Crown, 2007). Research does suggest that negative relationships with spouses can lead to military personnel developing mental health difficulties, which can affect their ability to complete their job and discourages them from reenlisting (Hoge, Castro, & Eaton, 2006). In the United Kingdom, research has highlighted the importance of personal relationships and contact with family and friends in providing support to military personnel (Greene, Buckman, Dandeker, & Greenberg, 2010).
Research from the United States that has compared differences between the rates of marriage and divorce between the military and general population indicates that U.S. military personnel are more likely to be married (Cadigan, 2000;Karney et al., 2012) and marry at younger ages (Adler-Baeder, Pittman, & Taylor, 2006;Hogan & Seifert, 2010;Karney & Crown, 2007;Lundquist, 2007) than are age-matched civilians. It is proposed that younger age at marriage is more common in the military as a result of job and financial security (Cadigan, 2000;Lundquist, 2007). Moreover, benefits for married military personnel, such as subsidized housing and being part of a supportive environment, may lead to marriage happening prematurely or even accelerating marriages in partnerships that might have otherwise dissolved (Cadigan, 2000;Karney & Crown, 2007;Lundquist, 2007). Lundquist (2007), however, found that 23 to 27 year old enlisted personnel (in the United States, enlisted personnel are all ranks below commissioned officer) are more likely to divorce than comparable civilians, even after controlling for demographic, religious, socioeconomic, and attitudinal factors. Hogan and Seifert (2010) found that active duty Armed Forces members who marry aged between 23-25 years have higher divorce rates compared with those who have been married but not served on active duty.
Key differences exist between the U.S. and U.K. military that may impact marital relationships; for example, U.S. operational deployments tend to be longer than U.K. operational deployments (approximately 12 months compared with 6 months). To date, in the United Kingdom, the marital status distribution of the U.K. military has not been compared with the marital status distribution of the general population. There is a clear need to do so to understand if the claims made by the media are correct and to better understand the marital relationships of the U.K. military to provide appropriate support to military personnel and their families. This article compares the marital status distribution of the U.K. military with the general population of England and Wales. On the basis of research from the United States, it is hypothesized that compared with civilians: 1. A larger proportion of U.K. military personnel will be married compared with the general population; this will be most evident in those younger than 30 years of age. 2. Despite the larger proportion of married military personnel, in comparison to the general population, a larger proportion of U.K. military personnel under 30 years of age will be divorced.  Hotopf et al., 2006). Nonresponse was mainly due to difficulties contacting personnel as a result of training, deployments, or being posted to a new location (Iversen, Liddell, Fear, Hotopf, & Wessely, 2006). There was no evidence of response bias in terms of health outcomes or fitness for deployment (Tate et al., 2007). Participants from Phase 1 were asked to participate at Phase 2. Phase 2 also included two further samples. The HERRICK sample (the name of operational tours to Afghanistan) was recruited to represent the United Kingdom's expanding involvement in Afghanistan and the replenishment sample to represent those who had joined the military since Phase 1. Phase 2 data were collected between November 2007 and September 2009 using selfcompletion questionnaires, which were sent to potential participants. The response rate for phase 2 was 56% (n D 9,984; Fear et al., 2010).

Data source: England and Wales general population
The Office for National Statistic data are available for people ages 16 years and older who lived in England and Wales, mid-2008 (n D 34,402,000) and provided marital status distribution by age group. The data are derived from statistics on marriage, divorce, and death registrations collected through administrative sources, maintained by the General Register Office and the Ministry of Justice (Office for National Statistics, 2011). Data from the ad hoc output Principal Marital projection datasheets units v2, from the Statistical Bulletin: Marital status population projections, 2008 based data from the Office for National Statistics (2010) were used as this provided data by single year of age (from 16 years), allowing the extraction of the data for 18 to 64 year olds.

Measures
In the KCMHR military cohort study, marital status was assessed using a seven-option question that asked the following: are you: married; living with partner; in a long term relationship; single and not in a long term relationship; separated; divorced; or widowed. Principal Marital projection datasheets units v2, from the Statistical Bulletin: Marital status population projections, 2008-based data from the Office for National Statistics (2010) used marital status categories: never married, married, divorced, or widowed; those who were separated but still legally married were categorized, by the Office for National Statistics, in the married group. To make comparisons between the Office for National Statistics and KCMHR data, the KCMHR marital status categories were categorized to replicate the Office for National Statistics marital status categories (see Figure 1).

Study samples
Military: Only data from Phase 2 of the cohort study was used for the purpose of this study. Of the 9,984 participants from Phase 2 of the KCMHR cohort study, 9,934 (99.5%) provided marital status information and were included in this comparison. Of these, 8,752 (88.1%) were male and 1182 (11.9%) were female. This sample includes regular and reserve U.K. military personnel and those who were serving and had left service at the time of questionnaire completion.
England and Wales general population: Extracting marital status for 18-64-year-olds from the Office for National Statistics data created a sample of 33,981,858 individuals. Of these, 16,962,772 (49.9%) were male and 17,019,086 (50.1%) female.

Data analysis
Sample weights for the military data were created to reflect the inverse probability of a participant from a specific subpopulation and specific engagement type (regular or reserve) being sampled. Response weights were also created to account for nonresponse. Response weights were defined as the inverse probability of responding once sampled and driven by factors shown to empirically predict response (gender, rank, age, and sample). Based on the assumption that the data are missing at random and that the observed variables modeled to drive nonresponse were correctly identified, the weighted analyses proved valid results. A combined weight was generated by multiplying the sample and response weights (Fear et al., 2010).
Weighted percentages were calculated for the military sample and compared with the Office for National Statistics percentages. To achieve the most meaningful comparison, marital status was investigated by age group (comparable between each sample; 18-29, 30-44, and 45-64 years) and gender. Percentages and total numbers are presented for both military and general population samples. 95% confidence intervals are also presented for the military sample due to the relatively smaller sample size. Because of type of data available from the Office for National Statistics, statistical analysis of the difference in prevalence between military sample and the Office for National Statistics percentages was not useful.

Ethical approval
The KCMHR cohort study received full ethical approval both from the MoD Research Ethics Committee 0732/ 117 and King's College Hospital Research Ethics Committee (NHS REC reference: 07/Q0703/36).

Results
Overall, military personnel were more likely to be married and less likely to be divorced compared with the general population of England and Wales (Table 1). The higher proportion of marriage in the military is most notable in the 18 to 29 years group. Although overall the proportion of divorce is lower compared with the general population, in the 18-29 years group the percentage of military personnel who are divorced is higher, however, this difference is small. These results support the hypotheses of this study. The prevalence of being widowed in both the general population and military population is low, however, lower in the military sample. This may be due to difference in the samples as the military sample has fewer participants in the older age groups (Table 1). Comparisons stratified by gender show that in contrast to the total sample, marital status distribution of males is similar to that of the overall sample (Table 2). Examining female military personnel only, indicates that they have a higher prevalence of never being married compared with women in the general population (Table 3). This is consistent across all ages except in 18-29-year-olds, and is most notable in the 45-65 years age group. Consistent with the total and male samples, female military personnel are less likely to be divorced compared with the general population, except in 18-29-year-olds where military women are more likely to be divorced. The difference in prevalence of divorce between women in the military and those in the general population ages 18 to 29 years is larger than in the total and male samples, it is, however, still small.

Discussion
This article provides the first comparison of military and general population marital status distribution in the U.K. Consistent with hypothesis 1, military personnel are more likely to be married, especially those under the age of 30 years compared with the general population. Hypothesis 2 is also supported as military personnel under the age of 30 years are more likely to be divorced, compared with the general population; however, this increase is small. Older than the age of 30 years, military personnel are less likely to be divorced compared with the general population. Female military personnel are more likely to have never been married, except for 18-29 years who are more likely to be married, compared with women in the general population. These results are consistent with existing literature from the United States indicating that compared with civilians, military personnel are more likely to be married, marry at younger ages, divorce at younger ages, and military women are more likely to have difficulties forming and maintaining romantic relationships (Adler-Baeder et al., 2006;Cadigan, 2000;Karney et al., 2012).
The increased proportion of marriage in young military personnel, compared with the general population, is likely to be attributable to the job and financial security provided by a military career (Kelty & Segal, 2013). Literature from the general population shows that financial and job security is perceived as being important in decisions to marry (Smock, Manning, & Porter, 2005). In the United States, it is reported that the military offers young junior enlisted personnel higher pay rates and better benefits than other jobs available to age matched non-serving individuals (Kelty & Segal, 2013). Consequently, a military career may afford young personnel financial and job security which could inform decisions to marry.
Further to financial and job security having the potential to impact decisions to get married, benefits available to married military personnel in the U.K. may be attributable to the higher prevalence of marriage and the lower prevalence of divorce in the military compared with the general population. These benefits include entitlement to subsidized housing normally on or near the military base (Defence Infrastructure Organisation, 2013), subsidized boarding school places (Harvey, 2011), and in the event that a serving military member dies, the married spouse is entitled to the war widower pension; unmarried partners are not entitled to this compensation (gov.uk, 2013). Dandeker, Eversden, Birtles, and Wessely (2013) found that many wives of U.K. Armed Forces personnel reported "perks" that helped moderate the impact of military life, including tax breaks, quality of living, subsidized schooling, and improved social status. They also found that job stability, financial security, and a good pension at the end of service enhanced quality of life . These "perks" might help maintain relationship stability for military personnel.
Similar or lower rates of divorce in the military compared with the general population are reported in the United States (Burland & Lundquist, 2013;Karney et al., 2012). U.S. literature supports the idea that the support, benefits and compensations provided by the military, for married personnel, increase stability in marital relationships (Burland & Lundquist, 2013;Karney & Crown, 2011;Karney et al., 2012). Social exchange theory (Thibaut & Kelley, 1959) stipulates that decisions to start, continue and end relationships are based on the couple involved weighing up the perceived rewards and costs. Relationships are formed when both partners perceive the possible outcomes to be better than any alternatives (Thibaut & Kelley, 1959). Karney and Crown (2007) suggested social exchange theory can help understand military marriages in terms of a cost/benefits process where potential hardships are compensated for by benefits that strengthen and stabilize the relationship. Burland and Lundquist (2013) reported that, there is a premium for U.S. soldiers from disadvantaged backgrounds who in the civilian world would have been more vulnerable to marital dissolution without the financial and support benefits received as part of their military career. Based on social exchange theory, continuation of a relationship may be motivated by the material and supportive gains rather than satisfaction and happiness with the relationship.
Our results indicate that there were proportionally more divorced 18-29-year-olds in the military compared with the general population. This is consistent with research indicating that marriage at a younger age is associated with marital instability (Burland & Lundquist, 2013;Karney & Crown, 2011;McCone & O'Donnell, 2006;Wilson & Stuchbury, 2010). Lundquist (2007) suggested that the increased divorce rates in younger military personnel in the United States could be a    consequence of the stress of military life, particularly in the context of younger less experienced personnel and newer marital bonds that are likely to be less stable.
Military women are more likely to be never married than are women in the general population. This is consistent with findings from the United States (Adler-Baeder et al., 2006;Karney & Crown, 2011;Karney & Crown, 2007;Karney et al., 2012;Segal & Segal, 2004). Adler-Baeder and colleagues (2006) compared the marital status of women in the military and general population and found that military women were less likely to be married, more likely to be divorced, and less likely to remarry after a divorce, compared with civilian women. They suggested that this direct comparison is misleading as it does not consider differences in employment status. Comparisons with civilian career women and female military personnel showed little difference by marital status. Adler-Baeder et al. (2006) suggest that this is due to the increased role (marriage/work) conflict for career women, which may also be true for military women. This is consistent with Kelty and Segal (2013) who, on the basis of their research investigating gender differences of marital status within the military, propose that military service is more compatible with the husband/ father role, than with the wife/mother role. Breen and Cooke (2005) suggested that relationships where women have higher labour force participation are likely to have greater marital instability, possibly because of the women's decreased need for the husband's economic production or the competition for occupational status within the relationship.

Strengths and limitations
The main strength of this research is the use of a large representative sample of the U.K. Armed Forces. The main limitation of this study is that it categorizes cohabiting or in long-term relationships in the never married category along with singles. This categorization should be considered when interpreting the results. Moreover, marital status categories include those who are remarried in the married category. Including those who are remarried within the married category could overemphasize marital stability in either group.
Armed Forces personnel were included in the Office for National Statistics data, however, they only make up 1.5% of the England and Wales population (Office for National Statistics, 2013). A further consideration of the U.K. military sample is the potential for bias due to the main reasons for nonresponse of the cohort questionnaire (e.g. personnel being on training, deployments or being posted to a new location) are all factors that could be considered as associated with additional stressors for marital relationships.
Consideration should be given to the differences in marital status response categories between the Office for National Statistics general population data and the KCMHR U.K. military data. As stated in the methods section response categories in the military data were recategorized to attempt to closely replicate the marital status items in the ONS data. The difference in these categories should, however, be considered when interpreting results as they may not have the exact same meaning, especially the "never married" category.
Analysis of the statistical significance of the difference between marital status distributions in each group was not possible because of the nature of the two data sets. However, the proportions presented allow for a comparison that adds to the current literature where no such investigation has been previously conducted.

Implications
This research indicates that young military personnel are more likely to be married. Awareness of this amongst military welfare services could improve support for young married couples who might benefit from additional relationship advice.
This comparison suggests that female military personnel may have challenges forming and maintaining romantic relationships compared with women in the general population. Being in the military, for women, appears not to be conducive to successful relationships where role conflict between family and work are likely to be increased, compared with males. Further research investigating the workfamily conflict experienced by women in the military might be beneficial to help inform policy for how to lessen the effect of such work-life balance challenges.

Conclusion
U.K. media and colloquial beliefs present an image of the U.K. military as being a group with troubled marriages marked by high divorce rates. A comparison of the distribution of marital status between the general population and the military in the U.K. indicates that overall military personnel are in fact more likely to be married and less likely to be divorced. Military women and military personnel who marry younger than the age of 30 years are, however, more likely to experience marital dissolution compared with the general population.

Funding
The U.K. Ministry of Defence funded the KCMHR cohort study where the military data used in this research were sourced from. The general population data were sourced from the Office for National Statistics. This research was conducted as part of the corresponding author's PhD, which was funded by the Royal British Legion.