A fully competent public health workforce, a key component of the nation's public health infrastructure, has become the focus of increasing attention.1,2 More specifically, workforce shortages have been recently identified as one of the urgent priorities that need to be addressed by US public health agencies.3–5 The ratio of public health workforce to US population has decreased drastically in recent years, from 220 per 100 000 population in 1980 to 158 per 100 000 in 2000. Moreover, the projected need by the year 2020 has been estimated to be 220 per 100 000.6 Employee turnover—which includes voluntary separation by the employee (quits); involuntary separations initiated by the employer (layoffs and discharges); and other separations due to retirement, death, or disability7—constitutes one of the major challenges facing the public health workforce, as it can be expensive, result in loss of expertise and institutional knowledge, and negatively affect organizational performance.7–11
According to a 2007 survey conducted by the Association of State and Territorial Health Officials, the average state public health worker (PHW) is nearly 47 years old and, as of 2002, a quarter of the public health workforce was eligible for retirement.12 However, retirement accounts for only one-third of total turnover in state and local health departments, which has been estimated at 10%.7 Although this rate is lower than the estimated 16.4% turnover rate for local and state governments,13 rates among state health departments have recently increased over time: 10.8% in 2009, 11.4% in 2010, and 11.5% in 2011.7 Understanding the factors driving nonretirement workforce loss can be crucial in developing policies to increase retention and prevent the high costs of recruiting and training new staff. Factors that can influence turnover include organizational infrastructure, management, workplace environment, advancement opportunities, training, competitive pay, and employee morale.7,9,10
The Public Health Accreditation Board has developed several standards and measures relating to public health workforce development, calling upon health departments to focus more on their current workforce while supporting the development of future PHWs.14 In addition, Healthy People 2010 first incorporated public health infrastructure into the national statement of health objectives, demonstrating recognition that infrastructure gaps jeopardize the accomplishment of all other objectives.
Comprehensive data on the public health workforce are fundamental to workforce development throughout the public health system. Although the existence of a public health workforce shortage in the United States is generally acknowledged, it has been difficult to quantify, given numerous challenges, including the enumeration of the existing workforce. Estimates have ranged from 291 000 to 516 000, and a recent study calculated the local, state, and federal shares at 50%, 30%, and 20%, respectively.15–18
The objective of this study was to improve our understanding of the workforce turnover that public health agencies would face in the short and medium term, focusing separately on retirement and other voluntary departures. We use data from the Public Health Workforce Interests and Needs Survey (PH WINS), which yielded the first nationally representative sample of state health agency central office employees,19 to conduct the following tasks: (1) estimating cumulative planned retirement among PHWs over the 2015-2019 period, by job classification; (2) estimating the rate of intended voluntary departure for reasons other than retirement, by job classification; (3) estimating the association between the intention to leave the public health workforce and job and pay satisfaction; and (4) assessing the main determinants of job and pay satisfaction. Results of our analysis will address gaps in our understanding of areas within public health that are more likely to face turnover issues and will provide insights into the factors driving turnover and potential solutions.
Study sample and measures
The sampling methodology and characteristics of PH WINS participants are described elsewhere in this supplement of the journal.19
PH WINS participants were asked whether they planned to retire before 2020 and, if so, in what year. Respondents were also asked whether they intended to leave their current organization within the next year and, if so, for what purpose. We used these questions to create a categorical variable of mutually exclusive voluntary departure intentions as follows: (1) if a respondent indicated the intention to retire within the next year or in any year before 2020, we classified them as Retiring by 2020; (2) among remaining participants, if they expressed the intent to leave their current organization to seek another job within public health, seek another job outside of public health, or leave for some other reason, we classified them as Leaving Within Public Health, Leaving Outside of Public Health, or Leaving for Other Reason, accordingly; and (3) all remaining participants were classified as Staying. In addition, for regression analyses described later, we created a binary indicator of Leaving, equal to 1 for all participants in the categories Leaving Within Public Health, Leaving Outside of Public Health, or Leaving for Other Reason, and “0” otherwise.
To assess job satisfaction, we used the Job in General (JIG) score, a validated measure calculated from responses to 9 questions about a worker's opinion about his or her job.20 The JIG score has a range of 0 to 48, where higher scores indicate higher job satisfaction. Although there are not exact cutoffs to assess whether a worker is satisfied with his or her job, according to the literature on this measure, scores of 29 or higher can be considered as “satisfied.”21 Therefore, we created a binary variable for job satisfaction, equal to 1 if the JIG score was 29 or higher and “0” otherwise. Participants were also asked directly whether they were satisfied with their pay and job security, using 5-point Likert scales ranging from “very dissatisfied” to “very satisfied.” Similarly to job satisfaction, we created binary measures of pay satisfaction and job security satisfaction by classifying as “satisfied” those participants who responded “somewhat satisfied” or “very satisfied.”
Descriptive analysis of turnover intentions
To assess whether public health agencies are facing higher voluntary turnover due to retirement among certain types of workers, we estimated cumulative intended retirement rates from 2015 to 2019 for the following 9 different job classifications: administration or management, health education, laboratory, information technology, social services, health services, environmental health, epidemiology, and other (eg, professionals and scientists, statisticians, students, engineers, veterinarians, and program managers). Similarly, we estimated 1-year rates of intended voluntary departure for each of the 9 job classifications listed earlier, identifying whether the purpose of departure was to leave for another job within public health, outside of public health, or for some other reason.
Association between voluntary departure and job satisfaction measures
We examined whether JIG satisfaction, pay satisfaction, and job security satisfaction are important determinants of intended voluntary departure by estimating linear probability regression models of a worker's intention to leave the workforce for any reason other than retirement. To understand the cumulative contribution of individual and job characteristics to explaining a worker's decision to leave, we estimated 4 models, each building upon the previous one. Model 1 included only demographic characteristics (gender, race/ethnicity, and age); model 2 added socioeconomic characteristics (education, years of experience, and supervisor status) and census region; model 3 added job classification; and model 4 added the binary job satisfaction measures described earlier.
Determinants of job satisfaction and pay satisfaction
Since PH WINS does not capture organizational-level information that would allow us to examine factors determining job security satisfaction, we focused on exploring the determinants of JIG satisfaction and pay satisfaction. First, to examine whether pay satisfaction can be attributed to perceived consistency between a worker's actual salary and his or her expected salary, given his or her education and years of experience, we estimated a linear probability regression model of pay satisfaction as a function of education and years of experience in public health, controlling for salary levels. If workers' salaries do not progress according to workers' expectations, we would expect to see lower satisfaction with higher education or higher experience, once the effect of absolute salary levels is removed.
Second, we hypothesized that JIG satisfaction is determined by workplace characteristics, such as relationships with supervisors, workplace environment, and employee motivation/morale. We explored the evidence for this hypothesis by estimating a linear probability regression model of job satisfaction (the binary JIG measure) as a function of 20 dichotomous variables measuring workers' opinions of their job environment.
All analyses were conducted using complex survey design procedures in SAS 9.4 and Stata/MP 13.
One in 4 PHWs reported plans to retire before 2020, and an additional 18% reported their intention to leave their current organization for reasons other than retirement, within 1 year of participating in the survey, nearly evenly distributed among those who intend to seek employment within public health (5.9%), outside public health (5.2%), and those leaving for other reasons (6.5%) (Table 1). There was significant heterogeneity in these intentions by demographic, socioeconomic, and job characteristics. Not surprisingly, our descriptive analysis found that older and more experienced workers were significantly more likely to plan to retire. Those more likely to report the intention to leave for reasons other than retirement included workers younger than 41 years, workers of African American, Hispanic, or other race/ethnicity, or those with salaries less than $55 000 per year with a master's degree with 10 or fewer years of experience in public health and those not satisfied with their pay or their JIG.
Rates of planned retirement varied significantly by job classification (Figure 1). Nearly one-third of health services workers (oral health, nutritionists, and clinical services), as opposed to only 13.0% of epidemiologists, reported that they plan to retire within the next 5 years. Other job classifications with high rates of planned retirement included administration and management (26.5%), health education (26.0%), and environmental health (28.7%). Similarly, there was significant heterogeneity in intended voluntary departure within 1 year, for reasons other than retirement (Figure 2). The highest rates were among health educators (25.2%) and epidemiologists (21.6%), although most in the latter group reported intentions to seek jobs within public health. Among those planning to leave, those most likely to seek jobs outside of public health were workers in information technology (6.5%) and social services (6.6%).
Determinants of intentions to leave
Results of the linear probability regression models of workers' intentions to leave their current job are shown in Table 2. By far, pay satisfaction and job satisfaction were the most important variables explaining variation in workers' intentions to leave, evidenced by the R2 statistic, which more than doubled (from 0.08 to 0.18) when adding these variables. These coefficients indicate that intention to leave decreases by 24 percentage points for workers satisfied with their JIG (P < .001), by 9 percentage points for workers satisfied with their pay (P < .001), and by 5 percentage points for those satisfied with their job security (P < .001). Projections generated using these results indicate that, if all workers were satisfied with their JIG and pay, the rate of intended departure would be 7.4%, less than half the current rate of 17.7%
Throughout all models, there were no significant differences in intentions to leave between men and women. In contrast, workers older than 41 years were significantly less likely to intend to leave than younger workers, and those differences remained throughout all models. Among socioeconomic characteristics, workers with a master's degree were 5 percentage points more likely to report their intention to leave their current job than those without a college degree and workers with more than 20 years of experience in public health were 8 percentage points less likely to intend to leave than workers with 5 or fewer years of experience. Job classification did not appear to be a significant determinant of workers' intentions to leave their job.
Determinants of pay satisfaction and job satisfaction
Results of our regression models of determinants of pay satisfaction and job satisfaction are shown in Table 3. Consistent with our hypothesis, higher education was associated with a lower probability of workers being satisfied with their pay, with the exception of MDs who, although less satisfied than workers without a college degree, were significantly more likely to be satisfied with their pay than workers with doctoral degrees (P = .02, not shown in the table). Similarly, more years of experience in public health were associated with lower pay satisfaction, although the magnitude of this association was not as large as that estimated for education.
Nearly all workplace characteristics included in our regression model were significantly associated to job satisfaction. The strongest relationships were estimated for variables in the Employee Motivation/Morale category; in particular, 3 variables were associated with increases of 10 percentage points or larger in the probability of being satisfied: I am satisfied that I have the opportunities to apply my talents and expertise, I am inspired to meet my goals at work, and I feel completely involved in my work. Other important predictors of job satisfaction were feeling that supervisors treat employees with respect (P = .001), work well with employees of different backgrounds (P < .001), or support employee development (P < .001). Variables found not to have a significant association with job satisfaction included those related to work-life balance, opportunity to demonstrate leadership skills, and learning from coworkers.
We analyzed turnover intentions reported in PH WINS. Our main findings were as follows: (1) more than 40% of US PHWs in state agencies either have plans to retire by 2020 or intend to leave their current organization within 1 year; (2) there is significant heterogeneity in turnover intentions by job classification; (3) pay and job satisfaction are the most important predictors of intentions to leave other than retirement; and (4) we identified several workplace characteristics significantly associated with worker satisfaction.
An effective public health system relies on the capacity of its workforce. Therefore, understanding the factors driving nonretirement workforce is key to an effective public health agency. Employee turnover can be expensive, and the loss of experienced personnel can prevent a health department's ability to respond rapidly and quickly to public health needs.2,7,9 Our study answers recent calls for research on factors that contribute to voluntary turnover; more specifically, what type of PHW is likely to leave voluntarily, reasons for departure, and what public health agencies can do to increase retention.
Our analysis revealed that the areas of administration and management, health services (physicians, nurses, physician assistants, etc), and environmental health are most likely to experience significant turnover due to retirement, whereas health education and epidemiology were the areas with the highest reported intentions to leave for other reasons. Of particular concern are the high rates of intended voluntary departure among young workers, who might be dissuaded by lower compensation in governmental public health than in the private sector, making it difficult to attract and retain new talent to the field.4
It is important to highlight that the type of turnover has different implications for public health agencies. High retirement rates can be addressed by ensuring that workers with relevant skills are being recruited, either right out of school or through targeted efforts for positions that require more experience. On the contrary, high rates of intentions to leave for reasons unrelated to retirement (eg, to seek another job) indicate that agencies have difficulty keeping those they already employ. This cannot be addressed merely by improving recruitment, but rather will require efforts to improve retention.
Not surprisingly, our analysis of voluntary turnover found that job in general and, to a lesser extent, pay satisfaction are the most important predictors of workers' intentions to leave their current job. In fact, our results suggest that if all workers were satisfied with their pay and job in general, the rate of intended departure would be 7.4%, less than half the reported rate of 17.7%. Studies in the psychology literature have found that for workers, pay-level satisfaction is raised when employees invest more mental energy into their enriched jobs and therefore expect higher levels of pay to compensate for their greater levels of responsibility.22
Also not surprising was our finding that more highly educated workers and those with longer work experience in public health are less likely to be satisfied with their pay. One exception to this pattern were physicians, who were more likely than those with PhDs, and just as likely as those with master's degrees, to be satisfied with their pay. This result suggests the existence of features in pay structures or promotion ladders that are specific to individuals with medical degrees. Although more research is needed to better understand these patterns and their impacts, our results suggest that public health agencies may need to establish pay structures that ensure that highly educated and highly experienced workers are compensated according to their expectations, especially if such work opportunities are available elsewhere.
Perhaps, the biggest insight derived from this study is our analysis of predictors of JIG satisfaction. Analyses shown in Appendix Table A1 suggest that education and salary are not important determinants of JIG satisfaction. Rather, as shown in Table 3, JIG satisfaction appears to be mostly a function of workplace characteristics, many of which are amenable to intervention by public health agencies. Examples of efforts that could increase JIG satisfaction, and thus reduce intentions to leave, include improving relationships between employees and supervisors, avoiding excessive workloads, improving communication between senior leadership and the general workforce, ensuring workers can apply their skills in their regular work, and improving workers' perceptions regarding the importance of their own work and how it contributes to the agency's goals. A recent survey conducted by the School of Public Health at the University of Illinois and the Center for State and Local Government Excellence provided 6 case studies of organizations that have been successful in recruiting and retaining their employees. One of them identified several key strategies to increase employee retention, which included providing a positive work environment (recognition of high-performing employees, flexible work schedules, holding periodic staff retreats), offering competitive pay, matching competitive salary offers, and encouraging and developing employee talent.11
Two important limitations of our study are the use of cross-sectional data and the lack of a causal design. Therefore, we cannot be certain that the relationships found in Table 3 are causal and that addressing factors related to those constructs would lead to higher satisfaction in pay or JIG and, in turn, to lower rates of voluntary departure. Studies using longitudinal data or exploiting natural experiments would provide more solid evidence on this respect. In addition, we did not directly test our hypothesis that pay satisfaction is a function of consistency between a worker's actual and expected pay. Further research using more appropriate data should explore this issue and should try to identify the size of discrepancies at different levels of education and experience. Finally, because data are self-reports of intentions to leave, there could be significant differences between reported intentions and actual departures, and our estimates could be biased if certain groups tend to over- or underreport their true intentions more than others.
PH WINS provides a significant opportunity to improve our understanding of voluntary turnover among the public health workforce. Insights derived from this study suggest that public health employers should focus their retention efforts on the areas of health education and epidemiology, on workers younger than 41 years, on those who make less than $55 000 a year, and on those with a master's level degree. Although pay satisfaction is a significant determinant of an employee's intention to leave, we identify several additional factors that agencies can target to reduce voluntary turnover, including improving relationships with supervisors and coworkers, preventing excessive workloads, and improving employee morale.
1. Tilson H, Gebbie KM. The public health workforce. Annu Rev Public Health. 2004;25:341–356.
2. Perlino C. The Public Health Workforce Shortage: Left Unchecked, Will We Be Protected? Washington, DC: American Public Health Association; 2006. APHA Issue Brief.
3. Hilliard TM, Boulton ML. Public health workforce research in review: a 25-year retrospective. Am J Prev Med. 2012;42(5)(suppl 1):S17–S28.
4. Beck AJ, Boulton ML, Lemmings J, Clayton JL. Challenges to recruitment and retention of the state health department epidemiology workforce. Am J Prev Med. 2012;42(1):76–80.
5. Drehobl PA, Roush SW, Stover BH, Koo D. Public health surveillance workforce of the future. MMWR Surveill Summ. 2012;61:25–29.
6. Rosenstock L, Silver GB, Helsing K, et al. Confronting the public health workforce crisis: ASPH statement on the public health workforce. Public Health Rep. 2008;123(3):395–398.
7. Newman SJ, Ye J, Leep CJ. Workforce turnover at local health departments: nature, characteristics, and implications. Am J Prev Med. 2014;47(5)(suppl 3):S337–S343.
8. Cho S, Johanson MM, Guchait P. Employees intent to leave: a comparison of determinants of intent to leave versus intent to stay. Int J Hosp Manag. 2009;28(3):374–381.
9. Price JL. Reflections on the determinants of voluntary turnover. Int J Manpower. 2001;22(7):600–624.
10. Drehobl P, Stover BH, Koo D. On the road to a stronger public health workforce: visual tools to address complex challenges. Am J Prev Med. 2014;47(5):S280–S285.
11. Darnell J, Cahn S, Turnock B, Becker C, Franzel J, Wagner DM. Local Health Department Workforce Recruitment and Retention: Challenges and Opportunities. Washington, DC: Center for State and Local Government Excellence; 2013.
12. The Association of State and Territorial Health Officials. ASTHO 2007 State Public Health Workforce Survey Results. Arlington, VA: The Association of State and Territorial Health Officials; 2007.
13. Hathaway K. Job openings continue to grow in 2012, hires and separations less so. In: Job Openings and Labor Turnover Survey (JOLTS) Annual Story. Washington, DC: US Department of Labor; 2013.
14. Bender KW, Kronstadt JL, Wilcox R, Tilson HH. Public health accreditation addresses issues facing the public health workforce. Am J Prev Med. 2014;47(5):S346–S351.
15. Gebbie KM. The Public Health Work Force: Enumeration 2000. Washington, DC: US Department of Health and Human Services, Health Resources and Services Administration, Bureau of Health Professions, National Center for Health Workforce Information and Analysis; 2000.
16. Center of Excellence in Public Health Workforce Studies. Public Health Workforce Enumeration, 2012. Ann Arbor, MI: University of Michigan Center of Excellence in Public Health Workforce Studies; 2013.
17. Hayes D. Insights in public health: strengthening the epidemiology workforce through mentorship: practicum and fellowship experiences in the Family Health Services Division at the Hawaii Department of Health. Hawai'i J Med Public Health. 2014;73(3):94.
18. Beck AJ, Boulton ML, Coronado F. Enumeration of the governmental public health workforce, 2014. Am J Prev Med. 2014;47(5):S306–S313.
19. Leider JP, Bharthapudi K, Pineau V, Liu L, Harper E. The methods behind PH WINS. J Public Health Manag Pract. 2015.
20. Ironson GH, Smith PC, Brannick MT, Gibson WM, Paul KB. Construction of a Job in General scale: a comparison of global composite and specific measures. J Appl Psychol. 1989;74(2):193–200.
21. Association of State and Territorial Health Officials. Sampling Methodology Report: 2014 Public Health Workforce Interests and Needs Survey (PH WINS). Arlington, VA: Association of State and Territorial Health Officials; 2015.
22. Williams ML, McDaniel MA, Nguyen NT. A meta-analysis of the antecedents and consequences of pay level satisfaction. J Appl Psychol. 2006;91(2):392–413.
job satisfaction; public health workforce; recruitment; retirement; workforce turnover