JOURNAL15.pdf | essaymoose

�

Succession Planning in Local Health Departments: Results From a National Survey

Julie S. Darnell, PhD, MHSA; Richard T. Campbell, PhD � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �

Context: Succession planning has received scant attention in the public health sector, despite its potential to generate

operational efficiencies in a sector facing chronic budgetary

pressures and an aging workforce. Objectives: We examined the extent to which local health departments (LHDs) are engaged

in succession planning and assessed the factors associated with

having a succession plan. Design: We conducted a national cross-sectional Web-based survey of workforce recruitment

and retention activities in a sample of LHDs responding to the

National Association of County & City Health Officials’ 2010

Profile Study and then linked these data sets to fit a multivariable

logistic regression model to explain why some LHDs have

succession plans and others do not. Setting and Participants: Top executives in a national sample of LHDs. Main Outcome Measure: Presence or absence of succession planning. Results: Two hundred twenty-five LHDs responded to the survey, yielding a 43.3% response rate, but no statistically

significant differences between respondents and

nonrespondents were detected. Only 39.5% reported having a

succession plan. Performance evaluation activities are more

common in LHDs with a succession plan than in LHDs without a

plan. In adjusted analyses, the largest LHDs were 7 times more

likely to have a succession plan than the smallest. Compared

with state-governed LHDs, locally governed LHDs were 3.5 times

more likely, and shared governance LHDs were 6 times more

likely, to have a succession plan. Every additional year of

experience by the top executive was associated with a 5%

increase in the odds of having a succession plan. Local health

departments that report high levels of concern about retaining

staff (vs low concern) had 2.5 times higher adjusted odds of

having a succession plan. Conclusions: This study provides the

first national data on succession planning in LHDs and sheds

light on LHDs’ readiness to meet the workforce-related

accreditation standards.

KEY WORDS: local health departments, succession planning, workforce development

Succession planning has received scant attention in the public health sector, although such activity is commonplace in the private sector and becom-ing increasingly prevalent in other kinds of public agencies.1-8 Consequently, very little is known about succession planning in public health departments be-yond accounts in Ohio,9 Washington,10 and Wyoming.11

This lack of attention is surprising, since local health departments (LHDs), like other sectors of the econ-omy, are faced with an aging workforce. In addition, increasing demand for LHDs’ services against a backdrop of chronic budgetary pressure means that LHDs need to generate operational effciencies—and avoid operational ineffciencies—whenever possible. Here, succession planning can be a success factor, as

Author Affiliation: School of Public Health, University of Illinois at Chicago (Drs Darnell and Campbell).

This project was supported by a grant from the Robert Wood Johnson Founda-tion. The authors would like to thank UIC colleagues Susan Cahn, DrPH candidate, for her help in developing and administering the survey, Clinical Professor Barney Turnock, MD, for advice on the project and for reviewing multiple drafts of the manuscript, and Yuanbo Song, PhD candidate, for generating the sample. They thank Joshua Franzel and his team at the Center for State and Local Government Excellence for feedback on the survey and early drafts of survey findings. They wish to thank the National Association of County and City Health Officials, espe-cially Carolyn Leep, for detailed comments on the survey. David Jemielity edited the manuscript. Finally, the authors are grateful to the local health department officials who filled out and returned the survey.

The authors declare no conflicts of interest.

Correspondence: Julie S. Darnell, PhD, MHSA, School of Public Health, University of Illinois at Chicago, 1603 W. Taylor Street, Room 758, Chicago, IL 60612 ([email protected]).

DOI: 10.1097/PHH.0000000000000120

141

142 ❘ Journal of Public Health Management and Practice

its potential benefts include “effciency drivers” such as improved operational continuity, organizational stability, systematic development of leadership com-petencies, preservation of institutional knowledge, fnancial stability, increased innovation, and decreased recruitment and orientation costs.12-16 In light of both the workforce challenges and the possible benefts of succession planning, it is important to assess the extent to which LHDs are engaged in succession planning and understand why some LHDs have implemented succession planning while others have not.

This study reports the frst national data about suc-cession planning in LHDs. It is also the frst to identify factors that are associated with succession planning. An understanding of the current magnitude of succession planning in LHDs across the country, the kinds of ac-tivities being undertaken, potential shortcomings of the processes currently in place, and the aspects of LHDs that make succession planning more or less likely ad-vances the dialogue concerning how to maintain a trained and competent public health workforce and sheds light on the readiness of LHDs to meet the workforce-related accreditation standards set forth by the Public Health Accreditation Board.

● Methods

Research design and data sources

We conducted a national cross-sectional survey of workforce recruitment and retention activities in a probability sample of LHDs and then linked the work-force survey data with the National Association of County & City Health Offcials’ (NACCHO’s) 2010 Pro-fle Study in a multivariate logistic regression model to explain why some LHDs have succession plans and others do not.

The sample for the workforce survey was drawn from the 2107 LHDs that responded to the NACCHO 2010 Profle Study.17 An LHD is defned as “an adminis-trative or service unit of local or state government, con-cerned with health, and carrying some responsibility for the health of a jurisdiction smaller than the state.”17

We stratifed the LHD population by jurisdiction size, governance structure, and type of services offered and then sampled units proportional to size within stratum, sampling the 38 LHDs serving populations of 1 million or more with certainty. This methodology yielded an initial sample of 601 LHDs. In cases in which multi-ple LHDs shared the same “top executive” (defned as the highest ranking employee with administrative and managerial authority or someone in a key lead-ership position who is familiar with human resources management practices), 1 LHD was drawn using sim-

ple random sampling, reducing the sample by 64 to 537.

We designed a Web-based questionnaire using Qualtrics software (Qualtrics, LLC, Provo, Utah). We developed drafts of the survey instrument on the ba-sis of questions from existing succession planning sur-veys found during our review of the literature.9 , 10 , 18 We shared drafts of the instrument with the 10-member NACCHO Workforce Committee and discussed it on one occasion via conference call. Written feedback was solicited from the Committee members and compiled by a NACCHO staff member. Local health department top executives and human service professionals se-lected by NACCHO and working in 7 LHDs of varying sizes, geographic locations, and governance structures pretested a revised draft instrument. The pilot partic-ipants completed the online survey and then partic-ipated in one-on-one cognitive interviews conducted by telephone to assess the respondent burden and the clarity, relevance, and comprehensiveness of the items. Each participant received a $100 gift card. In addition to current LHD offcials, 7 former health department off-cials who previously occupied leadership and manage-ment positions in state and local government settings but now served in faculty or staff positions in a school of public health provided written feedback on various drafts. In all, 22 current and former LHD offcials and 2 senior staff members at NACCHO reviewed various drafts. We incorporated the comments from these con-tent experts into a draft that was assessed formally at the pretest and fnal stages by the 3-member Question-naire Review Committee of the Survey Research Lab-oratory at the University of Illinois at Chicago. These technical experts reviewed the survey to ensure that ad-vanced principles of questionnaire construction were followed. Succession planning was one of the major topics covered by the workforce survey (n = 22 items). The fnal 44-item survey consisted predominantly of closed-ended items.

We administered the fnal survey in the feld from November 19, 2012 until December 31, 2012. The LHD’s “top executive” or someone in a key leadership posi-tion who is familiar with human resources manage-ment practices was invited to participate in the survey. We contacted all respondents at least 3 times and some up to 8 times by mixed methods, including regular and express mail, electronic mail, and telephone. All mate-rials noted that NACCHO endorsed participation. At-tempts to reach 19 LHDs were unsuccessful because of undeliverable addresses or invalid contact names, decreasing the fnal sample to 518 LHDs. We linked responses to the survey with data from the NACCHO 2010 Profle Study. The study protocol was approved by the Institutional Review Board at the University of Illinois at Chicago.

Succession Planning in Local Health Departments ❘ 143

Measures

Succession planning is defned as “intentionally iden-tifying, developing, and retaining individuals for fu-ture management and leadership roles.” Local health departments that have implemented “formal” suc-cession planning have a written document that es-tablishes a formal process for identifying, develop-ing, and retaining staff for future management and leadership positions. Local health departments en-gaged in “informal” succession planning use an infor-mal process for identifying, developing, and retaining staff for future management and leadership positions, without a written document or formal process.

The binary dependent variable used in the multivari-able model was presence or absence of succession plan-ning. Formal succession planning and informal succes-sion planning were combined into a single category. Local health departments that reported that their suc-cession plan had not been implemented (ie, inactive) were treated as not having a succession plan.

The explanatory variables included in the model re-fected characteristics that would be expected to in-fuence workforce policies and activities.10 , 15 , 19 We in-cluded 2 variables that denote the size and complexity of the organization: total population served and scope of services offered. Two variables represented the con-text for decision making: governance classifcation and having a board of health. Because previous research on succession planning in public health departments9 , 10

suggests that the top executive is vital to its successful implementation, we added a variable measuring the number of years that the top executive has served in his or her position. Finally, we included a variable measur-ing the level of concern about retaining well-qualifed staff.

All of the explanatory variables originate from the NACCHO 2010 Profle Study except for 2 variables: years that the top executive has served in his or her position and the level of concern about retaining well-qualifed staff, which were taken from the recruit-ment and retention survey. We collapsed total popula-tion served into 4 categories: small (<50 000), medium (50 000-499 999), large (500 000-999 999), and very large (1 million+). We defned scope of services as having no clinical services, home health only, or at least 1 clinical service. We used NACCHO’s classifcation of an LHD’s governance structure: a unit of state government, unit of local government, or unit governed by both state and local authorities. We categorized LHDs as having a board of health (yes/no). We added the number of years that the top executive has served in his/her po-sition as a continuous variable in 1-year increments. The assumption of a linear gradient was evaluated visually and judged not to be violated. We collapsed the response categories for the question about the

level of concern about retaining staff into dichotomous groups (no concern/slightly concerned vs moderately/ very/extremely concerned).

Statistical analyses

The unit of analysis is the LHD. We incorporated sampling weights to account for nonresponse and the complex survey design. To assess the potential for nonresponse bias, we modeled multiple predictors of response using logistic regression. We calculated bi-variate χ 2 tests with design-based F statistics to obtain unadjusted comparisons of LHDs with and without succession planning according to selected organi-zational characteristics and performance evaluation activities.

After conducting these crude analyses in which each independent variable was evaluated separately for its association with the outcome variable, we estimated a multivariable logistic regression model to explore the organizational factors that were associated with succession planning. Odds ratios were adjusted for the 6 factors hypothesized to be related to succession planning: jurisdiction size, governance classifcation, board of health, scope of services, tenure of top executive, and level of concern about retaining staff. All statistical tests were 2-tailed, and a P value of .05 was used to determine statistical signifcance. Standard errors were corrected to account for the complex survey design and adjusted for the clustering of LHDs within states. Statistical calculations were performed with Stata software (version 11.2-SE; Stata Corporation, College Station, Texas).

Model ft was checked in various ways. We assessed the model for specifcation error and found no evidence for either an omitted relevant variable or incorrect link function. We assessed goodness of ft using the Pear-son χ 2 goodness-of-ft test. Collinearity among the in-dependent variables was examined by calculating vari-ance infation factors.

● Results

Study sample

Two hundred twenty-fve LHDs responded to the sur-vey, yielding a 43.3% response rate. Because we have access to data describing the LHDs, we were able to estimate a multivariable logistic regression model with geographic region, population size of the jurisdiction, governance structure, and type of services as predictors of response. We found that none was an independent predictor of nonresponse, controlling for all else.

The characteristics of the sample of 225 LHDs are shown in Table 1. Most LHDs serve jurisdictions with small populations of less than 50 000. Most do not

144 ❘ Journal of Public Health Management and Practice

TABLE 1 ● Characteristics of Responding Local Health Departments � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � Characteristic Weighted %

Succession planning (n = 225) Formal/informal and active 39.5 None or inactive 60.5

Population size (n = 225) <50 000 58.8 50 000-499 999 35.2 500-999 999 4.2 1 million + 1.8

Scope of services (n = 225) No clinical services 60.8 Home health only 19.6 At least 1 clinical service 19.6

Governance classification (n = 225) Unit of state government 15.2 Unit of local government 73.1 Shared state/local government 11.7

Board of health (n = 224) Yes 76.9 No 23.1

Geographic region (n = 225) Northeast 15.6 Midwest 37.7 South 33.1 West 13.6

Mean (standard error) tenure of top executive 9.5 (.61) (in years) (n = 224)

Level of concern recruiting staff (n = 220) Not at all 6.2 Slightly 11.2 Moderately 23.3 Very 35.9 Extremely 23.4

Level of concern retaining staff (n = 221) Not at all 6.4 Slightly 15.3 Moderately 16.2 Very 35.8 Extremely 26.3

provide any clinical services (eg, substance abuse, men-tal health, comprehensive primary care). A majority are locally governed. Two-thirds are located in the South and the Midwest. The mean number of years that the responding top executive has served in this position is almost 10.

Succession planning amidst staff turnover

Local health departments replace key leadership/ management positions on a regular basis. The median number of years since the responding LHDs flled their

last key leadership/management position was only 1.3 years.

Local health department respondents reported high levels of concern about their ability to recruit and re-tain well-qualifed individuals. Nearly 60% of LHDs reported being very or extremely concerned about fnd-ing well-qualifed staff, while 62% reported similar lev-els of concern about retaining staff. The concern is widespread among LHDs of all sizes, geographic loca-tions, governance classifcations, and service offerings.

Despite the frequent turnover of key staff and high levels of concern about recruiting and retaining well-qualifed individuals, only about 40% of LHDs reported being engaged in either formal or informal succession planning. The remaining 60% said that they do not intentionally identify, develop, and retain individuals for future management and leadership roles.

Table 2 reports the unadjusted results comparing LHDs with and without succession planning by orga-nizational characteristics and selected workforce con-ditions. We fnd statistically signifcant differences for population size and governance classifcations. Also, compared with LHDs without a succession plan, a higher proportion of LHDs with a succession plan operate under a requirement to conduct an external search, although this difference fails to reach statisti-cal signifcance at P value of less than .05. We detected no statistically signifcant differences between LHDs with and without a succession plan according to their perceptions about the necessity to conduct an external search when there are strong candidates to replace the current top executive. Table 3 describes the frequency and scope of succession planning and career develop-ment activities in the subset of LHDs that said that they have a succession plan (n = 98). It shows the ex-tent to which LHDs with succession plans are engaged in typical succession planning and career development activities, and whether they had implemented these ac-tivities department-wide or only in selected divisions or programs. On the whole, the succession planning activities are targeted to specifc divisions or programs within the LHD rather than implemented throughout the LHD. The succession planning activities that were most broadly used were identifying the leadership po-sitions for which succession planning would be use-ful and identifying the competency, skills, and success factors for key leadership positions. Succession plan-ning activities that were much less commonly imple-mented department-wide were prioritizing positions in terms of impact or vacancy risk. Nearly one-third did not evaluate how leadership positions impacted the LHD’s strategic goals. Nearly all LHDs involve at least a portion of their staff in cross-functional projects, and just more than half use job rotations to some extent. Individualized development plans are reportedly not

Succession Planning in Local Health Departments ❘ 145

TABLE 2 ● Characteristics of Local Health Departments With and Without a Succession Plan (n = 225) � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �

Formal/Informal Does Not Have a Succession Plan Succession Plan Design-Based

Characteristic (Weighted %) (Weighted %) F Test, P

Population size (n = 225) .030 <50 000 49.7 64.8 50 000-499 999 41.6 31.1 500-999 999 5.1 3.6 1 million+ 3.7 .6

Scope of services (n = 225) .151 No clinical services 51.2 65.1 Home health only 19.7 19.6 At least 1 clinical service 26.1 15.3

Governance classification (n = 225) .008 Unit of state government 8.5 19.6 Unit of local government 73.8 72.6 Shared state/local government 17.7 7.8

Board of health (n = 224) .691 No 24.5 22.1 Yes 75.5 77.9

Geographic region (n = 225) .529 Northeast 19.6 13.0 Midwest 39.2 36.7 South 28.5 36.1 West 12.7 14.3

Mean (standard error) tenure of top executive (in years) 10.7 (1.1) 8.7 (.72) .108 (n = 224)

Level of concern retaining staff (n = 221) .113 Not at all/slightly 15.6 25.8 Moderately/very/extremely 84.4 74.2

Perceptions about the necessity to conduct an external .701 search for the top executive (n = 219) Not at all necessary 18.3 25.1 Slightly necessary 16.7 18.1 Moderately necessary 24.0 25.5 Very necessary 27.2 19.6 Extremely necessary 13.9 11.7

Requirementa to conduct an external search for the top 25.7 14.5 .065 executive (n = 207)

aRequirement: Law, regulation, rule, or policy.

used throughout the agency when individuals assume new positions. A minority of LHDs use stretch projects or 360-degree feedback assessments.

Most LHDs take a reactive rather than proactive approach to flling key leadership and management positions. Among LHDs that reported having either a formal or informal succession plan, 17.1% said that succession planning is done only after the current leader/manager has announced departure plans. About half (48.5%) reported doing succession plan-ning before a vacancy is anticipated but acknowledged that such planning is not ongoing. Only 34.4% said that succession planning is an ongoing activity.

The majority of LHDs (63.3%) with formal or infor-mal succession plans rate their succession planning as “average,” with only 11% rating their plans as “good” and more an one-quarter (26.2%) rating themselves as “poor.”

Performance evaluation activities in LHDs with and without a succession plan

Table 4 describes performance evaluation activities in all responding LHDs and compares the extent of these activities in LHDs with and without a succession plan. Our fndings show that a variety of performance evaluation activities are more common in LHDs with

146 ❘ Journal of Public Health Management and Practice

TABLE 3 ● Succession Planning and Career Development Activities in Local Health Departments That Have a Formal or Informal Succession Plan (n = 98) � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �

Department-Wide In Select Divisions Not At All Activity (Weighted %) (Weighted %) (Weighted %)

Succession planning Identify leadership positions for which succession planning would be useful (n = 93) 53.0 36.8 10.3 Evaluate impact of leadership positions on goals (n = 94) 44.4 23.9 31.7 Prioritize leadership positions by impact (n = 91) 39.8 25.6 34.7 Prioritize leadership positions by vacancy risk (n = 93) 36.9 28.6 34.5 Identify competency, skills, and success factors for key leadership positions (n = 90) 52.7 24.8 22.6 Assess current bench strength (ie, who is ready now for leadership positions) and 42.3 35.6 22.0

identify any skills gaps (n = 91) Determine whether critical leadership positions have a successor (n = 91) 37.0 37.4 25.6

Career development Job rotations in various parts of the local health department (n = 93) 14.5 37.8 47.8 Involvement in cross-functional projects, task forces, or teams (n = 93) 50.4 45.7 3.9 External management/leadership trainings (n = 91) 42.8 38.2 19.0 Internal workshops and courses designed to develop leaders (n = 93) 40.8 29.8 29.4 360◦ feedback assessment to identify areas of growth (n = 88) 19.5 15.4 65.1 Individualized development plans when individuals assume new positions (n = 91) 38.2 24.5 37.3 Stretch projects (n = 76) 25.5 14.5 60.0

a succession plan than in LHDs without a succession plan. Local health departments with a succession plan differed signifcantly from LHDs without a succession plan in terms of the occurrence and breadth of the following performance evaluation activities: (1) managers discuss future career opportunities with em-ployees; (2) managers discuss individual performance as a barrier to career mobility; (3) managers assist employees in setting appropriate career goals and in setting clear and specifc objectives to reach goals; and (4) the leadership team continuously monitors the effectiveness of efforts to strengthen the talent pool. The magnitude of the differences in some cases is huge: twice the proportion of LHDs with succession planning compared with LHDs without a succession plan re-ported that managers discuss individual performance as a barrier to career movement on a department-wide basis (60.4% vs 30.9%). Similar orders of magnitude were observed for discussing career development op-portunities (63.3% vs 33.5%). No statistically signifcant differences between LHDs with and without succes-sion planning were detected for biannual follow-up of career goals (P = .089) or using performance review data to guide workforce development (P = .371).

Local health department characteristics associated with succession planning

As shown in the multivariable logistic regression anal-ysis presented in Table 5, the largest LHDs are 7 times more likely to have a succession plan than the

smallest LHDs, adjusting for all else. Local health departments governed by local authorities are about 3.5 times more likely than state-governed LHDs to re-port having a succession plan, and LHDs governed by both were nearly 6 times more likely to report hav-ing a succession plan. Having a board of health is not associated with having a succession plan. For ev-ery additional year of experience by the top executive, we expect to see a 5% increase in the odds of having a succession plan, holding all other factors constant. Local health departments that report being moder-ately/very/extremely concerned about retaining well-qualifed staff have 2.5 times higher adjusted odds of having a succession plan than LHDs that reported be-ing slightly or not at all concerned about retaining staff.

● Discussion

The study provides the frst national portrait of succes-sion planning activities in LHDs. Very few LHDs have formal processes and written documents guiding their succession planning efforts, despite a clear need to re-place key leadership positions regularly. Most LHDs that are actively engaged in succession planning char-acterized their approach as “informal.” A majority of LHDs reported doing nothing to intentionally identify, develop, and retain individuals for future management and leadership roles or have succession plans that are not being implemented.

Succession Planning in Local Health Departments ❘ 147

TABLE 4 ● Performance Evaluation Activities in Local Health Departments With and Without a Succession Plan (n = 225) � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � Performance Formal/Informal Does Not Have a Evaluation Succession Plan Succession Plan Design-Based Activity (Weighted %) (Weighted %) F Test, P

Managers discuss future career development opportunities with each individual as part of their performance review (n = 218)