The aim of this study was to provide a descriptive analysis of a large sample of job vacancies in the current global health job market captured over a six-month period.
Study design
In this cross-sectional study, data was collected from global health job vacancies posted to web-based, open (non-subscription) job boards, employment email listservs for global health and development, and global health organizations and consulting firms over a six-month period from November 2015 to May 2016. Vacancy descriptions were analyzed to characterize the most common types, levels, and required skills and experience of positions in the global health job market.
Data collection
Job samples included in the master data set (S1) consisted of one posted job vacancy. A global health vacancy was defined as one soliciting applications for a paid non-internship position working with a project, program, or consultancy focused on a goal directly addressing human health. Positions with development programs not focused on health (e.g., agriculture, animal husbandry, microfinance, democracy/governance, and security), as well as positions only open to nationals from a particular country, were excluded. Both full- and part-time positions were included. Vacancies were gathered from twelve internet job boards, including Devex, DevNetJobs, Emory Public Health Employment Connection, Global Health Council, Idealist, Indeed.com (both the global health and international health job boards), Johns Hopkins Bloomberg JHSPHConnect, Peace and Collaborative Development Network, Public Health Institute newsletter, ReliefWeb, and USAJobs. Vacancies were also collected directly from the websites of 20 global health agencies and consulting firms, including Abt Associates, Camris International, CARE, Chemonics, Clinton Health Access Initiative, FHI 360, Gates Foundation, International Medical Corps, International Red Cross/Red Crescent, International Rescue Committee, John Snow International, Management Sciences for Health, Médecins Sans Frontières, PHI Global Health Fellows Program, Population Services International, RTI International, Samaritan’s Purse, Save the Children, Task Force for Global Health, U.S. Office of Foreign Disaster Assistance, World Health Organization, and World Vision. These job boards and employers were identified based on the authors’ collective knowledge of the global health industry and our 30 years of combined experience in global health job seeking. We collected 1254 total vacancy descriptions. Of these, 47 internships, unpaid (volunteer) positions, or positions for work not related to health were excluded. Of the remaining 1207 vacancies, 200 duplicate positions were eliminated, resulting in 1007 unique global health vacancies.
The data set for quantitative analysis (S2) was created by entering information on select characteristics of S1 into a spreadsheet. S2 data taken directly from each vacancy description included employer, position title, physical location of position, opening and closing dates (if applicable), education level required and preferred (for private- and nonprofit-sector vacancies), minimum number of years of overall professional experience and directly related experience required (for private- and nonprofit-sector vacancies), lowest and highest GS level (for US federal government vacancies), whether the position is term or permanent (for U.S. federal government vacancies), country or region of focus, whether proficiency in a second (non-English) language was required or preferred, up to two non-English languages desired, and whether overseas experience was required or preferred. For each vacancy, the two predominant areas of technical expertise required for the position were identified. After de-duplication, a unique alphanumeric ID was assigned, and private- and non-profit vacancies were classified by level as practicing clinician (clinical care or training responsibilities specified in vacancy announcement), entry-level (no more than three years of experience and no specialized technical knowledge or experience required), mid-level (two to five years of experience with some specialized technical knowledge or experience required), managerial (“manager” in position title and/or at least three years of project/program management experience required, plus financial, training, strategic planning, operations, or personnel management responsibilities specified in the vacancy announcement), technical/subject matter expert (at least two years of specialized experience required, plus consultative or training responsibilities, or management of technical aspects of projects/programs, specified in the announcement) or directorial (“director,” “chief of party,” “country representative,” or “president” in position title) according to job title, description, and the minimum number of years of experience required.
A qualitative data set (S3), consisting of samples with the text of each vacancy description (S1), was generated by simple randomized sampling of S1 without replacement, based on previously published sampling designs for mixed methods analyses [6,7,8]. To ensure an unbiased data set and sufficient sample size, 20% of the master dataset was used as the qualitative sample subset of S1. This had the effect of sampling from all six months and including samples from all sectors, which avoided both time- and sector-based biases that could have been introduced by a smaller subset.
Statistical analysis
S2 data was imported and descriptive analysis performed using SAS 9.4 (SAS Institute, Cary NC). All postings within the master data set were included in the analysis. Job and employer/agency counts, job levels, requirements and preferences for education level, non-English language proficiency and overseas experience requirements, geographic focus, top technical expertise areas, and physical location were calculated as frequencies and proportions. For each technical specialty area, vacancies with a focus in that area were compared to the remaining vacancies in the dataset that did not have that technical specialty as a focus of the job duties in order to determine whether those positions were more or less likely to require non-English language proficiency and overseas experience. The comparisons were assessed by calculating Chi-square statistics.
Thematic analysis
S3 data were imported into Dedoose Version 7.6.13 (SocioCultural Research Consultants LLC, Los Angeles, CA, USA) and analyzed using a thematic analysis approach. Content of the vacancies was classified either as a responsibility of the position or a desired qualification in the applicant. Position descriptions consisted of tasks, areas of responsibility, events, and deliverables that fell under the scope of the position as specified in the vacancy description. Candidate requirements included skill sets, experience with particular tasks or program areas, and individual character traits. A non-randomized sample of approximately 25% of the vacancies collected prior to February 2016 was used as a training and calibration set to develop a coding methodology and taxonomy for thematic analysis. This methodology was then used to analyze the S3 subset described above.