New Workforce Estimates Show Public Health Never Recovered from the Great Recession. Then Came COVID-19.
It’s 10 pm. Do you know where your public health workforce is?
These next 6 to 8 weeks are likely to represent the midnight hour of the first wave of COVID-19. Models differ, significantly, but as states begin to reopen, we know that thousands of more deaths are on the way. When this pandemic started as several outbreaks across the US, there was significant focus on the population-based prevention efforts that might contain or mitigate the spread of COVID-19, and the historic lack of investment in public health to date. Since then, there’s been a pretty substantial and understandable shift in the media to focusing on hospitals – the new front for COVID-19. As some media have pointed out, key workers at hospitals are among the many taking risk to help protect us all. There are those in the private sector, and many in the public sector as well, working across public transit, social services, and other governmental functions. All are critical for social function. One of those is public health. It’s hard to overstate the importance of governmental public health during a pandemic. Unfortunately, the workforce that drives its successes, and capacity, is in trouble.
The de Beaumont Foundation just released a graphic with the latest numbers on the state and local government and public health workforce, derived from ASTHO and NACCHO respectively.1, 2 I was fortunate to have helped develop some of these estimates, which show that the state workforce is now at 94,000 staff, down from over 108,000 prior to the Great Recession in 2008. Local Health Departments (LHDs) are a bit worse off, down to 153,000 staff from 184,000 prior to the Great Recession. Once you account for the fact that some state health agency staff actually work in local departments and handle the overlap between ASTHO and NACCHO estimates, the total number of staff in the state and local governmental public health workforce sits right around 200,000. This is down by 40,000-50,000 staff since before the Great Recession. While the public sector has mostly recovered in terms of staff and FTE, public health never has. We’ve seen what that’s meant for the early days of COVID-19. And I fear for what comes next in, and after, the pandemic. Even with extended disasters, recovery tends to take longer than the event itself. And public health will be expected to be there to help pick up the pieces, conduct massive contact tracing to help society “return to normal,” prepare for the next pandemic, and do all the normal critical things it does during a time of truly tremendous need, all while having a diminished workforce. It’s not clear to me that that is a sustainable proposition.
Incidentally, I find the idea of temporarily scaling up a public health workforce to needed levels of contact tracing a challenging proposition. Last month, a leading group called for perhaps more than 100,000 staff, or a 50% increase on the size of the current workforce, and former CDC Director Tom Frieden said 300,000 community-based contact tracers would be needed to be in line with Wuhan. I’d find more plausible an increase in temporary workers alongside a substantial and sustained increase in the overall public health workforce coupled with a hopefully-not-draconian leveraging of technology to scale contact tracing to a national level for extended COVID-19 response. But reasonable people can reasonably disagree on the right approach here – I digress. Perhaps we can all agree that a sustained surge capacity is just implausible with a chronically undersized and underfunded workforce.
I got a request recently about placing the size and scale of the public health workforce in historical context; I thought I’d share the results with you all. There are few people more prolific in this regard than Gebbie and Tilson’s work on enumeration in the earlier part of the millennium, as well as Beck’s enumeration work from a few years back. Some of these works include historical numbers saying that we had 500,000 people working in public health in the 80s. For these kinds of analyses, it is really hard to know what the data are actually based on, and whether we’re comparing apples-to-apples (spoilers: I think probably not).
So I thought I’d take us on a little journey into one of the most significant sources of these workforce data. Let’s start with a side quest in Tableau, using Government Employment & Payroll from the US Census. Below you’ll find a fairly crude visualization of the governmental “health” workforce at the state and local level in 2017. To make these estimates, the US Census collects self-reported data from over 90,000 governments in the US. As I’ve written about before, one challenge here is that their idea of health is a little bit different than mine, and perhaps yours. They separate public sector employment in “health” into hospitals, and then everything else. That “everything else” is where public health lives, but also behavioral health, outpatient clinical care, and the like. I’ve always been pretty suspicious of these workforce numbers and haven’t used them too much in my own research because I find they have poor correlation against other data out there, like from ASTHO and NACCHO. This map can help illustrate why that is, and will also allow you to explore public sector employment in your own county or city. Neat.
When you click on a county, it’ll take you to a table that shows all the governments in the county and the number of FTE they are reporting. I show you this not just because it’s fun, nor because it’s always nice to better understand your own government, but again to reinforce the idea that definitions matter.
We’re not capturing the public health workforce in these official government data – we’re just not. And it’s a decentralized problem – the Census is getting data from local governments who themselves are often not counting their public health workers or misclassifying them as something else. And some state governments employ their LHD staff directly, so that’s often not reported to the counties. Some states and localities do a great job reporting and some don’t, though it’s really hard to know which is which. All of this means that membership organizations, like ASTHO and NACCHO, are critical to understanding the size of our workforce. I am very appreciative that they collected those data and that membership answers those questions – because if they didn’t do it, we would literally have no idea of the size of our workforce.
As we move into the 2020s, no doubt the economic and health consequences of COVID-19 will be felt for years. Hopefully, we as a public health community can convince the public and policymakers that public health matters, so that 2030 sees a sustainably enlarged, well-retained, and well-trained public health workforce ready to face the challenges of the next decade.
 In the interest of disclosure, I will note that I have done contracted work with ASTHO in the past, and at present with NACCHO and de Beaumont. I helped develop the 2019 NACCHO workforce estimates and developed the combined staffing estimates for de Beaumont.
 Credit is due to the ASTHO R&E Team (Maggie Carlin, Alannah Kittle, Cara Person, Kristi Meadows) and NACCHO R&E Team (Aaron Alford, Debra Dekker, Karla Feeser, Kellie Hall, Shaunna Newton) for the data sources used to make the composite estimate.
 I constructed these combined estimates by adding ASTHO State Health Agency – Central Office staff estimates and NACCHO LHD staff estimates to avoid double counting where State Health Agencies employ staff at LHDs (who are captured both by ASTHO and NACCHO Profiles), and trying to correct for illogical entries and differential agency response over time, eg, where FTEs exceeded staff or where an umbrella agency reported Medicaid + Public Health in one year and just Public Health in the next. This resulted in 246,000 staff in 2008 (±10,000 LHD staff); 226,000 in 2010 (±6,000 LHD staff); 202,000 in 2013 (±5,000 LHD staff); 203,000 in 2016 (±6,000 staff); and 206,000 in 2019 (±13,000 staff). There is uncertainty in the NACCHO data due to non-response adjustment, as well as potential for data reporting issues. Those estimates use non-response adjustment with post-stratification and finite population correction based on agency size to create the 95% confidence intervals. There are no official uncertainty estimates for Central Office staff, as those are reported as point estimates to ASTHO. However, there remains the potential for data reporting issues.
Read All Posts in this Series:
- When the Going Gets Tough, the Tough Must Be Ethical
- What’s the Deal with Public Health Funding?
- A Workforce in Transition
- Meditations on the MPH, Part 2
- Meditations on the MPH, Part 1
- Jonathon P. (JP) Leider, PhD, is an independent consultant in the public health and health policy space, as well as a Senior Lecturer at the University of Minnesota and Associate Faculty at the Johns Hopkins Bloomberg School of Public Health. He has active projects and collaborations with foundations, national public health organizations, public health researchers and academics, and public health practitioners. His current projects focus on public health systems, the public health workforce, and public health finance. He holds a PhD in Health Policy and Management from the Johns Hopkins Bloomberg School of Public Health.
- Big Cities Health Coalition2021.06.30How Health Departments Are Addressing Substance Use Disorder and Overdose During a Pandemic
- Announcements2021.06.21AcademyHealth Call for Nominations
- Healthy People 20302021.06.16Podcast: Law and Policy as Tools in Healthy People 2030
- HRSA's Investment in Public Health2021.05.18Video Q&A — Preventive Medicine for Rural America: Why More Training Programs Must Be Here