The Science of Aging Well: How Your Beliefs Affect Your Health
These findings were presented during September 2021, to the National Association of Supplemental Benefit Insurance 2021 Virtual Forum and the International Claim Association Virtual 2021 Annual Education Conference.
Background: Has the Covid-19 pandemic impacted work disability rates?
In a previous installment of “From The CDA Research Files” we concluded that evidence existed that Covid-19 substantially suppressed short-term disability incidence rates (at least in the first stages of the pandemic), and that a “snapback” of rates after the initial, sharp dip was rather slow in playing out. This phenomenon was fairly consistently observed throughout the disability insurance industry.
We looked for other data sources to provide more clarity about what’s been happening over the past eighteen months in terms of work-related disability. And that’s when we added the Current Population Survey to our data set.
About the Current Population Surveyi
One of the measures the CPS tracks is work absence (see details here). The survey asks people who are employed whether they were at work during the survey reference week (typically, the week including the 12th of the month– see details here), and if not, to select a reason for being absent. Here are the various absence reasons:
In our research, we’ve focused on absences for reason 06 (own illness/injury/medical problems). To be sure, there are some limitations to using this data.
- For one thing, it applies only to absences during the reference week of any given month.
- For another, it doesn’t make distinctions by length of absence. So it lumps together both “incidental” sick-day absences (a couple of days long) and disability leaves (lasting more than a week or two).
- And finally, it doesn’t provide detail as to the nature of the illness, injury, or medical problem that’s keeping someone from work.
That said, it’s still useful data to have if we want to get more insight into how the Covid-19 pandemic impacted people’s ability to work. So, we recently took a look at reason 06 work absences over the eighteen months from January 2020 through June 2021 and how they compared to a pre-2020 baseline.
Annualized absence rates: Pre-2020 baselineii
Here’s an average, by month, of annualized absence rate over the five-year period of 2015 thru 2019:
Each of the years in this five-year period pretty closely follows this pattern of seasonality: highest rates in Q1, then a decline in Q2 and Q3, with an uptick again in Q4. In my previous work at two different disability insurers, I had several opportunities to track monthly short-term disability incidence for several large clients, as well as significant subsets of the book of business. I can tell you that I invariably saw a seasonality pattern very similar to this one.
Here’s what happened to absence rates starting in 2020
Note how January and February of 2020 were in line with the baseline period and then rates skyrocketed. As of June 2021 (the end of our study period), they were down from the same month in the previous year and also at a 16-month low, but still above the average June level for the five-year baseline period.
This picture is quite different from what we’ve seen for short-term disability incidence, where rates declined sharply in the early stages of the pandemic and then never fully snapped back to previous levels. Why is that?
One hypothesis that I think has lots of merit: The CPS absences are overwhelmingly two weeks or less in duration. This reflects the impact of the Families First Coronavirus Response Act, which was a broad mandate for paid sick leave of up to two weeks in duration. While the mandatory aspect of FFCRA expired at the end of 2020, employers who choose to continue providing paid sick leave based on the FFCRA framework are eligible for tax credits in 2021. (And note: Based on a survey by the Society for Human Resources Management, almost three-fourths of employers are continuing to offer voluntary FFCRA-like paid sick leave.)
What do you, the readers of this article, think is the case here? We’d love to hear from you.
Coming up next….
In the next installment, I’ll share what we found when we sliced the absence data by occupation and by whether or not people were working from home. What we saw was interesting, to say the least. Stay tuned!
i The CPS is a monthly survey conducted by the U.S. Census Bureau on behalf of the Bureau of Labor Statistics. It provides extensive information about employment, unemployment, work hours, earnings, and demographic characteristics of people in the workforce and is considered the definitive source of labor force statistics for the U.S. population.
Data tables from each month’s survey are released by the BLS on the first Friday after the first Thursday of each month. Later in the month, associated microdata is updated to a database called IPUMS, which is housed at and managed by the University of Minnesota. We used this database to access CPS data.
ii For our analysis, we used a measure I’ll refer to as annualized absence rate. This can apply to a single month, multiple months, a single year, or multiple years. Regardless of the timeframe, the calculation of the metric is the same:
Numerator: Take the number of employed people reporting they were absent because of their own illness, injury, or medical condition in each month and sum up for all months.
Denominator: For each month, take the total number of employed† people for that month and divide by twelve. Then sum up the result for all months.
Result: Multiply by 100.
†There’s a little tweak to be made in arriving at this number. It involves excluding people reporting an absence for reason 15 (other). This is because of a “hiccup” in the way BLS reported pandemic-related job activity over several months in 2020. You can read the details here.
Related posts
