Factors Impacting Employee Turnover in the Childcare Industry
Factors driving employee turnover in the childcare industry. Wages and payor type impacts turnover while benefits do not.
Employee turnover has been a persistent issue in the childcare industry for decades and has become even more so through the pandemic. This is important to study given the societal implications for a strong early childhood education foundation from both a childhood development standpoint (research shows quality early ed programming gives children a bump up in societal and educational success) and workforce participation standpoint (so dual income families can go to work). Research published in March 2021 by Regional Education Laboratory Program (REL) studied this issue with statistical rigor.
First, wage levels was the number one factor that correlated with turnover. My first thought was whether this would hold when controlling for regional cost of living; upon further reading, the relationship remained statistically significant when controlled for this factors. This is not very surprising and would point to a seemingly convenient solution by simply increasing wages. The problem is that childcare centers (especially smaller ones) are generally operating at a low margin and cannot afford many wage increases as labor is often 70%+ of the total costs of the operation. Providers can pass along the increased wages in the form of higher prices to parents, but parents can’t afford to pay more either as childcare tuition for just one child is 15%+ of their household income on average. This may speak to a somewhat broken free market model and would speak to higher government subsidies required. From personal experience, I would dispute this deduction partially because childcare businesses can operate at a higher margin if run well, professionally, which many are not.
Second, private pay centers (vs. fully government funded) programs exhibited higher turnover. This was a surprising finding given private pay centers are usually associated with higher demographics and higher paid teachers. I theorized that this relationship was not properly controlled. Fully government funded programs tended to be heavily weighted towards programs serving only 3-5 year olds, which exhibit higher pay and lower turnover. This is because more credentialed and qualified staff tend to teach 3-5 year olds, and the center can focus on a fully educational program vs. caregiving. Upon further reading, the research stated that once controlled for this factor, the relationship was no longer statistically significant. The problem with simply creating more 3-5 year old only government funded programs is that the 3-5 year old age group subsidizes the caregiving for the 0-2 year old age group (given lower staff-to-child ratios), and centers cannot remain financially viable serving only the 0-2 year old group. Thus, regions where 3-5 year old public programs are aggressively implemented have seen a catastrophic impact to the supply of 0-2 year old programs (e.g. New York City).
Lastly, nonwage benefits such as health insurance, retirement benefits and time off for professional development were unrelated to turnover. This is not that surprising since low wage workforce can access basic insurance through the ACA exchange and may not value health or retirement benefits.
By Charles Kim
https://ies.ed.gov/ncee/rel/regions/northeast/pdf/REL_2021069.pdf
Hi Charles! I found this post really thought provoking – using big data to understand what predicts worker turnover, especially in low-wage contexts, is much needed. And what I particularly love about your post is that it indicates the need to have both big data AND people with expertise in the field interpreting and reacting to that data. It was fascinating to read how you reacted to the findings and what you thought the authors should have included or explored. I found the lack of relationship between turnover and retirement/time off also surprising. I wonder if part of it is about labor market options? Like maybe workers know they wouldn’t find these benefits in other similar jobs, so it doesn’t influence their turnover intent. And/or perhaps it’s because so few workers actually have access to these benefits, so there’s no way to analyze whether these factors predict turnover?