Using NFIB Economist’s Estimates on Paid Sick Days: It’s Not Cricket
As a kid living near Manchester in the north of England, my first love was cricket. The sport (it is a sport) comes up nowadays when I use the phrase “it’s not cricket” — as in, it’s not acceptable, it’s not done.
In a report circulated to Philadelphia City Council and the media (but not online that I can find), Dr. William Dunkelberg estimated the cost to employers of enacting paid sick days legislation in Philadelphia. I’m going to use this blog post to try to persuade readers that, even if they oppose paid sick days, they shouldn’t use the Dunkelberg estimates because, well, “It’s not cricket.” The estimates are so transparently inflated that folks who live in a fact-based world shouldn’t use them.
Dr. Dunkelberg, the Chief Economist of the National Federation of Independent Business, conducted an analysis of implementing paid sick days, concluding that it would cost $350 million to $752 million to implement, and would reduce employment by 4,000 jobs.
So what’s wrong with this estimate? (This post draws from a more extended critique of Dunkelberg online.)
The most basic mistake is that Dr. Dunkelberg double counts the maximum cost of paid sick days. He assumes that workers will take all their legally permitted paid sick days each year, costing $350 million. He then says that some of the absent workers will be temporarily replaced. The maximum cost of this, if every worker is replaced, would be another $350 million. So that gets you to $700 million. Add $52 million in compliance costs at businesses that already have paid sick days and you get Dunkelberg’s $752 million figure.
But wait a minute. If workers aren’t sick, they get their job done at a cost to the employer of $350 million. If those workers are out sick and temporary replacements are hired, the work still gets done but somehow it costs the employer an additional $700 million? Wrong. The additional cost is $350 million — $700 million minus $350 million equals $350 million. I keep asking myself, am I missing something here? He can’t possibly have made this kind of mistake, can he? He can and he did.
Beyond this, consider two assumptions that drive Dr. Dunkelberg’s high-cost estimate.
First, he assumes that all workers take all of their sick leave. Evidence from national surveys and San Francisco indicates that people actually take a third to 40% of their permitted leave. Many workers view paid sick days as insurance — to be saved up in case they are needed, not used as personal days or extra vacation. Using the 40% figure, $350 million becomes $140 million.
Second, let’s consider how much employers actually hire replacements. Data from San Francisco indicate that employers do so less than 10% of the time. That means the additional, out-of-pocket, labor costs (for employers currently without paid sick days) fall to less than $14 million, a far cry from $700 million.
To be fair, there will be some lost productivity when workers are not replaced. It’s hard to say how large this will be. Many workers who are occasionally out sick “get their work done” anyway. (I don’t notice my work disappearing when I’m out.) Where that is not possible (e.g., nursing home care, customer service jobs, hotel housekeeping), other workers may pick up the slack. Based on this, the $14 million might climb to somewhere between $50 million and $100 million.
But wait, we haven’t even considered yet a series of positive benefits from paid sick days:
- Reduced turnover and recruitment and training costs
- Improved worker-supervisor relations and higher levels of work effort and commitment as workers’ reciprocate for paid sick days
- Reduced health problems due to contagion of other workers and of customers or clients
When you take all of these factors into account, there is a solid analytical and empirical reason for believing that implementing paid sick days would pay for itself — or better.
Bottom line, when carefully scrutinized, William Dunkelberg’s analysis of the costs of the proposed Philadelphia paid sick days ordinance is simply not credible. Regardless of whether you think paid sick leave is a good idea or not, if you agree with me about the Dunkelberg study, I hope you won’t use it. Because using something you know is wrong, “that’s not cricket.”
For another take on why advanced labor standards such as paid sick leave can be good for the economy, check out www.thirdandstate.org next week for a follow up blog post.