Sunday, September 9, 2012

The Costs of Mandatory Paid Sick Leave

As with any program, there will be costs involved with requiring companies to provide paid sick leave.  Rather than getting bogged down in the details (you can read the report for a blow-by-blow account), let me discuss the approach taken.  Most of it is pretty straightforward.

First, you need to estimate how many employees are not currently provided paid sick leave.  How is that done?  There are estimates of those without paid sick leave by industry nationally; take those numbers and apply them to the industrial make-up of Orange County.

Next, use an estimate of the number of sick days actually used.  A Bureau of Labor Statistics (BLS) report from Feb 2012 (an objective source) states that the average worker takes four sick days per year, so that was the number used.  This approach is in line with previous studies that are publicly available (i.e., assuming the same usage of paid sick days across industries).  Later, it was found that the BLS report also reported usage of paid sick days by industry, so an alternative estimate was calculed.  This provides an estimate of paid sick days (and thus hours) used by industry.  This resulted in a somewhat smaller estimated cost, so why use it?  To be as objective as possible.

Next, it was time to multiply the hours of paid sick leave by the average wage per hour by industry using offical data from the Florida Deoartment of Economic Opportunity.  Besides wages, businesses also pay payroll taxes, workmen's comp, and fringe benefits.  In addition, they also will face costs in terms of administering the program.  Rather than using my own estimate of all of these costs, I decided to use the one provided by the Institutie of Women's Policy Research (23% of the payroll, see report for source).

Finally, companies are likely to need to hire replacement workers or have other employees work overtime when people take paid sick leave.  How often would this occur?  This is where I could assume it occurred all the time, but instead assumed 10% of the time.  Why?  A IWPR study of the San Franscisco program stated that 9 out of 10 firms reported that they rarely or never hired replacements.  One can argue that they may have had to pay existing workers overtime,etc., so why not use a higher estimate?  A somewhat higher figure was easy to justify, but a specific number would be hard to justify.  Using an estimate of replacement workers from a proponent of mandatory paid sick leave makes it much easy to defend.

At this point, we have an estimate of the cost if all those currently without paid sick leave are covered by the proposal, but IWPR states that only 42% of those currently without paid sick leave in Orange County will gain access to paid sick leave under the proposal, so the estimated cost was reduced accordingly resulting in an estimated cost of $69.2 million per year if one uses BLS estimates of paid sick leave by industry or $82.3 million per year if one uses the BLS estimate of four days of paid sick leave used per worker.

Given that this study relied mainly on the approach used by IWPR, including estimates of costs, and used official government data that are publicly available, it's hard to claim that the costs were inflated.  In fact, they were underestimated.  In what way?  The figures presented don't include the administrative costs faced by those firms that already provide paid sick leave and must now show that they are in compliance nor the adminstrative costs of those that may need to demonstrate that they are not required to provide paid sick leave. In addition, since it uses the latest publicly available employment data for the county, it assumes no job growth.  Assuming that the economy does add jobs, the costs to businesses will increase.  Thus, the cost is likely to be higher than the estimates suggest.

Click here to return to the main post.