Policy Corner: Connecticut FMLA Program Would Quickly Go Broke

We would all like to be able to take up to 12 weeks of paid time off every year if something bad happens to us, or to a family member, or if we are blessed with a child, or face any number of other stressful circumstances.  In fact, that’s exactly the problem – we would all like to do that!  And once we started paying a payroll tax to underwrite that benefit, many of us would do that.  The result would be ever-increasing payroll taxes and still a high likelihood of the system going broke.

These are some numbers we’ve been working through here at Yankee Institute, trying to figure out what is likely to happen in the real world if Connecticut approves a paid family and medical leave act (“FMLA”) in something like the form which is currently moving through the legislature.  

We conclude that it’s very unlikely that the fund will remain solvent for long at a 0.5 percent payroll tax.

It is more likely that the fund will require significant increases in the payroll tax while still either requiring considerable reductions in benefit generosity and/or going broke in relatively short order.  

Senate Bill 1 and House Bill 5003 would mandate a 0.5 percent payroll tax for all employees to fund a $1,000/week paid FMLA benefit for all workers for up to 12 weeks per year every year. The bill has already been passed out of committee.

The mean annual wage in Connecticut is somewhat less than $60,000/year.^[1]  Because that average includes all income, not just the amount below the payroll-tax threshold, using it for these purposes technically (even if only in the short term) overestimates the amount of revenue that will be generated by this tax.^[2]  Using it, nevertheless, for simplicity’s sake yields this result:  the average Connecticut worker will pay $300/year as a result of the initial 0.5 percent payroll tax.  

Assuming very conservative administrative costs for the program of 10 percent of collections and benefit payments, it will take 44 workers’ payroll taxes to support one worker’s full use of the FMLA benefit each year.

It follows, then, that if just 2.27 percent of workers take full advantage of paid FMLA – and no one else takes any FMLA time at all – the program will nevertheless go broke almost immediately at a 0.5 percent payroll tax and sink further into the red every year from there.

This means that the state will have to start hiking this payroll tax almost as soon as it starts paying benefits.  How high will it get?  That depends on how many workers take partial or full advantage of the FMLA benefit each year. 

It’s not clear what the initial use rate will be. We can expect the rate to be fairly high, however, because under the proposed legislation, the benefit would pay out 100 percent of a worker’s weekly wage,up to a total of $1,000 per week for 12 weeks per year. This higher payment will induce more people to use the benefit than otherwise would, which in turn will drive up the costs of the program, which in turn will drive up the payroll tax.

Consider this close comparison.  Many employers offer sick days. These sick days usually offer full compensation for the day(s) missed, but with a limit on use of much less than 12 weeks a year. Assume that the average employee has five sick days a year.  Now assume that the average employee uses, again on average, just one of those sick days every year. That represents a 20 percent use of sick days. 

If FMLA use rises to 20 percent, then the payroll tax will have to rise by nearly 9 times, to nearly 4.5 percent.  At 4.5 percent the average Connecticut worker would pay $2,700 per year in state FMLA payroll tax – more in FMLA payroll tax than in state income tax.

If FMLA use rises to 20 percent, then the payroll tax will have to rise by nearly 9 times, to nearly 4.5 percent.  At 4.5 percent the average Connecticut worker would pay $2,700 per year in state FMLA payroll tax – more in FMLA payroll tax than in state income tax.

There are a variety of reasons to believe that use won’t reach 20 percent, especially right away. But even very constrained use will require increases in the payroll tax or bankruptcy of the fund. And as the payroll tax rises, inhibitions against using the benefit – and a desire to get something back for those increasingly burdensome deductions – will rise. This suggests markedly increased use of the benefit over time, and a spiral of ever-increasing payroll deductions.

From this, we conclude that Connecticut’s paid FMLA benefit, as currently structured, will likely become unstable very quickly, resulting in ever-rising payroll taxes even as benefits are slashed and the program itself veers into bankruptcy. We have not seen any effort by the state of Connecticut realistically to demonstrate that this level of payroll tax can possibly support this level of benefit into a sustainable future.  If the state were to undertake such a calculation, we would need it to show its work carefully; the state hardly has an impressive track record of properly calculating the cost benefit programs and funding them accordingly. But in this instance, it appears not even to have tried.

That is not, we think, the best possible basis on which to establish a new government program. Connecticut state government is beset by fiscal emergencies and unpayable bills on all sides. Why on earth is it seriously considering setting up another expensive failure now?

^[1]See https://www.bls.gov/oes/current/oes_ct.htm.

^[2]That figure decreases somewhat because the current proposal would charge the payroll tax on income only up to the Social Security maximum, which is now (in 2019) $132,900.  See https://www.ssa.gov/policy/docs/quickfacts/prog_highlights/index.html.  Because that cap will be raised automatically upon fund shortfalls, however, we don’t expect it to remain in place for long.  See proposed S.B. 1 § 2(c)(1)-(3) (2019).  (The legislature has designed this legislation so that the base upon which the payroll tax will operate will rise automatically to meet the needs of the program, without the legislature having to vote for – and thus take responsibility for – those tax increases.  Instead, increases in the top amounts subject to the payroll tax will occur automatically upon a determination of the administrator of the program that such increases are necessary to the solvency of the program, unless the legislature votes down the increase by a three-fifths vote.  Because we expect this to happen almost immediately, we also expect that this initial exclusion is effectively meaningless once benefit distributions start.)