Paid family leave: Balancing benefits and costs

By Harry J. Holzer of Georgetown University and Brookings.

"A strong consensus is developing among politicians and policy analysts that there are important benefits to paid family leave, and that some such policies should be widely implemented at the federal and/or state levels. Indeed, both presidential candidates last fall embraced paid family leave as a goal, though with quite different specific proposals on how to implement it. A number of states have implemented paid family leave or soon plan to, and the Washington, DC Council passed a generous bill in December 2016 providing paid leave to all workers in the District (though the Mayor has not yet indicated whether she will sign it.

There is also a growing consensus on the range of benefits paid leave provides to working parents (especially women), other family members, and even employers to some extent. Specifically, women benefit by remaining more attached to the labor force generally and also to specific jobs when they take time from work; other family members, especially newborn children, can get the care and attention they need for their health and well-being; and employers benefit from less employee turnover and healthier workplaces.

But, as economists always emphasize, there are important costs as well as benefits associated with paid family leave, and policymakers should strive for appropriate balance between them. The real and potential costs include:

• Tax burdens on employers and employees;
• Potential impacts on federal or state budget; and
• Disruptions to the workplace that also burden employers.

Tax burdens and potential shortfalls

Most paid family leave plans are financed through some version of a payroll tax, rather than a simple mandate on employers. This is a largely sensible approach, as it spreads the costs of providing leave over employers and all employees. In most states that have implemented paid family leave to date, something like a 1 percentage point tax on payroll has been sufficient to fund the amount of leave taken by employees.  This extra 1% of tax, relative to the 15.3% federal social security taxes already levied  on payroll plus other taxes (e.g., for unemployment insurance), is not a terrible additional burden but also not trivial.

Regardless of whether the tax is levied directly on employers or workers, economists believe much of the cost is ultimately borne by workers either way, in tax payments or lower wages. On the other hand, employers will bear some burden in the cases of very low (or minimum) wage workers, whose wages cannot be further reduced; and they will usually perceive that they bear the cost of taxes they directly pay, in any event.

Will these taxes be sufficient to fully fund the amount of leave taken? This depends on the generosity of the benefits, in terms of the replacement rate for lost earnings (up to a usually specified cap), plus the maximum duration of the benefit. Very high replacement rates and caps will raise the take-up rate on the benefit, and also encourage people to take the maximum allowable time out of work (though we don’t really have evidence on how much).

Thus, the first states to provide paid leave – California, New Jersey, and Rhode Island – provided replacement rates of about 50-60% up to modest dollar caps for weekly earnings, for a total of 4-6 weeks. In contrast, the bill just approved by the DC Council (replacing an earlier one vetoed by the mayor) provides up to 8 weeks of parental leave, 6 weeks of other family leave and 2 weeks or personal sick leave; they voted a replacement rate of 90% up to $900 per week of salary and 50% beyond that level, up to a cap of $1,000 per week in total benefits. In the latter arrangement, workers will have very little incentive to limit the durations of leave they take below the maximum allowed, and so the costs of the system might well exceed the 0.62% additional payroll tax allotted to pay for leaves there.

Thus, the possibility of fiscal shortfalls in DC and elsewhere exists, when the political pressure to limit taxes brushes up against the desire for lengthy and generously paid leaves with high replacement rates up to high caps.

Workplace disruptions and costs

Employers will likely experience some disruptions in their workplaces when their employees take paid leave, especially if employees have the right to return to their previous positions. The disruptions will be greater the longer the spells last, particularly in smaller establishments; and employers might hire temps, whose wages or salaries they must pay (on top of payroll taxes) and whose work performance will vary in quality.

Of course, many employers voluntarily provide these benefits to their highly-educated employees, in which case they must believe the benefits to them outweigh the disruption costs. This occurs most frequently with professional or managerial workers who are highly valued and expected to remain with the employer for many years, despite a leave or two over these years. Whether employers feel the same way about paid leave for less-educated employees whose work they value less, whose expected tenure on the job they expect to be shorter, and whom they might expect to give birth more frequently (especially if they have stereotypes about birth rates among less-educated women in general and minority women in particular) is an open but important question.

Under these circumstances, a mandatory paid leave policy might well lead employers to begin discriminating in hiring against less-educated women in the child-bearing ages, especially minority women. Though we have no direct evidence of such discrimination, we have some indirect evidence from the Americans with Disabilities Act, which caused employers to hire the disabled somewhat less frequently than before (accordingly to at least one widely respected study).

The paid leave costs employers believe they will bear will also likely be high in states or cities that have already imposed other large costs on employers of less-educated workers – such as high minimum wages, paid sick leave, and restrictions on hiring practices. For example, Washington DC and several other cities and states have already decided to raise their minimum wages to up to $15 per hour (though in stages); several of them have also mandated several paid worker sick days a year, and have outlawed asking about criminal records in the application (“Ban the Box”) or screening for marijuana use.

One can debate the merits of each of these practices individually, and the evidence to date is limited. My own reading of the literature suggests that very high minimum wages, especially if indexed to inflation, could generate substantial employment reductions in the long run, and that “Ban the Box” tends to discourage the hiring of less-educated minority men, who are now costlier to screen for criminal records.

But I strongly believe that, collectively, their impacts on the hiring of less-educated workers could be much more negative than the sum of the individual effects. Economists believe that changes in hiring practices and workplace organization are costly, and employers will undertake them only when the benefits of doing so are substantial and long-lasting. Imposing a large number of permanent labor costs on employers will render the benefits of less hiring much more attractive – perhaps leading to “tipping points” in many establishments.

Furthermore, these hiring changes can be implemented gradually but permanently over time – by employers who move (or who simply open their newer establishments) more frequently in lower-cost nearby locales. This is a particular risk for Washington, DC, since Arlington, VA, is right across the river and has retained a $7.25 minimum wage and imposed no other burdens on hiring. Absent a moderate and sensible federal minimum wage increase or paid leave policy, the variations across states and localities in both will become very large, and might trigger employer and worker relocations in opposite directions.

Even employers who remain in the high-cost jurisdictions will also gradually reduce their hiring over time. Recent work by Isaac Sorkin of the Chicago Federal Reserve Bank shows that new, less labor-intensive establishments gradually replace older, more intensive ones after lasting minimum wage increases. The remaining establishments might also implement new technologies (such as robots) over time to perform simple tasks now performed by low-wage workers.

Thus, the tendency of some states and localities – like Washington, DC – to impose very generous paid family leave policies on top of a range of other large costs on employers of less-skilled workers – reinforces my concerns that these policies might have unintended consequences, such as less employment of the very people whom we intend to help. I still support paid family leave, even in such locations, but urge even greater caution there regarding its costs.

In conclusion… 

I strongly believe that paid family leave — especially parental leave for newborn infants — provides important benefits to workers, their families and even their employers. Providing paid leave with moderate replacement rates up to a cap, for limited durations of time (perhaps 2-3 months for new parents and less in other circumstances) makes good sense, as does paying for the leave with a modest increase in the payroll tax. This tax should be primarily on employees, and fully funded to avoid burdening the federal or state governments with new fiscal shortfalls.

But it is important to strike the right balance between these benefits and the costs imposed on other workers, employers and perhaps the governments in question. Failure to consider these costs might well create a range of serious problems – such as fiscal shortfalls and lower employment for less-educated younger women.

It would be a great shame if such policies harm the very people whom we intend to benefit from them."