[The woman with 19 children is] an extreme case and that employers don’t need to really worry about this problem.

That’s true, employers don’t have to worry much about extreme cases, but the point is that employers will take into consideration the leave they reasonably anticipate an employee will use when hiring. I had assumed that having to pay salary while an employee took parental leave would be an additional cost that an employer would be wise to factor in.

If that were the case, a high rate of turnover would make the poor, young, female applicant even less attractive. People look after their own interests, and it’s only natural to expect that an employee would not quit a job just before qualifying for paid family leave, rather she would quit after receiving the benefit. So the responsible hiring manager would estimate the probability of the applicant contributing no value for some time, while getting paid, and then not returning. Clearly this adds costs for the business.

Your confident assurances of “possible” and “very small” negative effects for the business were not very convincing, but the citations you provided were much more helpful. Now I see that I was mistaken, the employer will likely be completely insulated from the costs of this benefit, since it is typically funded directly from employee salary via payroll tax. Mr. Condor has taken the trouble to examine the paper you mentioned about California’s Paid Family Leave, and he suggests that we do a cost/benefit analysis.

The paper indicates that California’s benefit plan is “funded entirely by an employee payroll tax” of 1.2%. The benefit is 55% of the employee’s usual weekly earnings, for up to six weeks, when they take time off from work to bond with a new child.

The plan also provides for paid leave to care for a seriously ill family member, up to twelve weeks. The employer can require that up to two of these weeks come from earned vacation time before paid family leave kicks in, but I will ignore this. Let’s assume employees always take the maximum benefit allowed.

California says 87% of recent claims were for newborn bonding, and the other 13% for sick family member care (189,317 and 26,496).

You mentioned that the average woman has 2.1 children. So that’s 2.1 × 6 weeks, about 13 weeks for newborn bonding. Bump it up to 17 weeks to include sick family member care.

The benefit is 55% of the employee’s usual salary, so the lifetime expected benefit for a typical California worker is about 9 weeks’ worth of salary (17 × 0.55).

That’s the benefit, what is the cost?

The payroll tax is 1.2%, so we can calculate how long the employee will work before contributing 9 weeks’ worth of salary, reaching the break-even point. It takes 9 weeks ÷ 0.012 = 750 weeks, less than 15 years. A typical California employee who works 15 years breaks even on paid family leave, and every year worked beyond 15 represents a net loss. If an employee has a 30-year career, which does not seem at all unlikely, she will pay twice as much as she receives in benefits. [UPDATE: it's not that bad]

Condor asks what seems to me an important question: “If offered the chance to opt out of the tax and forgo the benefit, how many employees would take that option?”

Unfortunately, employees are not offered this choice. But if they knew how badly they were being screwed by this rip-off wrapped in a humanatarian “Helps Working Families” label I think most of them would opt out.

This analysis is an estimate, and I would appreciate being made aware of any errors in my numbers. We should also consider features of the real world that will skew the numbers away from my estimate:

• Nobody has 2.1 kids, the real number will be higher or lower. I assume these will balance out, and using the average number of kids gives us the best estimate.

• Some of the average 2.1 children arrive as multiple births (1 in 30 infants born in the U.S. in 2009 was a twin). This reduces the benefit.

• Employees who don’t have kids or sick family members do not qualify. This reduces the benefit.

• Employees may choose not to take the leave even when they qualify (e.g. spouses who reckon that giving up 45% of their salary would be a hardship). This reduces the benefit.

• Employees do not have to take the full six or twelve weeks of leave. Employers can require that earned vacation time be used first. This reduces the benefit.

• I calculate the benefit in “weeks of salary.” Employees will tend to take the benefit when they are younger and their salary is lower. But they will pay the tax throughout their career, including the typical peak earning years that come later. This reduces the benefit.

So where does all that money taken from employees go? I wonder.

Do you still support this idea? b_b, you didn’t indicate how badly I distorted your position; I am curious to know what you think as well.

If a business can't handle the added expense of maternity leave, I would venture that that business is not socially useful enough to stay open.

This statement is so incredible it makes me wonder if you are taking the discussion seriously. Do you really believe that a hospital or bookstore or barbershop is not socially useful simply because it is struggling financially?