Paid leave advocates claim that the current paid leave proposal is universal, but it isn’t really. In order to be eligible for benefits, new parents need to have been working in the 3 to 6 months prior to birth. This work history requirement will probably exclude around one in three new mothers (I, II), including new parents who are still in education, who have a work-limiting disability, or who suffer an unluckily-timed spell of unemployment.

When I make this point, some advocates respond by saying I am being ridiculous. They say that it’s no scandal that a paid leave program only covers people who were previously working because that’s the whole point of it: to compensate people who are taking leave from their job. Below, I explain why this response is wrong and why including all new parents in a parental leave program makes sense.

1. Other countries do it.

For starters, the idea that parental leave programs should cover all parents is not something I invented. It’s pretty normal across the developed world.

For example, Finland, Norway, Sweden, and Germany all provide parental leave benefits for all new parents, even those who have never worked a day in their life. This is accomplished by establishing a minimum parental leave benefit and declaring that all new parents are at least eligible for the minimum. I advocate doing this exact same thing here.

2. It follows from the parental leave policy rationale.

There are two ways to think about parental leave benefits.

The first way is to think about it as a child care benefit similar to home child care allowance benefits that some countries pay to parents who choose to care for their own young kids rather than use child care services. Under this rationale, clearly all parents should be included in the parental leave scheme because all parents are providing child care to their newborn.

The second way is to think about it as a benefit for someone who is currently facing a work limitation, similar to unemployment or disability benefits. This rationale also points to including all parents because, even if a parent was not previously working, the presence of a newborn that they must care for prevents them from working right now.

When it comes to the second rationale, it seems like some advocates vaguely gesture towards the idea that the benefit should only flow to people who would be working if not for the newborn. The reason providing parental leave to all parents makes them uneasy is because they think to themselves “such and such person does not need a benefit because they would not have been working either way.”

The problem with this approach to the second rationale is that you actually have no idea whether someone would have been working in the counterfactual scenario. Whether they were working 3 to 6 months prior to giving birth certainly does not tell you whether they would have been working in the counterfactual scenario.

Some people who were working 3 to 6 months ago would not be working right now regardless of whether they had a child and yet they are eligible for parental leave benefits under the Democratic proposal. Likewise some people who were not working 3 to 6 months ago would be working if not for a child and yet they are ineligible for parental leave benefits under the Democratic proposal.

This is especially true in the case of pregnancy where employers often discriminatorily fire and discriminatorily refuse to hire visibly pregnant women. The fact that you weren’t working 3 months prior to giving birth may have nothing to do with whether you are the kind of person who would otherwise be working and everything to do with the fact that employers refuse to hire you because you are pregnant.

The proper approach to the second rationale should be to realize that the presence of the newborn creates a work limitation and to pay every parent in that situation. Trying to figure out whether someone’s worklesness in that period is overdetermined by other factors is bad on the merits and impossible to do.

3. It is a better failure outcome.

The public debate about welfare benefits is dominated by magical thinking about perfect welfare administration. In reality, benefit administrators — both public and private — make mistakes and especially struggle with complicated benefits that have lots of parameters and requires lots of proof from users.

In the case of parental leave, even the most simply-designed benefits are going to run into problems where eligible people are unable to prove their work history or prove their prior earnings level. A system that excludes all parents who do not meet the work history requirement would end up paying $0 to these parents even though they are eligible for the benefit. A system that makes all new parents eligible for at least a decent minimum benefit would end up paying these parents the minimum benefit.

Both outcomes are a failure. But one failure outcome is much better than the other.

4. Other benefits do work sort of like this.

I have seen some advocates acknowledge the wisdom of including all parents in the paid leave scheme but then say sort of fatalistically that this is America and this is not how we design benefits here: we don’t provide cash to anyone if they haven’t put in work. But as bad as the American welfare state is, this is not entirely true.

Elderly and disabled people who do not meet the Social Security work history requirement are eligible for Supplemental Security Income (SSI). Elderly people who do not meet the Medicare work history requirement are eligible for a collection of programs called the Medicare Savings Programs (MSP) that will cover their Medicare premiums. Survivor’s insurance used to be backed by AFDC (now TANF).

SSI, MSP, and AFDC function as the minimum benefits that are made available to people who do not meet the work history requirements for the higher earnings-related programs. These programs tend to be heavily income-tested and asset-tested, which is not good, but they do exist.

The only major program that has never had a minimum benefit backstopping the main one is Unemployment Insurance (UI). But there has been a recent push to plug that hole by creating a jobseeker allowance that would operate as a flat minimum benefit for unemployed people who do not meet the work history requirements of UI. Indeed, many of the same organizations and people who I’ve found to be dismissive of my point about parental leave are the ones pushing the jobseeker allowance!

In general, the idea that there is some time-immemorial American Way on welfare seems wrong and unnecessarily defeatist. The welfare state will be what we make of it and the policy organizations that craft these bad designs are making it badly and then pretending like they can’t do otherwise.