America wants women to stand on equal footing with men at work but also wants babies cared for by their mothers at least in the first weeks of their lives.

These sensible desires are sometimes hard to square with economic reality. Some people on the Left want to fix the problem by expanding government or suspending the obdurate realities of economics. But conservative scholars and lawmakers say the solution that actually works is not the creation of new federal programs but the loosening of restrictions on existing ones.

Sen. Marco Rubio, R-Fla., has embraced a proposal from Kristin Shapiro and the Independent Women's Forum, which would allow parents to use Social Security benefits for a brief paid parental leave. A new mother could draw benefits for a limited period and in exchange would delay her receipt of retirement checks by the same period of time.

The admittedly debatable premise is that society should provide for new mothers in the workforce. It’s incontrovertible that mothers and young babies benefit from spending time together in the earliest days of an infant's life. But any policy of mandatory employer-paid leave would prompt employers to hire fewer women of childbearing age. Any policy of government-paid leave would further strain the public finances.

President Trump promised universal paid leave in his State of the Union speech, and his daughter and most trusted adviser, Ivanka, holds it as her goal in life. A Democratic Congress, should one be elected in November, would be sure to pass something with costly mandates that over a couple of years grow into unsustainable government commitments. So Republicans and conservatives would be well served now to try to agree on a more practical plan, that pre-empts something more radical and ruinous. The plan by Shapiro and IWF is a good one, and Rubio is picking up the ball.

The Social Security-paid leave idea’s first virtue is that it doesn’t create new federal spending. The money would come out of Social Security funds today and reduce Social Security draws down the line. It leaves spending about the same in the long run.

The second advantage is that it doesn’t impose mandates on businesses, which are generally a bad idea because they discourage hiring. The cost of fulfilling mandates comes out of salary and other benefits.

The problem here isn’t that those people who don't have children pay for those who do. Society in general benefits from greater fertility and more births, if only to support a large older generation in retirement. The problem is that mandated benefits make employers less likely to give women jobs if they are of childbearing age. This also drives down women’s wages.

The Shapiro/IWF idea doesn’t stop employers from voluntarily offering additional paid leave. Those who couldn’t afford six- or eight-week paid benefits might manage a two- or four-week leave to supplement the Social Security benefit that is created.

The main argument against the proposal is that it would turn out to be a slippery slope. “There is really no way it would stay small and contained for very long,” conservative commentator Daniel Payne wrote on our op-ed pages, “if it even were small to begin with.”

This is not an unreasonable fear, but it’s also not convincing. It’s just as likely this policy could send us down the other side of the same slope, so that when lawmakers see how popular it is to free up Social Security benefits for maternity leave, they could allow still more people early access to their benefits for more reasons. Why not allow people to draw money to buy a house, for example? Maybe you could draw down some of your expected benefits today and invest them in a long-term fund for your retirement.

This proposal is the best we’ve seen for expanding paid leave to more parents. In the long run, it will stave off worse proposals and possibly open the door to better ones.

Correction: A previous version of this piece incorrectly solely credited Andrew Biggs of the American Enterprise Institute with the reform idea. Biggs helped run the numbers for the IWF proposal.